Pre-Normative Work on Sampling and Testing of Solid ...virtualmaze.co.in/sample/Biofuels Info/Rhombus Power - biodiesel... · 11 USTUTT Institute of Process Engineering and ... 12

FINAL TECHNICAL REPORT

CONTRACT N° : ENK6-CT-2001-00556

ACRONYM : BioNorm

TITLE : Pre-Normative Work on Sampling and Testing of Solid Biofuels for the Development of Quality

Assurance Systems PROJECT CO-ORDINATION : Prof. Dr.-Ing. Martin Kaltschmitt

Dr. Michaela Hein

Dipl.-Ing. Franziska Müller-Langer

Institute for Energy and Environment gGmbH

Torgauerstr. 116

D-04347 Leipzig

Germany PARTNERS :

1 IE Institute for Energy and Environment gGmbH, Germany

2 GLR Green Land Reclamation Ltd., United Kingdom

3 CTI Comitato Termotecnico Italiano, Energia e ambiente, Italy

4 TFZ Technology and Support Centre of Renewable Raw Materials, Germany

5 INETI Departamento de Engenharia Energética e Controlo Ambiental / Instituto Nacional de Engenharia e Tecnologia Industrial, Portugal

6 SFN Signalsfromnoise.com Ltd, United Kingdom

7 TNO Institute of Environmental Sciences, Energy Research and Process Innovation, The Netherlands

8 SLU Swedish University of Agricultural Sciences - Department of Bioenergy, Sweden

9 BLT Federal Institute for Agricultural Engineering, Austria

10 KCL Oy Keskuslaboratorio, Finland

11 USTUTT Institute of Process Engineering and Power Plant Technology, Germany

12 OFI Austrian Research Institute for Chemistry and Technology, Austria

13 TPS Termiska Processer AB, Sweden

14 TUV Vienna University of Technology - Institute of Chemical Engineering, Austria

15 NTUA National Technical University of Athens - Department of Chemical Engineering, Greece

16 FORCE FORCE Technology (former dk-Teknik Energy & Environment), Denmark

17 UOULU University of Oulu - Department of Chemistry, Finland

18 CIEMAT Centro de investigaciones energéticas, medioambientales y technologicas, Spain

19 ECN Netherlands Energy Research Foundation, The Netherlands

20 CRA Centre de Recherches Agronomiques de Gembloux, Belgium

21 DFLRI Danish Centre for Forest, Landscape and Planning, Denmark

22 HFA Holzforschung Austria, Austria

23 FCA Forestry Contracting Association Ltd., United Kingdom

24 FAT Agroscope FAT Taenikon Swiss Federal Research Station for Agricultural Economics and Engineering, Switzerland

25 AFOCEL Association Foret Cellulose, France

26 TW Tech-wise A/S (Elsam), Denmark

27 SKAB Skelleftea Kraft AB, Sweden

28 VT-TUG Graz University of Technology, Institute for Resource Efficient and Sustainable Systems, Austria

29 CPERI Centre for Research and Technology Hellas - Process Engineering Research Institute (CPERI), Greece

30 VTT Technical Research Centre of Finland, Espoo, Finland

31 SP Swedish National Testing and Research Institute, Sweden

32 IEC Technical UniversityBergakademie Freiberg - Institute for Energy Process Engineering and Chemical Engineering, Germany

33 IFE-A IFE-Analytik GmbH, Germany

34 CLGE Central Laboratory of General Ecology – Bulg. Acad. of Sci., Bulgaria

35 KONEKO KONEKO Marketing Ltd., Czech Republic

36 IPE LAS Institute of Physical Energetics - Latvian Academy of Sciences, Latvia

37 LEI Lithuanian Energy Institute, Lithuania

38 BREC EC Baltic Renewable Energy Centre, Poland

39 NYME University of West Hungary – Department of Energetics, Hungary REPORTING PERIOD : FROM 1st January 2002 TO 31st December 2004

PROJECT START DATE : 1st January 2002

DURATION : 36 months

Date of issue of this report : 11th April 2005

Project funded by the European Community under the 5th Framework Programme (1998-2002)

Contents of the Final Technical Report

BioNorm Project - Final Technical Report III

Contents of the Final Technical Report

Executive summary ...........................................................................................................IV

Part 1 – Publishable Final Report (Synthesis Report)

Part 2 – Detailed Final Report

Part 3 – Management Report

Executive summary

BioNorm Project - Final Technical Report IV

Executive summary

Due to its advantages (like wide availability, well-known with regard to supply and use and contribution to green house gas mitigation) biomass represents the most important source to address European policies increasing the share of renewable energy sources for a more sustainable energy system. For an improved integration of solid biofuels into the energy system a dynamic European market has to be created. Thus, accomplishable biofuel properties as well as procedures to test and control these defined parameters are required. Therefore, besides the elaboration of an overall Quality Assurance system for solid biofuels based among other on sampling and testing methods, the pre-normative research project (BioNorm) was aimed to directly contribute to the ongoing development of European standards for solid biofuels within CEN TC 335 “Solid biofuels”. Within this project the emphasis was laid on the identification and evaluation of the best appropriate sample-, test- and reference methods for the determination of specific fuel properties. Based on the experiences made respective best practice guidelines were compiled. These guidelines and research findings were incorporated for drafting Technical Specifications within the standardisation process.

In each of the different work packages and tasks respectively various kinds of solid biofuels covering a wide scope of fuels were analysed and investigated by carrying out selected methods. The applied methods and principles were initially identified to be basically suitable for a reasonable determination of (i) the number of increments and tests as well as specific increment sizes in terms of sampling and sample reduction, (ii) the physical/mechanical fuel properties (i.e. moisture content and bulk density, ash melting behaviour, particle size distribution, durability and particle density) and (iii) the chemical fuel characteristics (i.e. sulphur, nitrogen and chlorine as well as major and minor elements). The research has two emphasises. It was focused on the investigation of existing methods and equipment (e.g. applied for solid fossil fuels) with regard to their accuracy for solid biofuels and thus their applicability followed by appropriate adaptation and improvements of these methods and laboratory equipment. In addition, it was also focused on the development of new methods. Referring to this, common statistical parameters such as accuracy, reproducibility and repeatability were used for the assessment of the different sampling and testing methods. Besides technical, also work-efficiency, economical and environmental aspects were considered.

Supported by these research results, new methods for the development and implementation of Quality Assurance systems for the entire biofuel supply chain were evaluated based on an initial review of existing systems as well as extensive field trials at several companies.

The outcome of the pre-normative work of the BioNorm project includes among methods for sampling and sample reduction, improved and new developed methods and procedures for the determination of physical-mechanical and chemical biofuel properties as well as the development and implementation of a company specific Quality Assurance system. Coupled with this, basic recommendations are with regard to apply both the proposal of a standard/TS and the guideline in practice.

Executive summary

BioNorm Project - Final Technical Report V

In terms of the country conditions in the NMS/NAS recent trends have shown a continuous increasing international market of solid biofuels that also stimulates the domestic production of refined solid biofuels within these countries and thus increase the acceptance of renewable energy sources. Nowadays, these countries already use biomass (predominantly for domestic heat provision) and have very promising potentials concerning biomass and bioenergy utilisation. However, currently limited experience in utilisation of refined solid biofuels and missing R&D contributes to a lack of solid biofuels standards and Quality Assurance guidelines. In all of the NMS/NAS there are no specific biofuel standards and Quality Assurance implemented yet. Companies that produce refined solid biofuels for export currently apply national standards of the import countries. Referring to this, all NAS/NMS-partners clearly stated that common standards are urgently required for increasing the solid biofuel market. Hence, CEN standards currently being developed need to be quickly adopted.

Among the extensive and substantial outcomes during the pre-normative research and the progress of standardisation it has been acknowledged that there is still the demand on research in all purviews such as of fuel classification and Quality Assurance or sampling and testing methods and procedures. Moreover, further cooperation is recommended within the European countries, that already used biomass efficiently and their legislation and biofuels standards are harmonious developed.

Finally, the outcomes reveal that BioNorm has crucially contributed to clarify important issues and aspects associated with the ongoing European standardisation process for solid biofuels. This is in particular true referring to biofuel terminology and specification, appropriate procedure and methods for sampling and testing of fuel properties as well as the importance of company specific Quality Management systems. Furthermore, a basis for research exchange with the NMS/NAS has been established that will be further delved and extended in the future. However, BioNorm has also point out the urgent need on further investigations and method development. Therefore, continuing R&D support of the ongoing standardisation activities will be of high importance to close the gaps and to ensure that the European market for solid biofuels and thus the biomass industry will continue to grow rapidly in the future helped by acceptable standards.

Part 1 Publishable Final Report (Synthesis Report)

Part 1 Table of contents

BioNorm Project - Final Technical Report II

Table of contents

List of figures .....................................................................................................................IV

List of tables .......................................................................................................................IV

Glossary ............................................................................................................................... V

1 Introduction................................................................................................................... 1 1.1 Objectives and strategic aspects............................................................................... 1

1.1.1 Scientific and technological objectives ............................................................. 1 1.1.2 Socio-economic aspects and contribution to the EU policies............................ 2

1.2 Structure of project and synthesis report.................................................................. 3

2 Sampling and sample reduction .................................................................................. 5 2.1 Objectives................................................................................................................. 5 2.2 Methodical approach................................................................................................ 5 2.3 Conclusions and recommendations.......................................................................... 6

2.3.1 Sampling............................................................................................................ 6 2.3.2 Sample reduction ............................................................................................... 7

3 Physical / mechanical tests ........................................................................................... 9 3.1 Moisture content and bulk density determination.................................................... 9

3.1.1 Objectives .......................................................................................................... 9 3.1.2 Methodical approach ....................................................................................... 10 3.1.3 Conclusions and recommendations ................................................................. 11

3.2 Ash melting behaviour ........................................................................................... 13 3.2.1 Objectives ........................................................................................................ 13 3.2.2 Methodical approach ....................................................................................... 13 3.2.3 Conclusions and recommendations ................................................................. 14

3.3 Particle size distribution and dimension ................................................................ 15 3.3.1 Objectives ........................................................................................................ 15 3.3.2 Methodical approach ....................................................................................... 15 3.3.3 Conclusions and recommendations ................................................................. 16

3.4 Durability and raw density of pellets and briquettes.............................................. 17 3.4.1 Objectives ........................................................................................................ 17 3.4.2 Methodical approach ....................................................................................... 17 3.4.3 Conclusions and recommendations ................................................................. 18

4 Chemical tests.............................................................................................................. 20 4.1 Sulphur, chlorine and nitrogen content .................................................................. 20

4.1.1 Objectives ........................................................................................................ 20 4.1.2 Methodical approach ....................................................................................... 20 4.1.3 Conclusions and recommendations ................................................................. 22

4.2 Determination of major and minor elements.......................................................... 23 4.2.1 Objectives ........................................................................................................ 23 4.2.2 Methodical approach ....................................................................................... 23 4.2.3 Conclusions and recommendations ................................................................. 24



5 Fuel Quality Assurance .............................................................................................. 26 5.1 Objectives............................................................................................................... 26 5.2 Methodical approach.............................................................................................. 26 5.3 Conclusions and recommendations........................................................................ 27

5.3.1 Review of quality systems............................................................................... 27 5.3.2 Implementation of Quality Assurance in companies....................................... 28 5.3.3 Proposal for a draft standard for Quality Assurance ....................................... 30

6 Summary...................................................................................................................... 31

References .......................................................................................................................... 36



List of figures

Figure 1: Structure of the BioNorm project ..................................................................... 3

Figure 2: Interdependency among physical-mechanical fuel properties [19].................. 9

Figure 3: Procedure for moisture content round robin trials – moisture determination with rapid testers and sampling of reference material followed a standard sequence in order to obtain as uniform trial conditions as possible [14] ....... 10

Figure 4: Functioning principles of screening and image analyses used in round robin trials [23, 24, 25] ............................................................................................ 16

Figure 5: Liquid displacement methods applied for the determination of particle density [27] ................................................................................................................. 18

Figure 6: Methodology to develop and implement Quality Assurance (QA) [13] ........ 29

Figure 7: Supply chain of solid biofuels [12] ................................................................ 30

List of tables

Table 1: Recommended numbers and sizes for sampling .............................................. 7

Table 2: Recommended methods of sample reduction................................................... 8

Table 3: Recommended detection systems for major and minor elements [31, 32] .... 25

Table 4: Summary of preferred methods for investigated solid biofuels (continued next page) ..................................................................................... 32



Glossary

ANOVA analysis of variance

CCD charge-coupled device

CCP critical control points

CEN Comité Européen de Normalisation (European Committee for Standardisation)

CVAAS cold vapour atomic absorption spectrometry; system for determination of Hg

FAAS flame atomic absorption spectrometry; system for element determinations

FME failure mode and effect

F-tests statistical tool to test the homogeneity of two standard deviations

GC/MS gas chromatography - mass spectrometry

GFAAS graphite furnace atomic absorption spectrometry; system for trace element determinations

GHG greenhouse gas

GROT Swedish word for tops and branches of trees discarded during felling

IC Ion chromatography

ICP-MS inductively coupled plasma mass spectrometry; system for multi-element determinations of trace elements

ICP-OES inductively coupled plasma optical emission spectrometry; non destructive system for multi-element determinations

MAF Melt area fraction

MC moisture content

NAS Newly Associated States

NMS New Member States

QA Quality Assurance

QC Quality Control

QM Quality Management

SDTA simultaneous difference thermo analysis

SEM-EDS scanning electron microscope combined with energy dispersive X-ray spectrometry

TC Technical Committees

TGA thermo gravimetric analysis


BioNorm Project - Final Technical Report VI

t-tests test for a difference between two means

WP work package

XRF X-ray fluorescence spectrometry; non destructive system for multi-element determinations

Part 1 Introduction

BioNorm Project - Final Technical Report 1

1 Introduction

In its White Paper for a Community Strategy and Action Plan the European Union has set the target to increase the share of renewable energy sources to 12 % of the gross energy consumption by 2010. Biomass is one of the renewable energy sources that has the potential to major contribute for reaching this goal. Today, biomass contributes about two-thirds to the total renewable energy production. If produced in a sustainable way (i.e. primarily from forestry and agriculture) biofuels support to greenhouse gas mitigation. Moreover, being a domestic energy carrier, they help Europe to become more independent of imported fossil fuels and thus increasing the security of energy supply [1].

To take benefit of these advantages, biomass has to be integrated much better into the energy system. Therefore, a dynamic and sustainable international market for biofuels is urgently needed. However, for managing and obtaining a smooth business in Europe firm and accomplishable “market rules” and standards for common understanding referring to terminology, sampling and testing as well as Quality Assurance on biofuels are required. Within the ongoing European standardisation process the Technical Committee CEN TC 335 "Solid Biofuels" currently develops such "market rules" or standards [1, 3]. From the elaborated comprehensive knowledge basis best practice guidelines are derived. Project results are introduced to the Technical Specifications drafted by CEN TC 335. These Technical Specifications are accounted as tool of importance to support the biofuels trade in Europe.

1.1 Objectives and strategic aspects The main objective of BioNorm was to provide the scientific background for the European standardisation process of solid biofuels. The emphasis of the project was laid on a fuel Quality Assurance system for solid biofuels in order to support and extend the market. Such a system is based on extensive practical work on sampling and testing to improve the existing procedures and to allow for the provision of a high quality biofuel throughout the overall supply chain and thus improving the confidence of the consumer into the biofuel. Besides the work on terminology, specifications and classes, this includes the testing of physical-mechanical fuel characteristics (e.g. moisture content, bulk density and ash melting behaviour) as well as chemical fuel properties (among other elements particularly sulphur, chlorine and nitrogen content). The integration of different countries from the Newly Associated States (NAS) - meanwhile commonly New Member States (NMS) - ensured to take into account the standardisation requirements of these countries and to basically meet their specific needs. The work realised within the BioNorm project was closely linked to the work of CEN TC 335 “Solid Biofuels” guaranteeing an excellent exploitation of the results.

1.1.1 Scientific and technological objectives

Pre-normative investigations have revealed existing methods for sampling and testing as well as for Quality Assurance in use for solid fossil fuels (e.g. hard coal). However, shown

Part 1 Introduction


in several experiences these methods are only partly applicable to solid biofuels. Although CEN TC 335 "Solid Biofuels" will continue to develop standards on the basis of the best information currently available, it recognises the strong need for further research in some areas. This involve improvements of the reliability of selected sampling and testing methods, and the strong need for the development of an overall Quality Assurance system for solid biofuels taking practical needs into consideration.

Therefore, this project was aimed to carry out research on sampling and sample reduction procedures as well as physical-mechanical and chemical tests for a more successful practical implementation. Furthermore, the results were integrated into a fuel Quality Assurance system covering the overall supply chain. The research is aimed at comprehensive evaluation and identification of the best suitable methods for the respective determination, providing a knowledge basis for best practice guidelines. These guidelines and findings were also used for drafting of Technical Specifications within the standardisation process and thus contributing the work of CEN TC 335 "Solid Biofuels" to accelerate the development of improved standards with consideration of practical requirements. Coupled with this, the conveners of the different working groups of CEN TC 335 were involved deeply within this project.

1.1.2 Socio-economic aspects and contribution to the EU policies

The European Commission (EC) has given a mandate to CEN for the elaboration of standardisation in the field of solid biofuels. The implementation and application of European standards on sampling, testing and classification of solid biofuels as well as Quality Assurance is seen as an important element towards an increased utilisation and international trade of solid biofuels. Expanding markets are required to fulfil targets of the EC directives and to reach the political goals on national level.

An increased use of biomass for energy purposes is reflected in positive socio-economic effects on regional, national and global level. Regionally, it can create new market opportunities for farmers, contribute to the preservation of rural areas and improve the local infrastructure. On a national level, due to increasing biofuels application new jobs are created in the business sectors of engineering, manufacturing, energy production, energy distribution and consultancy. For instance, the broader implementation of small-scale biomass technology is labour intensive when compared to the larger-scale energy production using fossil fuel energy; thus creating employment especially in rural areas. It can be assumed that most of the new jobs will probably be related to fuel handling and plant operation. Being an domestic fuel, biomass also helps to reduce the significant dependence of the European on fossil fuel imports and thus, increasing the security of supply. Globally, the reduction of climate relevant greenhouse gas (GHG) emissions has a significant beneficial effect in long-term

Contributed by the close collaboration BioNorm work packages and CEN working groups (and thus ensuring the dissemination and exploitation of the pre-normative work), it was focused to develop highly sophisticated and therefore widely acceptable standards for the stimulation of the bioenergy markets throughout Europe. Users of these standards are mainly producers, traders and consumers of biofuels. Summing up all players involved into

Part 1 Introduction


the production, provision and use of solid biofuels throughout the overall value chain within Europe will benefit from this work realised within the BioNorm project.

Widely accepted standards resulting from the pre-normative work of BioNorm will also contribute to boost the market for devices for sampling and testing of solid biofuels. Additionally, the end use conversion technologies (e.g. combustion and gasification units) can be optimised with regard to the different fuel qualities. The same applies to the machinery and the devices required for the production and provision of solid biofuels, which in turn are needed to overcome the gap between biofuel provision and its utilisation. Finally, the results of the BioNorm project will help to contribute to the continued increasing biofuel market in Europe and therewith assist the European Commission to reach its environmental goals.

1.2 Structure of project and synthesis report According to the objectives named above, the project work plan has consisted of interrelated work packages, as illustrated in Figure 1. Referring to this, the BioNorm project deals with issues of sampling and sample reduction, with tests determining the physical-mechanical as well as the chemical properties and with aspects of biofuel Quality Assurance. These different work packages were strongly linked to each other by means of collaboration and common meetings. As sampling and testing represent activities of high importance for Quality Assurance within the solid biofuel supply chain, these work packages should support the elaboration of such a Quality Assurance system. In order to ensure that effects of the BioNorm are of greater impact on the continuous increasing biofuel market BioNorm was supplemented by involving also NMS/NAS in the elaboration of suitable test methods and for research exchange.

Fuel Quality Assurance

Sampling and Sample Reduction

Chemical Tests

Physical/Mechanical

Tests

Research Exchange

with NMS/NAS

Figure 1: Structure of the BioNorm project

The biogenious materials investigated within the BioNorm project were predominantly chosen because of their importance on the European market of solid biofuels. Accordingly, relevant solid biofuels based on woody, herbaceous and fruit biomasses were analysed and tested in context of sampling and sample reduction as well as physical/mechanical fuel properties (e.g. moisture content, size distribution and durability) and chemical fuel

Part 1 Introduction


characteristics (e.g. sulphur, chlorine, major and minor elements). Thus, in the following the focus is set on sampling and sample reduction, physical-mechanical and chemical test as well as on Quality Assurance. Starting with a short introduction of respective objectives, the respective work packages are summarised in terms of their methodical approach and scientific conclusions as well as recommendations. More comprehensive information on the different work packages and the corresponding tasks can be gained from the detailed final reports given in Part 2 of this final technical report. This is also true for the results of the review in terms of respective country conditions given for the NMS/NAS.

This synthesis report (Part 1) is finished by a summary involving final conclusions and recommendations of the whole BioNorm project.

Part 1 Sampling and sample reduction


2 Sampling and sample reduction

In this work package (WP I) five partners dealt with methods for sampling and sample reduction that are of importance with regard to testing methods of fuel properties in order to ensure that the required fuel properties are met (e.g. methods for reducing samples for sample preparation for physical-mechanical and chemical tests). The potential impact of this issue on the biofuel markets is considerable.

2.1 Objectives Often variabilities referring to sampling and sample reduction are much more significant than variabilities referring to testing methods. In average the possible error which can be made by an inappropriate sampling can be about ten times higher compared to testing. On this background, WP 1 is aimed on supporting information to Working Group 3 of the CEN/TC 335 ‘Solid Biofuels’ in order to improve the reliability of the draft European standards for solid biofuels. Therefore, investigation of methods for sampling (Task I.1) and sample reduction (Task I.2) were carried out on a representative bulk sample such as straw bales and a range of woody biofuels (i.e. sawdust, wood-chips and wood pellets).

The main objectives of this work package are [5, 8]:

• to assess the bias which may be introduced when samples of solid biofuels are taken from containers or stockpiles instead of from moving streams

• to define the number and size of increments needed to provide a representative sample

• to assess the variations introduced when reducing the size of samples of solid biofuels to form suitable test portions

• to determine the usefulness of a range of practical sample reduction methods

2.2 Methodical approach Determining the size and number of sample increments was the purpose of the experimental work on sampling. Comparing the efficacy of various methods was the purpose of the work on sample reduction [8].

For the woody materials (i.e. sawdust, GROT1 and pellets), investigations on sampling methods were executed on moving streams (conveyor) as well as on stationary piles (heap). The sample increments received in three different sizes were used for determinations of moisture content, ash content and particle size distribution. Investigations on sample reduction methods involved using the methods of riffling, coning and quartering as well as the "long-pile" and the rotary divider.

1 Swedish word for tops and branches of trees discarded during felling. GROT is a kind of wood chips.



Straw bales were sampled in two different ways: by using a coring device and a special hook. The resulting sample increments were tested for moisture content, chloride content, and ash content. The method of coning and quartering was also applied to reducing straw bale samples, with preliminary chopping of the straw in the increment. Further methods used were hand sampling with and without preliminary chopping of the straw as well as using a long pile, also with chopping of the straw [1, 4, 8].

In all uncertainties associated with testing within a single laboratory as well as associated with methods of sampling and sample reduction were measured. For this, statistical instruments (e.g. relative error, standard deviation) were applied, allowing the investigation of the variability between sampling increments (i.e. number of increments for a specified level of sampling variability), between samples, between determinations (i.e. number of determinations for a specified level of repeatability) as well as between the test results [5].

2.3 Conclusions and recommendations The experimental work has allowed the relative bias of different sampling methods and the influence on sampling variability of different increment sizes to be assessed. This work has also allowed the number of sampling increments required to give a satisfactory level of sampling variability to be estimated, for the biofuels investigated (i.e. woody-fuels such as GROT, pellets and sawdust as well as straw bales). On the basis of these experiments it can be concluded that none of the methods used in the experiments gave disastrous results. Thus, no method can be ruled out from practical use and should be included in the CEN standard [8].

Below the main conclusions and recommended options of sampling and sample reduction trials are briefly summarised [5, 6, 8, 9].

2.3.1 Sampling

In sum it is revealed: with exception of particle size distribution, no evidence of a relative bias between the methods of sampling and testing can be revealed for testing moisture, ash and chlorine content of the fuels investigated. Preferred numbers and sizes for sampling are summarised in Table 1.

Sampling from tipped lorry-loads is not biased relative to sampling from a conveyor for the methods recommended for GROT and sawdust.

Based on the work on GROT it can be concluded that for moisture and ash, the results on the relative bias do not provide any reason for preferring sampling from the conveyor over sampling from the heap. Different from that, particle size distribution shows relative bias implying the necessity for the CEN standard to define one preferred method.

For moisture and ash of sawdust analysis the results concerning the relative bias of the two methods show the same as for GROT. For particle size distribution, the results imply a relative bias between the two methods in case of smaller increment sizes (0.2 litres or 1 litre). Thus, if sampling from the stopped conveyor will be the reference method, then sampling from the heap is acceptable provided that sufficiently large increments are taken.



The results obtained for pellets in the investigation of relative bias indicate that the moisture and ash content vary from lorry-load to lorry-load. However, both are not affected by the handling, what is true for the particle size distribution.

For straw bales, the experiment results have shown that sampling with the hook is not biased relative to sampling with the coring machine. Taking five increments per bale will results a relative sampling error of about 10 %. Furthermore, doing one determination per sample will give a relative error of test results (i.e. repeatability) of about 5 %. Both relative errors might be acceptable for routine tests. Due to the effect of straw bales position, when using either the hook or the coring machine, for sampling straw bales should be turned on their side. In addition, increments should be taking from both sides, and from the full depth of the bale [5, 9].

Table 1: Recommended numbers and sizes for sampling

Biofuel Test method Sampling method Increment sizea

Number of increments

Number of determi-nations

moisture heap / conveyor 1 litre 5 1

ash heap / conveyor 1 litre 10 4 GROT size distribution

(passing 64 mm/2 mm) heap / conveyor 20 litre 5 4

moisture heap / conveyor 0.2 litre 2 1

ash heap / conveyor 0.2 litre 5 2 Sawdust size distribution

(passing 5.6 mm/0.5 mm) heap / conveyor 0.2 litre 5 2

moisture heap / bags 0.25 kg 5 1

ash heap / bags 0.25 kg 5 2 Pellets size distribution

(passing 5.6 mm/0.5 mm)heap / bags 1 kg 1 2

moisture core / hook - 5 1

ash core / hook - 5 1 Straw bales

chlorine core / hook - 5 1 a not analysed for straw bales

2.3.2 Sample reduction

Preferred methods (i.e. those gave least variations between sub-samples) in each of the experiments are outlined in Table 2. Some general conclusions may be drawn tentatively from this. “Riffle” is the preferred method to use with the coarser materials (GROT and pellets), if a rotary divider is not available. If available, a rotary divider is the preferred method for determinations of moisture and particle size distribution on pellets. An explanation should be sought for the poor performance of the rotary divider in comparison with the riffle when determining ash. “Coning” is the preferred method with sawdust, but the riffle performed nearly as well. The special method used to sub-sample straw, handful sampling (on straw coarse cut prior to reduction), is the preferred method for determination



of liberated or partially liberated properties of straw, but not for determination of moisture [8, 10].

In general is was revealed, when a sample is taken for the purpose of determination of moisture-content, sample reduction should be avoided, since the materials dried noticeably during the reduction process and the result of the test will be affected.

Table 2: Recommended methods of sample reduction

Biofuel Moisture content Ash content Chloride content Particle size distribution

GROT riffle riffle - riffle Sawdust riffle / coning / long pile coning - coning Pellets rotary* riffle - rotary* Straw coning / long pile hand 2 hand 2 - * or “riffle” if no rotary divider is available

However, it is important that technicians should regularly and routinely check the achieved repeatability with whatever sample reduction methods applied for [5, 8].

Finally, it is suggested that there is further need to [5, 6]:

• extend the work on sampling other solid biofuels in order to cover the breadth of materials to be found throughout Europe (e.g. wood chips, bark, reed canary grass, olive waste and briquettes)

• develop a guideline on how users of CEN standards can ensure reliable results of sampling and sample reduction procedures as well as can decide the frequency of sampling

Part 1 Physical / mechanical tests


3 Physical / mechanical tests

Determining physical and mechanical fuel properties is matter of this extensive work package (WP II), in which 17 partners across Europe were involved. The physical and mechanical fuel properties are largely determining each other. Moisture is the parameter with the highest influence on other physical properties (cf. Figure 2). Testing of these properties is routine.

Figure 2: Interdependency among physical-mechanical fuel properties [19]

However, the precision and reproducibility of the results is often very scanty. In addition, rapid and reliable test methods are urgently required for e.g. on-site fuel acceptance or rejection. Therefore, within this work package existing procedures were improved concerning measuredness, and new procedures were developed respectively for moisture content and bulk density (Task II.1), the determination of ash melting behaviour (Task II.2), particle size distribution and dimension (Task II.3) as well as the determination of the durability and raw density of pellets and briquettes (Task II.4).

Except for ash melting behaviour all tests were conducted by international round robin trials by several participants. One important aim of the round robin trials is the comparability and transferability of the results by testing the quality of methods (i.e. repeatability) applied for physical-mechanical properties. The results provided a knowledge basis for best practice guidelines. These guidelines and findings were also used for drafting of Technical Specifications within the standardisation process.

3.1 Moisture content and bulk density determination

3.1.1 Objectives

Moisture content is a primary property for a successful utilisation of solid biofuels within the entire supply chain (e.g. for transportation costs, storage management, calorific value and conversion at end use). Today, several national standards for testing moisture content



are available. However, for common European standards or guidelines reliable tests have been identified. Thus, the research is aimed (i) to study different test methods for the determination of moisture content and non-water losses in comparison with the oven drying method (reference method) and (ii) testing of rapid on-sites applications using the oven drying method as reference (rapid methods). A wide scope of fuels from all over Europe was analysed to cover the broad range of biomass. This is also true for the applied technologies and measuring principles.

Strongly linked to the moisture content of solid biofuels is the bulk density, which is an important property with regard to space for storage and transportation and for volume based payment of biofuels. Furthermore, bulk density is influencing rapid moisture content measurements. The applied national and international standard methods (e.g. German, Swedish and American standard) for determining bulk density are highly inconsistent in practice and confusing in their size and the way of handling [14, 15, 18].

3.1.2 Methodical approach

The investigation was dominated by carrying out round robin trails to determine rapid moisture content (cf. Figure 3) and bulk density.

Figure 3: Procedure for moisture content round robin trials – moisture determination with rapid testers and

sampling of reference material followed a standard sequence in order to obtain as uniform trial conditions as possible2 [14]

2 Please see the final report concerning moisture content for detailed information about deployed test

equipment/devices like Mettler (thermogravimetric) or Moist (dielectric microwave).



Moisture content The total moisture content and volatile components were determined by means of: (i) the oven drying method, (ii) the xylene distillation method and (iii) the freeze drying method. Furthermore, (iv) the new analytical GC-MS-method (“gas chromatography - mass spectrometry”) for measuring the non-moisture volatile matters (e.g. terpenes) released during the oven drying process was applied. About 21 biofuel types (e.g. wood chips, bark, pellets, sawdust, olive stones, miscanthus, cork) were analysed using these reference methods.

The rapid moisture test research was carried out using: (i) the thermogravimetric method by oven drying and halogen drying, (ii) the electrical method based on di-electrical properties of moisture subdivided in capacitive method and the microwave method as well as (iii) the optical measuring method based on infrared reflectometric determinations. About 11 laboratories participated in the round robin. All partners followed detailed sample preparation and measuring guidelines in order to reduce any effects from sample handling and preparation [14, 15]. The evaluation of results was done using ANOVA (i.e. ANalysis Of VAriance) for homogeneity as well as significance tests of variances and mean values (e.g. F-tests and t-tests).

Bulk density For bulk density determination several test methods with research focus on container size and shape, shock impact on the measurement and varying moisture content were analysed by testing about 48 biofuels in total. Assuming that the volume remains constant (i.e. negligence of shrinkage) all original bulk density data were initially calculated to dry matter basis. These data were used for container and treatment evaluation. Furthermore, for the identification of any influence during measurings through fuels bulk density itself, all wood chips were allocated in high and low density fuels (at 180 kg/m³, dry basis).

For container comparison one cube (100 l of nominal volume) and three cylinders (100 l, 50 l and 15 l of nominal volume) were used. All measurings were performed applying two different treatments: (i) “without shock impact” (i.e. normal filling, levelling and weighting) and (ii) “with shock impact” (i.e. weighting after having dropped the container three times from 150 mm height and having it refilled). The influence of moisture content was tested by previous progressively drying of several fresh biofuels samples in order to provide identical sub-samples at varying moisture content levels. Measurements repeatability was tested by comparison of relative repeatability limits (coefficient of variation) [17, 18].

3.1.3 Conclusions and recommendations

Basic results of methods for testing moisture content and bulk density of solid biofuel, which are identified to be the most promising for standardisation in terms of e.g. suitable performance and high reproducibility, are briefly summarised below.

Moisture content – reference method The applied moisture determination results gave overall comparable results. A statistical significant difference was found among the oven drying method between the standard



method at 105 °C compared to the temperatures 80 °C and 130 °C, respectively. According to this, a slightly increased amount of evaporated matter (volatiles and moisture) results for each temperature step.

The freeze drying method determined significantly lower moisture content values compared to the standard method, whereas the results from the xylene method were deviating only for some of the fuels. This difference was partly attributed by the small sample size used for xylene distillation.

The GS-MS-method has shown promising results as a standard reference method. So it was found that the amount of volatile matter was in order of 0.06 to 11.33 % calculated as percentage of moisture content. Furthermore, it was revealed that the mass loss differences found among the different methods were primarily due to α- and β-pinenes. Though, a more throughout optimisation of the method is required [14, 15].

Moisture content – rapid methods Among the seven tested rapid test devices about four types were identified to be particularly applicable to measure the moisture content of solid biofuels. These are:

• as on-site types: the thermogravimetric Mettler-Toledo HB45 as well as the capacitive devices Pandis FMG3000 and the Schaller FS2002-H.

• as on-line type: the optical MESA MM710. Although MESA was only tested in a reduced moisture range of 10 to 40 % MC but the method is applicable to the full range of 0 to 100 % MC.

However, the small nominal size and high time need (both Mettler-Toledo) as well as the need of fuel specific calibrations (Pandis, Schaller and MESA) should be considered when selecting the device. For the electrical devices including the bulk density in the calibration function can significantly increase measuring exactness. Moreover, reducing the scope of fuels increases the power of the calibration functions.

The Moist100, Wile25 and ACO estimated the moisture content with a higher variation and therefore cannot be recommended for moisture content determination in solid biofuels [14, 15].

Generally, further work is required to optimise and evaluate e.g. blank values, recovery, adsorption and extraction efficiency as well as response factors [16].

Bulk density The following was concluded for best practice in bulk density determination of solid biofuels [17, 18]:

• For all tested solid biofuels a measuring container size of 50 l is acceptable. A cylindrical container shape should be preferred for practical reasons due to higher stability and easier manageability.

• A standardised shock impact on the filled container significantly increases the measured bulk density (e.g. 6 % for wood pellets, 10 to 12 % for fuel chips and



18 % for chopped miscanthus), while there was only found a minor improvement for the relative repeatability limits.

• The fuels moisture content during the measurement is of high importance, and consequently has to be recorded since an obvious increase in bulk density (wet basis) was observed with increasing moisture content. Thus, the comparability of bulk density data is only given if any inconsistency in moisture content between the samples is accounted for by the use of a correction factor of 0.712 % for each 1 % moisture difference.

Moisture content effects are largely restricted to the moisture content range up to 25 %. Beyond this point possible effects can be neglected.

3.2 Ash melting behaviour

3.2.1 Objectives

The determination of biofuels ash melting behaviour is of high importance for all thermal conversion processes. The temperature range of sintering, softening and melting can vary broadly, depending primarily on ash composition. Nowadays, it is impossible to forecast ash melting behaviour by calculations based on standard elemental analysis. Standards for evaluating ash melting behaviour of coal and coke are also used for biofuels today. For biofuels known to be problematic in terms of their ash, the implementation of existing laboratory and recommended new methods were tested and assessed. The tested fuels were wood/bark (relatively unproblematic ash), olive stones (unknown ash behaviour), straw and Lucerne (sticky and very sticky ash) [20].


Various methods for the characterisation of biomass ash melting behaviour were investigated. The production of biofuel ash has been carried out according to CEN TS 14775 at 550 °C. The methods used by the seven involved partners are characterised by computer-controlled analysis during thermal conversion of biofuels. Except for MAF (“melt area fraction”), they are established techniques adopted for this specific use.

The improved DIN3 and the MAF4 are methods based on analysis of images that are acquired from a high temperature light microscope by means of a CCD camera. Using these methods, ash melting behaviour (i.e. characteristic temperatures for sintering, deformation, sphere, hemisphere and flow) can be gathered from the specific contours of the analysed test pieces during the melting process. According to their melting behaviour, ashes can be classified into little, moderate and severe sintering/slagging, respectively.

Other applied methods are based on bed sintering methods that determine the onset of agglomeration by monitoring diffential pressure and temperatures in the bed. It is 3 also called heating microscope method 4 new developed by the BioNorm partner FORCE Technology, Denmark



distinguished between batchwise CFBA (“controlled fluidised bed agglomeration”) and continuously fed types. Both methods are used for complementary as well as for comparison with the other methods reported.

Simultaneous TGA (“thermogravimetric analysis”) and SDTA (“simultaneous differential thermal analysis”) analyses imply continuous measurement of the sample weight and the sample temperature during heat-up. For a combustion ash, the melting will result in several peaks overlapping each other, corresponding to the melting of the different chemical compounds in the ash.

Tests were also performed by SEM/EDS (“scanning electron microscope” combined with a “energy dispersive x-ray analyser”). Based on that method the distribution of chemical species (e.g. Al, Ca, Cl, Fe) in individual particles from a sample. The crystalline phases present in an ash and the content of amorphous material (i.e. the crystalline) were determined by using XRD method (“X-ray diffraction”).

Besides named biofuels, synthetic ash samples (pyrolytic Orton Cones, typically used by potters) were tested using the laboratory methods DIN, MAF, TGA/ SDTA and XRD [20, 21, 22].


Although biofuels ashes were tested by seven methods, only from improved DIN, MAF and CFBA temperature information can be derived that can be compared directly. For synthetic samples it was revealed that only the DIN and the MAF method are able to quantify melting temperatures. However, DIN and MAF are ash testing methods suitable for standardisation whereas CFBA may be used as reference for agglomeration/sintering. The recently developed MAF method shows large potential to become reproducible and repeatable. The MAF method has been proven to determine the temperature of 10 % and 50 % melt in the ash samples tested. Though, some modifications are needed for problematic ashes, e.g. straw ash, which tends to form “cakes” (agglomerates) during handling. Moreover, further improvement is required concerning melt viscosity and the possible sample shrinking.

The technologies of TGA/SDTA cannot be standardised at present, but they have the potential to confirm various phenomena associated with ash melting behaviour. SEM-EDS is predestined for providing valuable information on the ash compositions and species. Testing the XRD method revealed that the amount of amorphous phase is evaluable if the connection between amorphous phase and melt phase is clarified before.

Besides the suitability of test methods, issues such as really required information to forecast ash melting and the relation of acceptable analysis cost. It is generally accepted that testing of initial melt temperature as well as the rate of melt formation is essential. With regard to the costs, bed sintering methods and methods as TGA/SDTA that produce results that have to be evaluated by specialists can be eliminated. The improved DIN method is an attempt to get around with comparatively simple and low-cost manner in the standards. This is also true for the MAF method.



However, there certainly are some problems about the identification of the characteristic temperatures for biomass ashes. This is also true for variations (not studied within BioNorm) in the density of the test specimen that is produced by compression. Furthermore, the reproducibility has to be improved [20, 21, 22].

3.3 Particle size distribution and dimension

3.3.1 Objectives

For an unobstructed biofuel use (i.e. fuel handling, storage, transhipment and energetic conversion) particle size dimensions and their distributions in a bulk have often been identified as a key quality parameter. But, the determination of these parameters is still uncommon, since the provision of reproducible and comparable data on physical biofuel properties is a quite difficult issue. This is primarily due to high variability of measuring devices and principles and thus irregular international test standards. Coupled with test methods, there are several procedure-based sources of deviation, which are largely underestimated [24]. Particle sizes were investigated for 13 different conventional practice wood chip samples produced from debarked logs, whole trees, logging residues and wood thinning as well as two special prepared standard samples [23].


For biofuel size classification several methods were tested in an international round robin that was conducted with two standard samples used as reference. For these standard samples the size parameters (i.e. maximum length, width and thickness) were determined by hand measurement using a digital calliper gauge. Furthermore, their weight was determined. Following a consistent length, standard samples were sorted into different size classes and coloured before sending to the round robin laboratories. In addition to the standard samples the conventional wood fuel samples were produced using four different chipper types, such as disc, drum, spiral chipper and shredder.

At the laboratories samples were analysed using four different methods represented by (i) horizontal screens, (ii) vertically vibrating screens as modification of the horizontal screen, (iii) rotary screens as well as (iv) image analysis system. The measures were carried out at different instrumental and procedure variations. For all screenings five progressively round hole diameters are applied.



Horizontal screens Rotating screens (cylindrical rings)

Image analysis (photo-optical)

(instrumental variations: e.g. one-to-three dimensional shaking

at different frequencies and amplitudes; procedure variations:

e.g. screening duration, filling high, moisture of sample)

(procedure variations: e.g. rotating speed, inclination angle, moisture of

sample, uniformity of feeding)

(instrumental variations: e.g. camera position and resolution, recording

modes)

Figure 4: Functioning principles of screening and image analyses used in round robin trials [23, 24, 25]

In all trials about 8 litres of samples were passed. The usually used fractionating criterion is that of length distribution. Each sample was tested in three replications by means of every equipment, except for the image analysis system (five replications). The results of the test methods were compared on the basis of mean particle size and median value. Furthermore, the respective influences of various modes were investigated (i.e. moisture content, frequency, screening duration, feeding rate, rotation speed, inclination angle, mechanical wear as well as sample pre-treatment) [24, 25].


The results revealed that the analysis of biofuels particle size is associated with high measuring uncertainties. This is basically due to the fact, that the tested major measuring principles (i.e. horizontal and rotary screening as well as image analysis) produced results which were largely incompatible. Results acquired from the image analysis system showed highest conformity to the reference values (standard samples). For all horizontal and vertical screening machines the median value of the size distribution (according to particle length) was only between one third to half of the reference median value. This is attributed to the high particle misplacement particularly found in larger fractions. For rotary screening the median particle length is between the results of image analysis and horizontal screening. Based on this, comparable measurements must consistently be made using only one of the three principles, while for the same principle modifications of the equipment type (e.g. different dimensional shaking operations) are usually acceptable.

With regard to influencing factors, for horizontal screening a critical shaking frequency (~ 190 rpm) was identified, while the chosen initial sample volume was found to be less important. For the screening duration a larger effect was observed; a fixed minimum time requirement of about 15 min was identified to be meaningful. For rotary screening the influencing factors are mainly the rotation speed and the inclination angle of the rotating drum. Also the feeding rate and the moisture content of the sample play an important role,



as reflected by the measured differences in the calculated median particle size. Generally, it should be attend to sample preparation due to the possible inhomogeneous moisture content within the sample. Coupled with this, the fixation of a tolerable moisture range is required if results from different fuels and test methods shall be compatible.

However, high reproducibility is only given, when all relevant measuring variables and the influencing factors are carefully considered and standardised accordingly [23, 24, 25].

3.4 Durability and raw density of pellets and briquettes

3.4.1 Objectives

The physical fuel quality of densified biofuels like pellets and briquettes is primarily characterised by durability (i.e. fuel resistance towards shocks and tensions) and particle density (i.e. ratio between mass and volume of a sample that is appropriate to estimate durability). Thus, durability is important with regard to handling, transportation and end-conversion processes where higher durability is required without falling into small pieces and dust [28]. The performance of existing methods have been tested and evaluated under technical and work-efficient aspects for different biofuels representative of the European market. Tested fuels are 15 pellets and 5 briquettes based on mixed wood, hard wood, soft wood as well as on agricultural residues [26, 27].


Several test methods, devices and different procedures were investigated and evaluated by means of round robin tests involving five partners. Following the guidelines, each partner had to determine the moisture content of the appropriate fuels just before the tests.

Durability Regarding durability determination, pellets were tested and compared: (i) using a tumbling device according to the American ASAE standard, and (ii) using a pneumatic tester based on the Austrian ÖNORM standard. The durability of briquettes was determined in a defined rotating drum in different test modes. The number of rotations is investigated in order to define the best appropriate rotating period. After treating in durability testers the samples (of pellets and briquettes) were screened with a sieve. The durability is calculated from the mass of sample remaining on the sieve and expressed as a percentage of the initial mass.

Particle density Methods applied for particle density were (i) the stereometric method as well as methods based on liquid displacement such as (ii) the hydrostatic method and (iii) the buoyancy test method. Stereometric methods are based on measuring of sample dimensions (e.g. diameter, length, width, height) using e.g. a calliper rule or a palmer. The volume of the sample is estimated by calculating the volume of the nearest regular geometrical shape (e.g. cylinder, parallelepiped, cubic). Liquid displacement methods (see Figure 5) are based



on the Archimedes principle (i.e. an object is buoyed with a force equal to the weight of the liquid it displace). The tests were performed for coated (by using paraffin) or non-coated samples in pure water as well as for non-coated samples in a mixture of water wetting agent. Accordingly, five methods were applied. Coupled with this, the increase of the volume, due to the increase of moisture content, was neglected.

Hydrostatic method Buoyancy test method

Figure 5: Liquid displacement methods applied for the determination of particle density [27]

The influence of tested methods and selected fuels on the particle density results was evaluated by calculating mean values and standard deviations (through 15 replications and participating laboratories). Furthermore, the repeatability and the reproducibility were calculated [26, 27, 28].


Durability For the estimation of the briquettes durability the most repeatable and reproducible method is to tumble the briquettes for 105 rotations corresponding to 5 min treatment. Nevertheless, the briquettes durability testing leads to highly variable results. But the variability of the method is influenced by the fuel itself and is smaller for briquettes of high durability. Furthermore, if all the tested briquettes are considered, it seems illusive to reach a higher accuracy than 10 %.

Also for determination of pellets durability, using the tumbling device (ASAE standard) shows better results compared to the pneumatic tester (ÖNORM standard). Though, there is no accurate relation between the results of pellets durability obtained by both devices. It clearly appears that the level of durability influences the variability of results: i.e. the lower the pellet durability, the higher the variability. Taking into account the number of replication needed, an accuracy level of 1 % could be reached in practice with the tumbling device.



Particle density The different results revealed that the hydrostatic and buoyancy methods based on liquid displacement give lower repeatability and reproducibility than the stereometric methods and thus performed better. For all tested methods, it clearly appears that the fuels type influences the variability of the results.

For each analysed parameter the stereometric methods led to higher variability, higher bias between participated laboratories, higher values of repeatability and reproducibility and needed more replications to reach a given exactness.

The liquid displacement methods showed similar standard deviation. The values of repeatability and reproducibility were similar. This is particularly true for briquettes. Methods with addition of wetting agent give higher results than those obtained with paraffin coating methods. In the case of pellets, the buoyancy method using non coated samples and wetting agent mixed with water gives the lowest values in repeatability and reproducibility. However, the choice of the liquid displacement method used may be based on the material available at the laboratory. Thus, e.g. hydrostatic with paraffin coating needs to have a balance that can weight in a range of 10 to 15 kg with an accuracy of 0.1 g. Indeed applying buoyancy without paraffin coating decreases the time of sample preparation but the water has often to be changed because briquettes start to disintegrate very fast.

Generally, it is concluded that for given fuel types, the number of replications needed to reach a given accuracy level is by far smaller than for the others. It is suggested to further determine on which other parameters the number of replications (or the expected accuracy) may be based.

Comparing particle density and durability by using the most accurate methods for both parameters, no relation has been found between those two parameters. Thus, the particle density cannot be used to estimate the durability [26, 27, 28].

Part 1 Chemical tests


4 Chemical tests

For biofuel quality, in addition to physical-mechanical fuel properties also chemical properties are of high relevance. Appropriate test methods were evaluated within this work package (WP III) where about thirteen partners were involved.

Producers, traders and users of solid biofuels have to rely on reliable standardised methods for the determination of biofuels properties. This is particularly true with regard to fuel supply (i.e. handling and storage aspects) and to the utilisation of solid biofuels (i.e. predominantly thermo-chemical conversion in combustion systems), both associated with solid and gaseous emissions. Chemical biofuels characteristics are determined applying standards originally developed for the analysis of coal. Therefore, the exactness of existing procedures and the adaptation for solid biofuels had to be improved or new test methods had to be developed. Thus, this work package was concerned with the selection, evaluation and validation of suitable test methods for the determination of sulphur, chlorine and nitrogen content (Task III.1) and for major and minor elements (Task III.2). The results of these investigations provided the basis for preparing best practice guidelines and in turn were also destined for the respective CEN standards.

4.1 Sulphur, chlorine and nitrogen content

4.1.1 Objectives

To assess biofuel quality it is of high importance to know the concentration of sulphur, chlorine and nitrogen in biofuels. This is particularly true for biomass conversion processes. Contents of sulphur and chlorine in biofuels are of relevance for corrosion and fouling, for emissions of SOx, HCl and PCDD/F as well as for aerosol formation. The concentration of nitrogen causes NOx emissions. Standards defining analytical methods for the determination of these elements are not available for solid biofuels so far. Hence, different approaches and procedures are in use. The individual laboratory methods and system devices applied are originally designed for the analysis of coal samples, whereby certain deviations result. The analysed solid biofuels were selected with the intention to cover the most frequently traded biofuels in Europe and additionally cover a wide range of concentration of investigated elements (S, Cl, N). Testing methods were carried out for coniferous wood without bark, woodchips, bark, hardwood with glue, straw, rape straw, olive residues and herbaceous perennial cynara as well as hemp [36].


Sample preparation and homogeneity tests Depending on delivered biofuel size, sample preparation was performed stepwise using coarse cutting mills equipped with a 10 mm sieve and laboratory cutting mills equipped with tungsten carbide tools and sieves of 1 mm and 0.5 mm. The preparation is followed



by homogeneity tests by means of bomb combustion and determination of chlorine and sulphate using ion chromatography (IC). The results were evaluated with statistical methods. After distribution of the prepared homogeneous samples to the participating laboratories the elemental analysis for S, Cl and N was carried out [36].

Method investigation and improvement For the determination of sulphur and chlorine existing procedures include methods based on (i) combustion (e.g. closed in a oxygen bomb or according to Wickbold, open in a high temperature tube furnace or automated) and (ii) digestion (e.g. high pressure acid in a closed vessel, in a open vessel according to Eschka) as well as (iii) AOX analyser, (iv) water soluble chlorine, (v) XRF (“X-ray fluorescence spectrometry”) and (vi) INAA (“irradiation neutron activation analysis”). Typically carried out subsequently, for S and Cl quantification methods such as IC (“ion chromatography”), ICP (“inductively coupled plasma”) and titration were used [35, 36]. Two standard methods applied for determining the nitrogen content in organic samples were used: (i) automated analysers and (ii) the Kjeldahl method. Initially, the analysis have taken into account all available methods for testing the prepared biofuel samples.

Based on the intermediate test results, not suitable methods for testing solid biofuels were excluded since they are not sensitive enough, difficult to handle or of poor reproducibility and repeatability. The most promising methods were selected to investigate in detail and to get improved. This involves [36]:

• sulphur: combustion in an oxygen bomb and quantification of sulphate by IC as well as acid digestion in closed vessels and quantification by ICP

• chlorine: combustion in an oxygen bomb and quantification of chloride by IC as well as the water soluble chlorine method

• nitrogen: automated analysers and the Kjeldahl method

Sensitivity analysis For the identification of critical points in the analytical procedures, a sensitivity analysis was performed for selected parameter. This is applied for the influences of (i) moisture content, (ii) particle size, (iii) receiving solutions, (iv) oxygen pressure and (v) time for pressure release on the respective elemental determination. Moreover, for the separation of systematic deviations of independent analytical steps (i.e. combustion or digestion followed by quantification), both steps were investigated separately. Therefore, two approaches were performed:

(a) A bomb digestion was prepared by one laboratory and send to all the other laboratories for the concentration analysis of sulphur and chlorine.

(b) A bomb combustion was carried out by all laboratories and the content of sulphur and chlorine was measured by one laboratory.

The developed methods were finally tested using hemp as an additional sample [36].




Concerning sample preparation, for S, Cl and N analysis a particle size of < 1 mm is sufficient in most cases. Smaller particles increase the repeatability but also increase the risk of contaminations with metals (i.e. from the inner materials of the used mills). The reproducibility of chemical analyses is usually improved when larger sample amounts are used (e.g. 1 g for the determination of S and Cl) [34]. Furthermore, in laboratory practise it should be especially considered that (i) the samples are always be within the calibration, and (ii) if low nitrogen concentrations are to be analysed, an increase in sample amount may improve the result.

Besides this, the method evaluation for element analysis led to the conclusions and recommendations given below. Basically, except the Kjeldahl method all of the identified most promising methods named above are recommended for standardisation. They are described in detail in the corresponding best practice guideline and draft standards [35, 36].

Chlorine and sulphur determination Sample combustion in an oxygen bomb and the quantification of sulphate and chloride in the receiving solution is currently the best method. It enables the application of procedures presently standardised on European level.

For chlorine, the determination applying the method of water soluble chlorine led to similar results compared to the bomb combustion method (at least for all untreated biofuels investigated). For samples characterised by a high ash content even higher values were obtained.

In general, the required repeatability and reproducibility can only be obtained when strictly abided to the standard procedures. Solid biofuels that are characterised by low concentrations of sulphur and chlorine are difficult to analyse. Since currently the reproducibility is not satisfactorily, there is more need on research and method improvement. Coupled with this, the INAA method may be of interest for further scientific investigations, as for instance absolute values can be obtained. Furthermore, this method could contribute to improve and validate analytical methods.

Nitrogen determination From the comparison of the obtained results for automated analysers (different and same brands and types) and for the Kjeldahl method the following was revealed.

With respect to certain types or brands of automated analysers available no systematic deviations were found. However, there were differences between the participating laboratories (persons operating the systems). Thus, for the obtained results the operation and especially the calibration is critical. Nevertheless, automated analysers are suitable for the nitrogen determination in solid biofuels and thus, involved in respective draft standardised. Since recognising comparable results for different designs of automated analysers, no specific system design is recommended in the draft standard. However, the applied apparatus should meet the functional requirements. No influence of moisture content on the nitrogen concentration could be found. Hence, it is not necessary to used dried samples which are difficult to handle since they are in general very hygroscopic.



Against the standardised methods developed so far, the work has generally revealed that there is further need on improvement with regard to the test methods, the lack of reference material and laboratory experiences as well as on own laboratory practice. Final recommendations involve to further investigate the chemical fuel concentrations of bromine and iodine (especially e.g. for recovered fuels, fruity biomass and seaweed) [35].

4.2 Determination of major and minor elements

4.2.1 Objectives

For appraising quality of solid biofuels important chemical parameters are the content of major (i.e. (Al, Ca, Fe, K, Mg, Na, P, Si, Ti) and minor (i.e. As, Ba, Cd, Co, Cr, Cu, Hg, Mn, Mo, Ni, Pb, Sb, Tl, V, Zn) elements. Major elements are of key relevance referring to ash melting, deposit and slag formation as well as to corrosion. Minor elements are of special importance for particulate emissions as well as the environmental impact of produced ashes and their subsequent utilisation [32]. Specific determination procedures for these elements in biofuels exist in few cases, but often are associated with considerable deviations between the results of testing laboratories. For the majority of investigations coniferous wood and bark mixtures and wheat straw are employed as test biofuels. Furthermore, olive residues and certified biomass reference materials were used [31].


Sample preparation and homogeneity studies To obtain satisfactory homogeneity spectrometry determinations (XRF and ICP-MS) were performed on three different particle sizes. For the actual tests about 10 kg wood and straw material were milled to homogeneous material of particle sizes < 0.25 mm and filled in bottles of 20 g each. Afterwards, prepared materials were analysed by spectrometry again and statistical F-tests were carried out to check within and between bottle homogeneity.

Method testing and improvement Similar to the approach of Task III.1, in a first instance several advisable analytical methods were tested in order to choose the most suitable methods for further investigations. For this about 16 digestion and 8 determination (i.e. by means of several detection systems) methods were scrutinised. The digestion methods involved wet chemical digestion in closed vessels with different mixtures of HNO3, H2O2, HF, HCl and HClO4, heated conventionally or by microwaves. Furthermore, digestion tests were performed with or without neutralisation by H3BO3 as well as dry-ashing with subsequent dissolving in acids or dry-ashing with subsequent fusion in LiBO2 or Li2B4O7. The wet digestion procedures were followed by determination using (i) atomic absorption spectrometry (AAS) such as FAAS (“flame”), GFAAS (“graphite furnace”), CVAAS (“cold vapour”) and (ii) inductively coupled plasma spectrometry such as ICP-OES (“inductively coupled plasma optical emission spectrometry”) or ICP-MS (“inductively coupled plasma mass spectrometry”). These determinations showed good agreement between the different results for many elements investigated. Other determination methods



applied are hydride generation AAS (“atomic absorption spectrometry”), XRF (“X-ray fluorescence spectrometry”) as well as direct Hg determination.

Based on the results achieved with this procedure, in the second step the optimisation as well as improvement of a wet chemical digestion method was focused investigating the amount of HF required as well as different digestion temperatures (105 °C, 190 °C and 220 °C). For a broader applicability of the examined methods, olive residues as well as a certified biomass reference material were tested in addition to straw. Wood and bark mixture was excluded from these examinations due to restricted time and budget.

Method validation Finally, based on the outcome of previous investigations different digestion methods for major and minor elements were validated. For that accuracy (i.e. precision and trueness) and application ranges as well as estimations of detection limits were examined [31, 32].


In term of sample preparation, particle sizes < 1 mm or < 0.25 mm should provide satisfactory homogeneity to be used for the analyses of wood and bark mixtures or straw. Regarding this, it is noted to pay attention during size reduction in order to avoid contamination from the inner materials of the used mills. Thus, materials of that mill parts having contact with the sample should be chosen depending on the elements to be determined. If, for instance, minor elements such as Cr and Ni have to be determined, it is disadvised to use stainless steel materials, and recommended to use e.g. tungsten carbide or titanium instead. Generally, high-speed mills should not be used due to the higher abrasion rate.

In conclusion of the validation results, for solid biofuel analyses wet digestion with H2O2 / HNO3 / HF / H3BO3 are proved to be the most suitable for the determination of major elements. For minor elements wet digestion using H2O2 /HNO3 /HF are recommended.

Depending on the specific element to be determined, the application ranges revealed the suitability of tested methods for a wide concentration range, including potential concentrations in both natural and contaminated solid biofuels [32]. According to detection limits, the most suitable determination methods are summarised for the different elements in solid biofuels in Table 3.



Table 3: Recommended detection systems for major and minor elements [31, 32]

Detection system Major elements Minor elements

FAASa Ca, Fe, K, Mg, Na, Si Mn, Zn

ICP-OES Al, Ca, Fe, K, Mg, Na, P, Si, Ti Mn, Zn, Ba, Cr, Cu, Mn, Ni, V, Zn

GFAASb - Cd, Cr, Cu, Ni, Pb

ICP-MS P, Ti As, Ba, Cd, Co, Cr, Cu, Mn, Mo, Ni, Pb, Sb, V, Zn

Direct Hg determination - Hg

CVAAS - Hg a because of relatively poor detection limits, Al determination by FAAS is not recommended for solid

biofuels containing low concentrations of this elements b because of relatively poor detection limits, As, Co, Mo, Sb and V determination by GFAAS is not

recommended for solid biofuels containing low concentrations of these elements

Basically, applying XRF detection would be a suitable and fast alternative method for the determination of several major and minor elements. However, due to the requirement of reliable calibration standards, which are not available for solid biofuels so far, XRF systems for element determination presently not be recommended. Thus, the development of such standards is desirable [31].

Part 1 Fuel Quality Assurance


5 Fuel Quality Assurance

Closely linked to the previous work packages, this work package (WP IV) dealt with Quality Assurance systems for the provision of solid biofuels. Six partners were involved in the elaboration of this topic.

5.1 Objectives Standardisations of solid biofuel properties contribute to promote a more widespread use of biofuels by providing a base to facilitate the business of operators within the market. Besides this, a Quality Assurance system that takes the whole provision chain of solid biofuels into account is essential for an increased use and trade of biofuels. This system provides tools to demonstrate that the level of quality demanded by customers is achieved. Hence, activities on standardisation have to be accompanied by the development of a Quality Management (QM) system with the main focus on Quality Assurance (QA) and Quality Control (QC). However, currently no Quality Management system exists in practice in Europe, which fulfils these demands. Therefore, the aim of WP IV was to fill gaps of knowledge concerning Quality Assurance in the field of solid biofuels taking requirements from practice into consideration [11].

This has been achieved in three tasks taking into account that the resulting documents are applicable for operators dealing with solid biofuels:

• Objective of the first Task IV.1 was a review of existing relevant Quality Management systems.

• Emphasis of Task IV.2 was the development, implementation and improvement of a guideline for Quality Assurance by field trials to gain experience from practice. It was aimed to provide a methodology on how to develop and implement a Quality Assurance system (including Quality Control) within a company dealing with solid biofuels.

• The development of a proposal for a standard for Quality Assurance was object of Task IV.3 in order to directly support the Technical Specification on Quality Assurance for solid biofuels.

Considered biofuels are conform with the scope of Technical Specification “Fuel specifications and classes” [13].

5.2 Methodical approach Following the objectives, initially there was an extensive review of existing, relevant QM systems (e.g. in accordance with the international standard ISO 9001:2000) already applied by producers, operators and users of solid biofuels. This review covers solid biofuels and excludes recovered fuels. Existing QM systems were analysed in ten different cases on the basis of a question list in order to identify (i) how QA and QC are currently performed, (ii) what are the main components as well as (iii) what are the needs and demands.



Furthermore, pros and cons of QM systems were discussed. Based on these results conclusions were drawn for the design of a guideline for QM with emphasis on QA and QC for solid biofuels.

The guideline sets out a step-by-step methodology in order to support each operator within a supply chain of solid biofuels to design a manual for Quality Assurance. Draft manuals were implemented and tested out in practical conditions. This was accomplished in “field-trials” at the industrial premises of a range of producers, traders and users of solid biofuels, referred to as host companies. The choice of host companies was well balanced covering main types of European biofuels and conversion processes and considered geographical distribution over Europe. This involves companies that produce refined solid biofuels and companies that buy such up-grade biofuels for end use. Furthermore, a wide range of possible circumstances has to be covered by including all cases between small-scale users (especially domestic, who require high-grade fuels in installations with narrow fuel specifications) and large-scale users (who can benefit from lower-cost fuels by using fuel flexible conversion units). The field trials were divided into three main steps [13]:

(1) to develop a methodology on how to implement a QM system with emphasis on QA and QC in companies, and to write a first draft of a guideline for the development and implementation of such a system in companies

(2) to investigate the implementation in different companies throughout Europe

(3) to develop, evaluate and update the guideline for the development and implementation of QA and QC in companies

Finally, a proposal for a standard for Quality Assurance was elaborated.

5.3 Conclusions and recommendations According to the different tasks, the results were briefly summarised as follows based on [11, 12, 13].

5.3.1 Review of quality systems

The review has clearly shown the need of the adoption of a QM system for solid biofuels with emphasis on QA and QC. Accordingly, aspects such as defining the product quality, traceability, documentation, statistical control, testing and sampling are of importance.

Because ISO 9001 is the most commonly applied Quality Management standard, it is considered meaningful to prepare the guideline for Quality Assurance and Quality Control using the terminology as used in ISO 9000.

Since biofuel product quality is defined in different manners, it is therefore concluded to define product specifications on a case-by-case basis by the market itself, not by prescribing normative fuel parameters by an external organisation. Being of value for biofuel end-users it is recommended to implement a system of tracing back the origin of the biomass raw material used, e.g. by means of delivery notes or batch identification numbers. Linked to this, a QM system for solid biofuels should demand that the different



owners within the entire provision chain are responsible for the documentation and filing of all relevant information on treatment, storing and transport of the biofuel under their specific ownership. In addition, the current owner should be responsible to provide all information back through the provision chain at all times via a batch number.

One important part of a QM system is the design of an appropriate sampling and testing procedure that provides confidence to customers that promised fuel specifications are always met. The costs of implementing such procedures within a biofuel production process depend on the selection of process parameters to be tested and the frequency of testing. In turn, the need for sampling certain process parameters depends on the type of production process, the variability in used raw materials and the allowable product variations. Therefore, the sampling and testing procedure is needed to be custom made.

5.3.2 Implementation of Quality Assurance in companies

The field trials have underlined the need for a guideline with a general methodology applicable by each operator throughout the supply chain of solid biofuels. The guideline is aimed to assist all operators to develop an appropriate manual for Quality Assurance according to their specific needs. It provides a better understanding of the issues, thereby enabling to design appropriate measures to control and assure quality.

Quality Assurance aims to provide confidence that a steady quality is continually achieved in accordance with costumer requirements. However, the supply chains of solid biofuels can consist of different process chains and/or process steps, which can be distributed among different companies or organisational units. In this definition, the customer is the next operator (company or organisational unit) within the supply chain. The customer requirements, i.e. demanded quality, include not just the quality of product but also the quality of the company’s performance. The quality of solid biofuels can be defined in terms of a number of key properties that relate to the suitability of the fuel for a specific use. Therefore, the quality of performance refers to documentation, timing and logistical issues.

The methodology, briefly described hereafter and illustrated in Figure 6, can operators apply to design a suitable QA system for a specific chain. Besides an efficient control of the processes considered, the methodology ensures also the control over the overall provision chain by integration of previous and subsequent process steps of other organisational units. It is recommended to document each implementation to these different steps of the methodology within a specific company in a (site-)specific manual, which can serve as an appropriate tool to illustrate to different parties that all the processes and their interaction are fully under control.



1. Description of process chain

2. Determination of customer requirements

3. Analysis of quality influencing factors

4. Selection of appropriate QA measures

Elaboration of a process (site)-specific manual

Figure 6: Methodology to develop and implement Quality Assurance (QA) [13]

Step 1: Description of process-chain. The first step is to describe the production process in the company including the different sub-processes at a very detailed level. Coupled with this, it should carefully be assessed which steps of the production process are significant for realising an appropriate QA. This is of vital importance for the following drafting of the QA manual.

Step 2: Determination of customer requirements. The position of the process unit(s) considered within the overall provision or process-chain is essential for QA because the customer requirements depend on previous and subsequent process steps. In principle the requirements of the next operator of the overall provision chain have to be fulfilled. Due to a lack of knowledge the fuel end user is unable to provide adequate Technical Specifications of the fuel. Hence, key properties of the final product should be determined either by a standard that specifies which fuels can be used, or by the fuel supplier that through knowledge of the user needs is able to define such specifications. Such an indirect demand is to be understood as requirement from a subsequent step.

Step 3: Analysis of quality influencing factors. After the description of the process unit(s) and the analysis of customers requirements the process should now be examined concerning factors considered to be most influential in terms of fuel quality. The following factors are involved in general in determining fuel quality and refer to the management of the company:

• effectiveness of preliminary inspections of fuel sources

• effectiveness of checking of incoming loads

• appropriateness of applied methods to handle, store and process materials

• quality-control measures adopted (frequency of testing)

• company management and responsibility

• qualification and knowledge of staff

Step 4: Selection of appropriate Quality Assurance measures. In accordance with the results of step 1 to step 3 appropriate QA measures (i.e. measures giving confidence) have to be identified and applied. Aspects to be taken into account are e.g.:

• allocation of responsibilities and elaboration of work instructions



• identification and documentation of Critical Control Points (CCP`s) and application of Quality Control (QC) measures

• training of staff and proper documentation of processes and test results

• failure mode and effect (FME) analysis

The field trials have clearly shown, that all parties who are active in the overall provision chain for solid biofuels (including suppliers of raw materials, producer of biofuels, transport and perhaps one or more retailers) should be integrated within an overall QA system. Based on the methodology described above an appropriate QA system can be implemented easily in a company dealing with solid biofuels, which fulfils those demands. The methodology ensures thereby a system that is simple to operate without undue bureaucracy, and which offers savings in costs.

5.3.3 Proposal for a draft standard for Quality Assurance

The proposal of a standard based upon the Technical Specification (TS) "Solid biofuels - Fuel Quality Assurance" under development in TC 335/WG 2 was aimed to guarantee solid biofuel quality through the whole supply chain and to provide adequate confidence that specified quality requirements are fulfilled. Main stage of this chain from the origin to the delivery of the solid biofuel can be gathered from Figure 7.

Raw material

Indentification and collection of raw material

Production/preparation of solid biofuels

Trade anddelivery of solidbiofuels

Reception of solid biofuel by end-user

End - user

Combustionunit or other conversion unit

Supply chain activities covered by this Technical Specification

Figure 7: Supply chain of solid biofuels [12]

The objective of this document is to serve as a tool to enable efficient trading of biofuels. Thereby the end-user can receive a fuel, which corresponds to his needs, and the producer/supplier can produce a fuel with defined and consistent properties. Besides instructions how to fulfil the requirements of the standard, general information is given about QA and QC explaining the approach and benefits of implementing an adapted QM system. This also involves QA measures (e.g. production requirements, transportation, handling and storage as well as quality declaration) and QC (e.g. specifications of biomass origin, source and of traded from as well as determination properties).

Part 1 Summary


6 Summary

Due to its advantages (like wide availability, well-known with regard to supply and use and contribution to green house gas mitigation) biomass represents the most important source to address European policies increasing the share of renewable energy sources for a more sustainable energy system. For an improved integration of solid biofuels into the energy system a dynamic European market has to be created. Thus, accomplishable biofuel properties as well as procedures to test and control these defined parameters are required. Therefore, besides the elaboration of an overall Quality Assurance system for solid biofuels based among other on sampling and testing methods, the pre-normative research project (BioNorm) was aimed to directly contribute to the ongoing development of European standards for solid biofuels within CEN TC 335 “Solid biofuels”. Within this project the emphasis was laid on the identification and evaluation of the best appropriate sample-, test- and reference methods for the determination of specific fuel properties. Based on the experiences made respective best practice guidelines were compiled. These guidelines and research findings were incorporated for drafting Technical Specifications within the standardisation process.

In each of the different work packages and tasks respectively various kinds of solid biofuels covering a wide scope of fuels were analysed and investigated by carrying out selected methods. The applied methods and principles were initially identified to be basically suitable for a reasonable determination of (i) the number of increments and tests as well as specific increment sizes in terms of sampling and sample reduction, (ii) the physical/mechanical fuel properties (i.e. moisture content and bulk density, ash melting behaviour, particle size distribution, durability and particle density) and (iii) the chemical fuel characteristics (i.e. sulphur, nitrogen and chlorine as well as major and minor elements). The research has two emphasises. It was focused on the investigation of existing methods and equipment (e.g. applied for solid fossil fuels) with regard to their accuracy for solid biofuels and thus their applicability followed by appropriate adaptation and improvements of these methods and laboratory equipment. In addition, it was also focused on the development of new methods. Referring to this, common statistical parameters such as accuracy, reproducibility and repeatability were used for the assessment of the different sampling and testing methods. Besides technical, also work-efficiency, economical and environmental aspects were considered.

Supported by these research results, new methods for the development and implementation of Quality Assurance systems for the entire biofuel supply chain were evaluated based on an initial review of existing systems as well as extensive field trials at several companies.

According to the work content of the BioNorm project, the results of the respective sampling and testing methods are summarised for the range of investigated and analysed solid biofuels in the following (see Table 4).

Part 1 Summary


Table 4: Summary of preferred methods for investigated solid biofuels (to be continued next page)

Origin and source of solid biofuelsa Woody

biomass Herbaceous

biomass Fruit biomass Blends and mixtures

Recommended methods / determination devices / procedures

Best practice

guidelines available

CEN TC 335 Draft Standards available

Sampling

• number and size of increments to take for a bulk sample given for methods determining moisture, ash content, size distribution and chlorine content ∗ GROT and sawdust: heap / conveyor ∗ pellets: heap / bags ∗ straw bales: core / hook

x

Sampling and sample reduction

Sample reduction

• GROTb • sawdust • pellets

• straw bales - - • methods determining moisture, ash content, size distribution, chlorine content

∗ GROT: riffle ∗ sawdust: riffle / coning / long pile ∗ pellets: rotary divider / riffle ∗ straw bales: coning / ling pile / hand

x

Moisture content

• reference methods: ∗ oven drying, freeze drying, xylene distillation, GS-MS method

• rapid methods: ∗ on-site types: thermogravimetric Mettler, capacitive Pandis and Schaller ∗ on-line type: optical MESA

x x

Bulk density

various kinds of • bark • chips • pellets • residues • needles

various kinds of • triticale • grain • pellets • rape cakes

various kinds of • stones/

shells • olive cake

• peat

• cylindrical container shape à 50 litres • standardised shock impact for increasing bulk density • for data comparability correction factor of 0.712 % per 1 % moisture

difference for inconsistent moisture content between samples • neglecting of effects based on moisture content above 25 % (MC)

x x

Ash melting behaviour

various kinds of • bark

• straw • Lucerne

• olive stones -

• ash sample preparation at 550 °C (CEN TS 14775) • quantify melting temperatures: improved DIN and MAF • reference for agglomeration /sintering: CFBA • ash composition and species: SEM-EDS • promising for further ash behaviour but not suited for standard: TGA/SDTA

x

Particle size distribution and dimension

various kinds of • chips

- - -

• for comparability measurements have to consistently made using on of the three methods: ∗ horizontal screening (critical shaking frequency ~ 190 rpm, fixed minimum

time 15 min) ∗ rotary screening ∗ image analysis

x x

Durability • briquettes: tumbling drum for 105 rotations corresponding to 5 min treatment • pellets: tumbling device (ASAE standard) • depending on fuel properties: the lower the durability, the higher the variability

x x

Physical-mechanical tests

Particle density

various kinds of • pellets • briquettes

various kinds of • pellets

- - • methods based on liquid displacement: buoyancy and hydrostatic • choice of method according to the material/devices available at the laboratory • pellets: buoyancy method using non-coated samples and wetting agent mixed

with water

x x

Part 1 Summary


Origin and source of solid biofuelsa Woody

biomass Herbaceous

biomass Fruit biomass Blends and mixtures

Recommended methods / determination devices / procedures

Best practice

guidelines available

CEN TC 335 Draft Standard available

Sulphur, nitrogen and chlorine content

• coniferous wood

• woodchips • bark • hardwood

• wheat straw

• rape straw • cynara • hemp

• olive residues -

• sample preparation to particle sizes of 1 mm • sample amounts of 1 g for sulphur and chlorine • increase of sample amounts in case of low nitrogen concentration • ·determination methods:

∗ combustion in an oxygen bomb and quantification of sulphate and chloride by IC

∗ acid digestion in closed vessels and quantification by ICP for sulphur ∗ water soluble chlorine ∗ automated analysers for nitrogen

x x

Major elements

• sample preparation to particle sizes of < 1 mm or < 0.25 mm • detection systems:

∗ FAAS for Ca, Fe, K, Mg, Na, Si ∗ ICP-OES for Al, Ca, Fe, K, Mg, Na, P, Si, Ti ∗ ICP-MS for P, Ti

Chemical tests

Minor elements

• coniferous wood-bark mixture

• wheat straw

• olive residues - • sample preparation to particle sizes of < 1 mm or < 0.25 mm

• detection systems: ∗ FAAS for Mn, Zn ∗ ICP-OES for Mn, Zn, Ba, Cr, Cu, Mn, Ni, V, Zn ∗ GFAAS for Cd, Cr, Cu, Ni, Pb ∗ ICP-MS for As, Ba, Cd, Co, Cr, Cu, Mn, Mo, Ni, Pb, Sb, V, Zn ∗ Direct Hg determination & CVAAS for Hg

x x

Fuel Quality Assurance considered for all relevant solid biofuels through the entire

supply chain

• development and implementation of Quality Assurance (QA) systems in companies ∗ step 1: description of process chain ∗ step 2: determination of costumer requirements ∗ step 3: analysis of quality influencing factors ∗ step 4: selection of appropriate QA measures

x x

a For more detailed information on biofuels analysed and tested please refer to the respective final reports in Part 2 “Detailed Final Reports” b kind of wood chips

Part 1 Summary


The outcome of the pre-normative work of the BioNorm project includes among methods for sampling and sample reduction, improved and new developed methods and procedures for the determination of physical-mechanical and chemical biofuel properties as well as the development and implementation of a company specific Quality Assurance system. Coupled with this, basic recommendations are with regard to apply both the proposal of a standard/TS and the guideline in practice.

In terms of the country conditions in the NMS/NAS recent trends have shown a continuous increasing international market of solid biofuels that also stimulates the domestic production of refined solid biofuels within these countries and thus increase the acceptance of renewable energy sources. Nowadays, these countries already use biomass (predominantly for domestic heat provision) and have very promising potentials concerning biomass and bioenergy utilisation. However, currently limited experience in utilisation of refined solid biofuels and missing R&D contributes to a lack of solid biofuels standards and Quality Assurance guidelines. In all of the NMS/NAS there are no specific biofuel standards and Quality Assurance implemented yet. Companies that produce refined solid biofuels for export currently apply national standards of the import countries. Referring to this, all NAS/NMS-partners clearly stated that common standards are urgently required for increasing the solid biofuel market. Hence, CEN standards currently being developed need to be quickly adopted.

Among the extensive and substantial outcomes during the pre-normative research and the progress of standardisation it has been acknowledged that there is still the demand on research in all purviews such as of fuel classification and Quality Assurance or sampling and testing methods and procedures. In this context, essential recommendations on further work are:

• to precise and revise data on sampling and testing methods

• to extent sampling and sample preparation to a broader range of solid biofuel materials, i.e. less common biomass with high variations in fuel properties such as olive and grape residues, shells and stones of fruit biomass as well as road side green

• to extent testing on further fuel parameters (e.g. particle size and shape factors, bridging behaviour, fuel impurities, fungi release)

• to improve and develop reference test methods and rapid test methods (e.g. for ash melting behaviour, chlorine potassium, nitrogen and sulphur as well as heavy metals and with regard to reference materials)

• to investigate aspects of Quality Planning and Quality Control on interactions of the various quality related activities in the quality policy of a company as well as to adapt measures of with regard to Quality Management systems

Moreover, further cooperation is recommended within the European countries, that already used biomass efficiently and their legislation and biofuels standards are harmonious developed.

Part 1 Summary


Finally, the outcomes reveal that BioNorm has crucially contributed to clarify important issues and aspects associated with the ongoing European standardisation process for solid biofuels. This is in particular true referring to biofuel terminology and specification, appropriate procedure and methods for sampling and testing of fuel properties as well as the importance of company specific Quality Management systems. Furthermore, a basis for research exchange with the NMS/NAS has been established that will be further delved and extended in the future. However, BioNorm has also point out the urgent need on further investigations and method development. Therefore, continuing R&D support of the ongoing standardisation activities will be of high importance to close the gaps and to ensure that the European market for solid biofuels and thus the biomass industry will continue to grow rapidly in the future helped by acceptable standards.

Part 1 References


References

1 Kaltschmitt, M. & Hein, M. (2004): BioNorm Second Progress Report. Institute for Energy and Environment, Leipzig, Germany, February 2004

2 Hein, M. & Kaltschmitt, M. (2004): BioNorm - A Project to Support the Ongoing European Standardisation Process. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

3 Bodlund, B. & Sjöberg, L. (2004): Standardisation as a Measure to Increase the Markets - Status and Prospects. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

4 Kaltschmitt, M. & Hein, M. (2003): Mid-term report of the BioNorm Project. Institute for Energy and Environment, Leipzig, Germany, August 2003

5 Sym, R. (2004): Methods for Sampling and Sample Reduction of Solid Biofuels - RTD results. Proceedings of Conference "Standardisation of Solid Biofuels", Leipzig, Germany, October 6-7, 2004

6 Sym, R. (2004): The results of WP1 “Sampling and sample reduction”. Presentations of Conference "Standardisation of Solid Biofuels", Leipzig, Germany, October 6-7, 2004

7 Willart, W. (2004): Methods of Sampling and Sample Reduction and for Preparing Sampling Plans and Sample Certificates – Status of Standardisation. Proceedings of Conference "Standardisation of Solid Biofuels", Leipzig, Germany, October 6-7, 2004

8 Sym, R. & Preddy, S. (2005): WP I – Sampling and sample reduction. BioNorm Final Report, January 2005

9 Pike, Dr. D.C. et al (2005): WP I – Sampling results. Annex R of the BioNorm Final Report, January 2005

10 Pike, Dr. D.C. et al (2005): WP I – Sample reduction results. Annex S of the BioNorm Final Report, January 2005

11 Langheinrich, C. & Kaltschmitt, M. (2004): Implementation and Application of Quality Assurance Systems. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

12 Valtanen, J.; Alakangas, E. & Levlin, J.-E. (2004): Fuel Quality Assurance. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

13 Langheinrich, C. (2005): WP IV – Fuel Quality Assurance. BioNorm Final Report, January 2005

Part 1 References


14 Daugberg Jensen, P.; Burvall, J. & Samuelsson, R. (2005): WP II.1a – Deliverable D4, Final report reference moisture content and rapid moisture content determination methods. BioNorm Final Report, January 2005

15 Daugberg Jensen, P.; Samuelsson, R.; Jirjis, R.; Burvall, J.; Hartmann, H.; Böhm, T.; Temmermann, M. & Rabier, F. (2004): Moisture Content - RTD Results and Status of the Standardisation. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

16 Samuelsson, R. (2004): Moisture Content - RTD Results and Status of the Standardisation. Presentations of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

17 Hartmann, H. & Böhm, T. (2005): WP II.1b – Bulk density determination. BioNorm Final Report, January 2005

18 Böhm, T.; Hartmann, H.; Daugberg Jensen, P.; Temmermann, M.; Rabier, F.; Jirjis, R.; Hersener, J.-L. & Rathbauer, J. (2004): Bulk Density - RTD Results and Status of the Standardisation. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

19 Hartmann, H. (2004): Physical-Mechanical Fuel Properties - Significance and Impacts. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

20 Hjuler, K. & Larfeldt, J. (2004): WP II. 2 - Ash melting behaviour. BioNorm Final Report, December 2004

21 Hjuler, K. & Larfeldt, J. (2004): Ash Melting Behaviour - RTD Results. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

22 Hofbauer, H. (2004): Ash melting behaviour - Status of standardisation. Presentations of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

23 Hartmann, H. & Böhm, T. (2005): WP II.3 - Particle size distribution and dimension. BioNorm Final Report, January 2005

24 Hartmann, H.; Böhm, T.; Daugberg Jensen, P.; Temmermann, M.; Rabier, F.; Golser, M. & Herzog, P. (2004): Size Classification - RTD Results and Status of the Standardisation. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

25 Hartmann, H.; Böhm, T.; Daugberg Jensen, P.; Temmermann, M.; Rabier, F.; Golser, M. & Herzog, P. (2004): Size classification - RTD results and status of the standardisation. Presentation of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

Part 1 References


26 Rabier, F. & Temmermann, M. (2004): WP II.4 - Durability and particle density for pellets and briquettes. BioNorm Final Report, December 2004

27 Temmermann, M.; Rabier, F.; Daugberg Jensen, P.; Hartmann, H.; Böhm, T.; Rathbauer, J.; Carrasco, J. & Fernandez, M. (2004): Particle density of pellets and briquettes - RTD results and status of standardisation. Presentations of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

28 Temmermann, M.; Rabier, F.; Daugberg Jensen, P.; Hartmann, H.; Böhm, T.; Rathbauer, J.; Carrasco, J. & Fernandez, M. (2004): Durability of Pellets and Briquettes - RTD Results and Status of the Standardisation. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

29 Obernberger, I.; Brunner, T. & Bärnthaler, G. (2004): Chemical Fuel Properties – Significance and Impact. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

30 Englisch, M.; Haraldsson, C.; Bakker, F.; Westbourg, S.; Thomsen, E.; Niebergall, K.; Versterainen, R.; Carrasco, J. & Agrifiotis, C. (2004): Determination of Total Sulphur and Total Chlorine - RTD Results and Status of the Standardisation. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

31 Bärnthaler, G. & Obernberger, I. (2005): WP III.2 - Determination of major and minor elements. BioNorm Final Report, January 2005

32 Bärnthaler, G.; Zischka, M; Haraldsson, C. & Obernberger, I. (2004): Determination of Content of Major and Minor Elements - RTD Results. Proceedings of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

33 Bärnthaler, G. (2004): Determination of major and minor element contents in solid biofuels. Presentations of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

34 Englisch, M. & Bärnthaler, G. (2004): WP III - Preparation of analysis samples. Presentations of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

35 Englisch, M. (2004): Task III.1 - Determination of total sulphur and total chlorine. Presentations of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

36 Englisch, M. (2005): WP III.1 - Report on Chemical Tests - Sulphur, Chlorine and Nitrogen. BioNorm Final Report, March 2005 (Draft Version)

37 Kaltschmitt, M. & Bodlund, B. (2004): Open questions of standardisation in EU-25. Presentations of the Conference “Standardisation of Solid Biofuels”, Leipzig, Germany, October 6-7, 2004

Part 2 Detailed Final Reports



Table of contents

Part 2.1 Sampling and Sample Reduction ................................................................... 1 1 Summary................................................................................................................. 1 2 Objectives ............................................................................................................... 2 3 Description of fuels investigated ............................................................................ 2 4 Description and discussion of results ..................................................................... 2

4.1 Sampling................................................................................................................. 2 4.2 Sample reduction .................................................................................................... 6

5 Scientific conclusions............................................................................................. 7 6 Recommendations .................................................................................................. 7 7 References .............................................................................................................. 8 8 Annexes .................................................................................................................. 8

Part 2.2 Physical / Mechanical Tests ............................................................................ 9

Part 2.2.1 Task II.1a - Moisture Content..................................................................... 9 1 Summary................................................................................................................. 9 2 Objectives ............................................................................................................. 10 3 Description of fuels investigated .......................................................................... 11 4 Description and discussion of results ................................................................... 12

4.1 Reference test methods......................................................................................... 12 4.1.1 Total moisture .............................................................................................. 13 4.1.2 Moisture in the analysis sample ................................................................... 19 4.1.3 Volatile compounds ..................................................................................... 21

4.2 Work plan for rapid on-site test methods ............................................................. 23 4.3 Results of rapid moisture testing .......................................................................... 25

4.3.1 Fuel type....................................................................................................... 25 4.3.2 Moisture level............................................................................................... 26 4.3.3 Thermogravimetric method.......................................................................... 27 4.3.4 Capacitive method........................................................................................ 28 4.3.5 Optical method ............................................................................................. 30

5 Scientific conclusion ............................................................................................ 32 5.1 Reference tests...................................................................................................... 32 5.2 Rapid-MC tests..................................................................................................... 32

6 Recommendations ................................................................................................ 33 7 References ............................................................................................................ 34

Part 2.2.2 Task II.1b - Bulk Density Determination................................................. 35 1 Summary............................................................................................................... 35 2 Objectives ............................................................................................................. 36 3 Description of fuels investigated .......................................................................... 36 4 Description and discussion of results ................................................................... 38

4.1 Tested methods and influences............................................................................. 38 4.1.1 Container comparison .................................................................................. 38 4.1.2 Shock impact ................................................................................................ 38 4.1.3 Moisture content influences ......................................................................... 39

4.2 Work plan ............................................................................................................. 39



4.3 Data Evaluation .................................................................................................... 39 4.4 Results and discussion.......................................................................................... 41

4.4.1 Influence of container shape and size .......................................................... 41 4.4.2 Influence of shock impact ............................................................................ 42 4.4.3 Effects on repeatability ................................................................................ 43 4.4.4 Influence of moisture content....................................................................... 46

5 Scientific conclusions........................................................................................... 47 6 Recommendations ................................................................................................ 47 7 References ............................................................................................................ 48

Part 2.2.3 Task II.2 - Ash Melting Behaviour ........................................................... 49 1 Abstract................................................................................................................. 49 2 Introduction .......................................................................................................... 50 3 Methods included ................................................................................................. 50

3.1 Improved DIN method ......................................................................................... 51 3.2 MAF ..................................................................................................................... 53 3.3 TG/DTG ............................................................................................................... 55 3.4 Bed sintering methods .......................................................................................... 57 3.5 SEM-EDS and XRF ............................................................................................. 59 3.6 Fuels ..................................................................................................................... 60 3.7 Synthetic ash......................................................................................................... 63

4 Results .................................................................................................................. 64 4.1 Synthetic samples ................................................................................................. 64 4.2 Ash samples.......................................................................................................... 69

4.2.1 DIN............................................................................................................... 69 4.2.2 MAF ............................................................................................................. 70 4.2.3 TGA/DTG .................................................................................................... 71 4.2.4 Bed sintering methods.................................................................................. 71

5 Discussion and comparison .................................................................................. 72 6 Conclusions .......................................................................................................... 74 7 References ............................................................................................................ 75

Part 2.2.4 Task II.3 - Particle Size Distribution and Dimension ............................. 76 1 Summary............................................................................................................... 76 2 Objectives ............................................................................................................. 77 3 Description of fuels investigated .......................................................................... 77 4 Description and discussion of results ................................................................... 78

4.1 Tested equipment.................................................................................................. 78 4.1.1 Horizontal screening .................................................................................... 79 4.1.2 Vertical vibration screening ......................................................................... 79 4.1.3 Rotary screening........................................................................................... 80 4.1.4 Image analysis .............................................................................................. 80

4.2 Round robin trials (equipment comparison)......................................................... 80 4.3 Influencing factors................................................................................................ 85

4.3.1 Influence of shaking frequency (horizontal screening)................................ 85 4.3.2 Influence of sample volume (horizontal screening)..................................... 86 4.3.3 Influence of the screening duration (horizontal screening).......................... 87 4.3.4 Influence of samples moisture content......................................................... 88



4.3.5 Influence of inclination angle and rotation speed (rotary screening)........... 90 4.3.6 Influence of feeding rate (rotary screening)................................................. 91 4.3.7 Influence of mechanical wear (rotary screening)......................................... 92 4.3.8 Influence of sample drying........................................................................... 92

5 Scientific conclusions........................................................................................... 93 6 Recommendations ................................................................................................ 94 7 Acknowledgements .............................................................................................. 94 8 References ............................................................................................................ 95

Part 2.2.5 Task II.4 - Durability and Raw Density of Pellets and Briquettes ........ 96 1 Summary............................................................................................................... 96 2 Objectives ............................................................................................................. 97 3 Description of fuels investigated .......................................................................... 97 4 Description and discussion of results ................................................................... 99

4.1 Particle density ..................................................................................................... 99 4.1.1 Tested Methods ............................................................................................ 99 4.1.2 Analysed parameters .................................................................................. 102 4.1.3 Results and discussion................................................................................ 102

4.2 Durability............................................................................................................ 106 4.2.1 Briquettes durability test ............................................................................ 107 4.2.2 Pellets durability test .................................................................................. 107 4.2.3 Results and discussion................................................................................ 108

4.3 Relation between durability and particle density ............................................... 114 4.3.1 Briquette ..................................................................................................... 114 4.3.2 Pellet........................................................................................................... 115

5 Scientific conclusions......................................................................................... 115 5.1 Particle density ................................................................................................... 115 5.2 Durability............................................................................................................ 117 5.3 Relation between particle density and durability ............................................... 118

6 Recommendations .............................................................................................. 118 7 Acknowledgements ............................................................................................ 118 8 References .......................................................................................................... 119 9 Glossary.............................................................................................................. 120

Part 2.3 Chemical Tests ............................................................................................. 121

Part 2.3.1 Task III.1 - Sulphur, chlorine and nitrogen content............................. 121 1 Summary............................................................................................................. 121 2 Objectives ........................................................................................................... 122 3 Selection and preparation of the biofuel samples............................................... 123 4 Description of experiments and results .............................................................. 126

4.1 Methods for the determination of sulphur, chlorine and nitrogen in solid biofuels 126

4.2 Sequence of experiments.................................................................................... 128 4.3 Results of the method evaluation for chlorine analysis...................................... 130 4.4 Results of the method evaluation for sulphur analysis....................................... 131 4.5 Results of the method evaluation for nitrogen analysis...................................... 132 4.6 Performance of the method for total sulphur and total chlorine......................... 134



4.7 Results of the sensitivity analysis for the bomb method .................................... 136 4.7.1 Oxygen pressure......................................................................................... 136 4.7.2 Influence of the pressure release time ........................................................ 137 4.7.3 Amount of water in the bomb and alternative receiving solutions ............ 138 4.7.4 Influence of sulphur and chlorine combined in the combustion residues .. 139 4.7.5 Influence of analysis sample particle size .................................................. 139

4.8 Chlorine determination using the Eschka method.............................................. 140 4.9 Other methods for chlorine and sulphur (not investigated in detail).................. 140

4.9.1 Neutron activation analysis (reference method)......................................... 140 4.9.2 X-ray fluorescence analysis (XRF) ............................................................ 141

5 Conclusions and recommendations .................................................................... 141 5.1 Chlorine and sulphur determination ................................................................... 141 5.2 Nitrogen determination....................................................................................... 141

6 References .......................................................................................................... 143 7 Annex ................................................................................................................. 144

7.1 Chlorine determination using the Eschka method.............................................. 164 7.1.1 Background ................................................................................................ 164 7.1.2 Comparison of the Eschka methods used by FORCE and CIEMAT......... 164 7.1.3 Follow-up tests ........................................................................................... 165 7.1.4 Results ........................................................................................................ 167

Part 2.3.2 Task III.2 – Determination of major and minor elements ................... 168 1 Summary............................................................................................................. 168 2 Objectives ........................................................................................................... 169 3 Description of fuels investigated ........................................................................ 170 4 Description and discussion of results ................................................................. 170

4.1 Sub-task III.2.1 “Definition of the state-of-the-art” ........................................... 170 4.2 Subtask III.2.2 “Sample homogenisation and homogeneity tests”..................... 170 4.3 Subtask III.2.3 “Method testing and improvement – part 1” ............................. 173 4.4 Subtask III.2.3 “Method testing and improvement – part 2” ............................. 174 4.5 Subtask III.2.4 “Method comparison and validation” ........................................ 177

4.5.1 Definitions, purpose and calculation of validation parameters .................. 177 4.5.2 Results and conclusions from method validation....................................... 187

4.6 Subtask III.2.5 “Method evaluation and compilation of best practice guidelines” 188

5 Scientific conclusions......................................................................................... 188 6 Recommendations .............................................................................................. 189 7 References .......................................................................................................... 191 8 Glossary.............................................................................................................. 192 9 Annex ................................................................................................................. 194

Part 2.4 Fuel Quality Assurance............................................................................... 208 1 Summary............................................................................................................. 208 2 Objectives ........................................................................................................... 209 3 Description of fuels investigated ........................................................................ 210 4 Description and discussion of results ................................................................. 211

4.1 Review of quality systems.................................................................................. 211 4.1.1 Approach .................................................................................................... 211


BioNorm Project - Final Technical Report VI

4.1.2 Results ........................................................................................................ 212 4.2 The implementation of Quality Assurance in companies – field trials .............. 214

4.2.1 Approach .................................................................................................... 214 4.2.2 Results ........................................................................................................ 215 4.2.3 Conclusions about the field trials ............................................................... 221

4.3 Proposal for a draft standard for Quality Assurance .......................................... 222 4.3.1 Cooperation with WG2 .............................................................................. 222 4.3.2 Elements of the proposal of a standard ...................................................... 222

5 Scientific conclusions......................................................................................... 224 6 Recommendations .............................................................................................. 225 7 Acknowledgements ............................................................................................ 225 8 References .......................................................................................................... 225 9 Glossary.............................................................................................................. 226

Part 2.5 Research exchange with NMS/NAS ........................................................... 227 1 Summary............................................................................................................. 227 2 Objectives ........................................................................................................... 228 3 Description of fuels investigated ........................................................................ 228 4 Description and discussion of results ................................................................. 228

4.1 Country Reports.................................................................................................. 229 4.1.1 Market Situation for Solid Biofuels in the NMS/NAS .............................. 229 4.1.2 Bulgaria ...................................................................................................... 232 4.1.3 Czech Republic .......................................................................................... 232 4.1.4 Latvia.......................................................................................................... 233 4.1.5 Lithuania .................................................................................................... 235 4.1.6 Poland......................................................................................................... 236 4.1.7 Hungary...................................................................................................... 237

4.2 National Platforms.............................................................................................. 238 5 Scientific conclusions and recommendations..................................................... 239 6 Acknowledgements ............................................................................................ 239 7 References .......................................................................................................... 240 8 Glossary.............................................................................................................. 241

Part 2 Sampling and Sample Reduction


Part 2.1 Sampling and Sample Reduction

Final report prepared by: Roger Sym1), S. Preddy2) Contributing co-authors: Kathrine Hansen2), Helle Junker2), Anna-Karin Nilsson4), Jan Burvall3), Jaap Koppejan3), Max Nitschke2), Dr. D.C. Pike1) 1) Green Land Reclamation Ltd, United Kingdom 2) Tech-wise, Elsam A/S, Denmark 3) Swedish University of Agricultural Sciences, Sweden 4) Skellefteå Kraft AB, Sweden

1 Summary

Work-package 1 (WP1) of the BioNorm Project, which started in January, 2002 and will end in December, 2004, will provide information to support draft European Standards (EN) for sampling and sample-reduction for solid biofuels, which is the responsibility of CEN Technical Committee CEN/TC335. Under its mandated programme agreed with the European Commission, TC335 has speedily to produce interim Technical Specifications (TS). Working Group 3 (WG3) of TC335 has already prepared draft TS for sampling and sample-reduction, but without having the benefit of data from statistically designed experiments carried out on a range of solid biofuels. WP1 will provide such data, which can be expected to be used when the TS are upgraded to full EN.

The partners in WP1 are two British companies: Green Land (co-ordinator) and Signalsfromnoise (statistical consultant); two Swedish organisations: Skellefteå Kraft and the Unit of Biomass Technology and Chemistry at the Swedish Agricultural University, Umeå; and two parts of the Danish electricity-utility Elsam, including its technical arm, Techwise.

The purpose of the experimental work on sampling is to determine the size and number of sample-increments for a range of woody biofuels (sawdust, wood-chips and wood-pellets) taken at two sampling-points in each case, and to compare two methods of obtaining increments from straw-bales. The purpose of the work on sample-reduction is to compare the efficacy of various methods.

The outcomes of the work include recommendations for methods of sampling and sample-reduction for the several combinations of point of sampling, material and proposed test-method. Reference is made to further work that could increase the value of this part of the BIONORM Project.



2 Objectives

The aim of WPI is to provide supporting information to CEN/TC 335 ‘Solid Biofuels’ Working Group 3 ‘sampling and sample-reduction’, to improve the reliability of the draft European Standards for solid biofuels.

The first objective is to assess the bias that may be introduced when samples of solid biofuels are taken from containers or stockpiles instead of moving streams, and to define the number and size of increments needed to provide a representative bulk sample.

The second objective is to assess the variations introduced when reducing the size of samples of solid biofuels to form suitable test portions and to determine the usefulness of a range of practical sample-reduction methods.

3 Description of fuels investigated

The fuels investigated were sawdust, wood-pellets, GROT, and straw. These four materials were chosen for their importance in the European solid biofuel market. The woody materials represent the common forms of tradeable solid biofuels in Europe, and straw is an important material in many applications.

The sampling and sample-reduction experiments on woody materials were all carried out at Skellefteå, and the associated testing was undertaken at the Swedish Agricultural University, Umeå. Both of those cities are on the Baltic coast of northern Sweden, a part of Europe where wood is extensively used as fuel.

Skellefteå Kraft, a large municipal utility-company, owns and operates a large (36 MWe) biomass-fuelled power station at Hedensbyn in its home city, which is integrated with a plant that produces over 100,000 tonnes a year of wood-pellets. The raw materials for the power station include sawdust from about 60 large sawmills, some other forestry-residues and peat. Among the other residues are chips produced from the branches and tops of felled trees. In the Swedish language, that material is called “GROT”. So it was relatively easy to find convenient sources of sawdust, GROT and pellets at Skellefteå.

The sampling and sample-reduction experiments on straw-bales were all carried out at Elsam’s premises in Denmark, where straw is an important fuel

4 Description and discussion of results

4.1 Sampling

Task I.1 comprised a series of increment trials on the four chosen materials, with three relevant test-methods per material.

The experiment was designed to allow the relative bias of different sampling methods and the influence on sampling variability of different increment sizes to be assessed, and also



to allow the number of sampling increments required to give a satisfactory level of sampling variability to be estimated.

Two methods of sampling were applied to each material in order that the relative bias of methods can be investigated. The straw bales were sampled using a coring machine and a hook. The GROT and sawdust were sampled from the heap formed by tipping a lorry-load (using a shovel) and from a stopped conveyor (using a sampling frame). The wood-pellets were sampled from the heap formed by tipping a lorry-load into a hopper (using a shovel) and from bags taken from a conveyor (using a riffle to sub-divide the contents of a bag). Photographs of the methods used can be found in the Annexes.

With the GROT, sawdust and wood-pellets, increments of three different sizes were taken so that the influence of increment size on the variability due to sampling can be investigated.

The test-methods applied to the materials were determinations of moisture-content, ash-content, particle-size distribution, or chloride-content, because of their importance in practice and their ability to show the effects of different sampling methods.

With all four materials the increments were divided into two sub-samples so that the variability due to testing (i.e. repeatability) can be estimated.

Five replicates were performed of each experiment, i.e. five straw bales, and five lorry loads each of GROT, sawdust, and wood-pellets.

Table 1: Relative bias of sampling methods

Evidence of a relative bias between the two methods Material Sampling method

Moisture Ash Chloride Particle size Core No No No Straw bales Hook No No No Heap No No Yes GROT Conveyor No No Yes Heap No No Yes Sawdust Conveyor No No Yes Heap No No Yes Pellets Bags No No Yes

Table 2: Recommended numbers of increments for sampling straw bales

Test method Sampling method


Relative error due to

sampling [%]

Number of determin-

ations


testing [%]

Total relative error sampling

and testing [%]

Core 5* 11% 1* 4% 11% Moisture Hook 5* 11% 1* 4% 11% Core 5* 8% 1* 4% 9% Ash Hook 5* 7% 1* 5% 8% Core 5* 10% 1* 2% 10% Chloride Hook 5* 9% 1* 2% 10%

* Recommended options



Table 3: Recommended numbers and sizes of increments for sampling GROT


Increment size



sampling

Number of determin-

ations


testing

Total relative error

sampling and testing

Heap 20 litre 5 2% 1 1% 2% Conveyor 20 litre 5 2% 1 1% 3% Heap 4 litre 5 2% 1 2% 3% Conveyor 4 litre 5 2% 1 2% 2% Heap 1 litre* 5* 3% 1* 3% 4%

Moisture

Conveyor 1 litre* 5* 2% 1* 3% 4% Heap 20 litre 10 9% 4 9% 12% Conveyor 20 litre 10 8% 4 7% 11% Heap 4 litre 10 9% 4 10% 13% Conveyor 4 litre 10 6% 4 6% 9% Heap 1 litre* 10* 9% 4* 13% 15%

Ash

Conveyor 1 litre* 10* 9% 4* 9% 15% Heap 20 litre* 5* 5% 4* 5% 7% Conveyor 20 litre* 5* 4% 4* 7% 8% Heap 4 litre 5 7% 4 12% 14% Conveyor 4 litre 5 8% 4 12% 14% Heap 1 litre 5 12% 4 18% 22%

Passing 64.0 mm sieve

Conveyor 1 litre 5 10% 4 10% 14% Heap 20 litre* 5* 6% 4* 2% 7% Conveyor 20 litre* 5* 5% 4* 2% 5% Heap 4 litre 5 5% 4 2% 6% Conveyor 4 litre 5 4% 4 3% 5% Heap 1 litre 5 6% 4 3% 6%


Conveyor 1 litre 5 6% 4 2% 6%

(1) Range of relative errors for 4.00mm to 64.0mm sieves.



Table 4: Recommended numbers and sizes of increments for sampling sawdust.


Increment size



sampling

Number of determin-

ations


testing



Heap 5 litre 2 2% 1 1% 2% Conveyor 5 litre 2 2% 1 1% 2% Heap 1 litre 2 3% 1 1% 3% Conveyor 1 litre 2 2% 1 1% 2% Heap 0.2 litre 2* 2% 1* 1% 2%

Moisture

Conveyor 0.2 litre 2* 2% 1* 1% 2% Heap 5 litre 5 4% 2 4% 5% Conveyor 5 litre 5 4% 2 5% 6% Heap 1 litre 5 12% 2 8% 15% Conveyor 1 litre 5 3% 2 5% 6% Heap 0.2 litre 5* 3% 2* 5% 6%

Ash

Conveyor 0.2 litre 5* 7% 2* 7% 10% Heap 5 litre 5* 2% 2* 3% 3% Conveyor 5 litre 5* 2% 2* 3% 4% Heap 1 litre 5 3% 2 4% 5% Conveyor 1 litre 5 2% 2 2% 3% Heap 0.2 litre 5 3% 2 5% 6%


Conveyor 0.2 litre 5 4% 2 4% 6% Heap 5 litre 5* 1% 2* 1% 1% Conveyor 5 litre 5* 1% 2* 1% 2% Heap 1 litre 5 1% 2 1% 2% Conveyor 1 litre 5 1% 2 1% 1% Heap 0.2 litre 5 1% 2 1% 2%


Conveyor 0.2 litre 5 2% 2 1% 2% * Recommended options



Table 5: Recommended numbers and sizes of increments for sampling pellets.

Test method

Sampling method

Increment size



sampling

Number of determin-

ations


testing



Heap 4 kg 5 2% 1 1% 2% Bags 4 kg 5 3% 1 1% 3% Heap 1 kg 5 2% 1 1% 2% Bags 1 kg 5 2% 1 1% 2% Heap 0.25 kg 5* 2% 1* 1% 2%

Moisture

Bags 0.25 kg 5* 2% 1* 1% 2% Heap 4 kg 5 3% 2 4% 5% Bags 4 kg 5 1% 2 4% 4% Heap 1 kg 5 2% 2 3% 3% Bags 1 kg 5 2% 2 3% 3% Heap 0.25 kg 5* 1% 2* 4% 4%

Ash

Bags 0.25 kg 5* 4% 2* 6% 7% Heap 4 kg 1 6% 2 1% 6% Bags 4 kg 1 1% 2 1% 1% Heap 1 kg 1* 3% 2* 1% 3% Bags 1 kg 1* 1% 2* 1% 2% Heap 0.25 kg 1 5% 2 1% 5%


Bags 0.25 kg 1 3% 2 1% 3% Heap 4 kg 1 1% 2 1% 1% Bags 4 kg 1 1% 2 1% 1% Heap 1 kg 1* 1% 2* 1% 1% Bags 1 kg 1* 1% 2* 1% 1% Heap 0.25 kg 1 1% 2 1% 2%


Bags 0.25 kg 1 1% 2 1% 1% * Recommended options

The Task report on the sampling trials (Deliverable I.1.D4) can be found in the Annexes.

4.2 Sample reduction

Task I.2 comprised a set of experiments that were designed to compare the variabilities achieved when using different methods of sample-reduction.

The same four materials were used in the sample-reduction experiments as in the sampling experiments.

Three methods of sample-reduction were applied to each material. Hand sampling, coning and quartering, and the long pile method were applied to the straw. Riffling, coning and quartering, and the long pile methods were applied to the GROT and the sawdust. Riffling, a rotary divider, and the long pile methods were applied to the wood-pellets. Photographs of each of the methods are located in the Annexes.

Two variations of the hand sampling method were tried with the straw. In Hand 1 the straw was not coarse cut prior to sample-reduction, and Hand 2 in which it was.

The same test-methods were applied to the materials as in the sampling experiments. Each experiment involved dividing a sample of material into sixteen sub-samples, by repeated action of the method if necessary.



The statistic chosen to describe the variability achieved in a sample-reduction experiment is a relative error, as in the sampling experiment. This is the standard deviation of the sixteen results expressed as a percentage of the corresponding average value. It can be called an estimate of repeatability.

Table 6: Preferred methods of sample reduction.

Moisture Ash Chloride Particle size distribution

GROT Riffle Riffle Riffle Pellets Rotary* Riffle Rotary* Straw Coning/Long pile Hand 2 Hand 2 Sawdust Riffle/Coning/Long pile Coning Coning * or Riffle if no rotary divider is available

The Task report on the sample-reduction trials (Deliverable I.2.D4) can be found in the Annexes.

5 Scientific conclusions

The experimental work has allowed the relative bias of different sampling methods and the influence on sampling variability of different increment sizes to be assessed, and has also allowed the number of sampling increments required to give a satisfactory level of sampling variability to be estimated, for the materials included in the experiment (GROT, sawdust, wood-pellets, and straw).

None of the methods of sample-reduction used in the experiments gave disastrous results, so no method can be ruled out from use in practice on the basis of these experiments.

6 Recommendations

The results of the experiments should be included in the appropriate CEN Standards on sampling of solid biofuels, so that users of the Standards can use the information to help guide their choice of the number and size of sampling increments required in their applications.

All the methods of sample-reduction used in the experiments should be included in the CEN Standard on sample-reduction of solid biofuels, along with the preferred methods listed in Table 2.1 above.

When a sample is being taken for the purpose of determination of moisture-content, sample-reduction should be avoided, as moisture will be lost and the result of the test will be affected.



7 References

1 CEN 2002a. Solid biofuels – Sampling – Part 1 Methods for sampling. CEN/TC335/WG3 working document N41.

2 CEN 2002b. Solid biofuels – Sampling – Part 2 Method for sampling particulate material delivered in lorries. CEN/TC335/WG3 working document N42.

3 CEN 2002c. Solid biofuels – Sampling – Part 3 Method for preparing sampling plans and sampling certificates. CEN/TC335/WG3 working document N43.

4 CEN 2002d. Solid biofuels – Method for sample reduction. CEN/TC335/WG3 working document N44.

8 Annexes

The detailed descriptions of the work and of the results are contained in the following Annexes. Because of practicability for these Annexes please refer to the “Deliverables”.

• Annex A Definitions

• Annex B Biofuels used in sampling and sample reduction experiments

• Annex C Sampling experiments on wood chips

• Annex D Sampling experiments on sawdust

• Annex E Sampling experiments on pellets

• Annex F Sampling experiments on straw

• Annex G Testing samples of wood chips

• Annex H Testing samples of sawdust

• Annex I Testing samples of pellets

• Annex J Testing samples of straw

• Annex K Sample reduction experiments on wood chips

• Annex L Sample reduction experiments on sawdust

• Annex M Sample reduction experiments on pellets

• Annex N Sample reduction experiments on straw

• Annex O Statistical analysis of data from sampling experiments

• Annex P Data from sampling experiments

• Annex Q Data from sample reduction experiments

• Annex R Results of sampling experiments

• Annex S Results of sample reduction experiments

• Annex T Results for inclusion in CEN Standards

• Annex U References

Part 2 Physical / Mechanical Tests


Part 2.2 Physical / Mechanical Tests

The following final reports were prepared according to the respective tasks of this work package. Following the two basic objectives for Task II.1 there are two separate final reports for “Moisture Content” and “Bulk Density Determination”.

Part 2.2.1 Task II.1a - Moisture Content

Final report prepared by: Peter Daugbjerg Jensen1), Jan Burvall2), Robert Samuelsson2), Contributing co-authors: Hans Hartmann3), Thorsten Böhm3), Michael Temmerman4), Fabienne Rabier4), Jean-Louis Herserner5), Barrie Hudson6), Josef Rathbauer7) Macin Pisarek8) Marzena Hunder8) Michael Lecourt9), Raida Jirjis2), Dulce Helena Boavida10) Julio Antonio Calzoni11) Giuseppe Toscano Dibiaga12)

1) Forest & Landscape, Denmark 2) Swedish University of Agriculture Sciences, Sweden 3) Technologie- und Förderzentrum im Kompetenzzentrum für Nachwachsende Rohstoffe,

Germany 4) Centre wallon de Recherches agronomiques, Belgium 5) Ingenieurbüro Hersener, Schweiz 6) Forestry Contracting Association Ltd, UK 7) Bundesanstalt für Landtechnik, Austria 8) EC Baltic renewable energy centre, Poland 9) AFOCEL, Laboratorie Bois-Process, France 10) INETI, Portugal 11) Comitato Termotecnico Italiano, Italy 12) University of Ancona, Italy

1 Summary

Task II.1 Moisture content and bulk density in the BioNorm project has three focuses:

• Test of reference method

• Test of rapid on-site moisture measuring equipment

• Bulk density test methods

The results from the reference test method and rapid on-site test methods are treated in the present report, whereas the results from the bulk density study are discussed in a separate report.

The oven drying method for moisture determination was tested as reference method. This method is widely used in standards for moisture determination of biofuels. The investigation was performed at the Swedish University of Agricultural Sciences, (SLU) on



a wide scope of biofuels collected by the partners in BioNorm Task II.1. The trials aimed at a qualitative and quantitative determination of the carbon, or dry matter and energy losses during oven drying at different temperatures using various biofuels. The relevant chemical composition properties of the biomass were analysed in order to find basic correlations to any non-water losses during oven drying tests. Statistically significant higher moisture levels were found for determinations in the drying oven compared to xylene distillation and freeze-drying. An analysis of the amount and composition of volatile compounds released during oven drying in air at 105 oC was determined by the GC-MC method. The test revealed that the primarily volatile compounds were α- and β-pinene.

For rapid MC determination 63 wood chips were dried down to eight more or less equidistant moisture steps covering the full moisture range, from fresh comminuted chips down to about 10 % moisture content (w.b.). They were then applied using seven rapid-MC determination devices in a round robin involving 11 European partners. The tested equipment is based on the following methods: thermogravimetrical, optical and electrical (microwave and capacitive). Calibration curves were developed for all tested devices using the drying oven (at 105°C) as reference method [1]. The bulk density of all tested fuels were determined and in case they had a significant influence on the result the bulk density was included by applying multiple regression analysis models.

The most promising method was found to be the optical, which measures the surface moisture by the reflection of infrared radiation (Mesa MM710). Also the thermogravimetric equipment measures the moisture content with a very high precision. However, due to the limitations in sample size the device is only suitable for biofuels with small particle sizes (e.g. sawdust). Among the four tested capacitive devices the Pandis FMG3000 and the Schaller FS2002-H measured the moisture content at the highest accuracy. Common for the capacitive devices is that the precision decreased by increasing moisture levels, thus the highest accuracy is found to be below the fibre saturation point. Implementing calibration functions for specific fuel types e.g. chips from coniferous trees or individual species e.g. Norway spruce results in an increasing accuracy of the devices. The microwave-based device was not found to be applicable for moisture determination of the tested fuel types.

2 Objectives

Moisture content is a fundamental property for a successful utilisation of biofuels throughout the supply chain, from costs of transportation, through storage management to energy values and combustion optimisation at the conversion plant. A number of national standards for determination of moisture content [2, 3] are valid today. However, to perform reliable tests with common acceptance a requirement for common European standards or guidelines for the measurement of moisture content has been identified.

The research presented here aimed at evaluating the performance of relevant and useful laboratory methods for moisture content determination, which can be used for either reference or rapid applications. To cover the broad range of European materials a wide scope of fuels from all over Europe were included. This is also true for the applied technologies and measuring principles.



From the elaborated comprehensive knowledge basis, best practice guidelines are derived and the results from the study are introduced to the technical specifications drafted by CEN TC 335. The technical specifications are regarded as an important tool to support the trade of biofuel in Europe.


Two different sets of biofuels were investigated in Task II.1. For the test of the reference method a selection of biofuels was sent to SLU (Table 1).

Table 1: List of the biofuel material used for tests of the reference method together with its origin country. The approximate size of the material used in the experiments is given.

Fuel Country State of comminution

Spruce wood chips Austria Grinded, <10 mm Pine wood chips Austria Grinded, <10 mm Miscanthus Denmark Grinded, <10 mm Hard wood Germany As received, <30 mm Rape cakes Germany As received, <10 mm Triticale Germany As received, <3 mm Cork Portugal As received, <1 mm Eucalyptus Portugal Grinded, <10 mm Olive stones Portugal As received, <5 mm Pine bark Sweden Grinded, <10 mm Milled peat Sweden As received, <2 mm Sawdust Sweden As received, <5 mm Wood pellets Sweden As received, <30 mm Wood pellets II Sweden As received, <30 mm Spruce bark Sweden Grinded, <10 mm Logging residues Sweden Shredded, <40 mm Spruce needles Sweden As received, <5 mm Pitchy wood chips* Sweden Grinded, <10 mm Salix chips Sweden Grinded, <10 mm Birch bark Sweden Grinded, <10 mm Birch wood chips Sweden Grinded, <10 mm * yellowish-red and strong tar smelling resinous wood of pine

The materials with large particle sizes were grinded using a knife mill, followed by a thorough homogenisation. The fuels were divided into 32 separate sub-samples in such a way that sub-sample 1 and 2, sub-sample 3 and 4, etc. constituted replicates. Sample division was carried out using three different riffle boxes with slits (20 - 75 mm) adapted to the particle size of the samples. The samples were stored in plastic buckets (3-5 litre) or pots (1-1,5 litre) with airtight lids (see BioNorm deliverable II.1.D1 Report on reference test methods for moisture content for further details). Materials with moisture content above 15 % was stored in a refrigerated room at ~ 0 oC, while low moisture fuels were stored at room temperature.



In the test of the rapid on-site equipment a total number of 98 biofuels from a broad range of European wood fuels were used. The fuels covered both deciduous and coniferous wood species comminuted to various size levels, shells, kernels, straw- and wood pellets etc. A short overview of the applied fuels is given in Table 2 (for more detailed description see deliverable II.1.D2 Report on rapid on-site test methods).

Table 2: Applied fuels for the test of on-site equipment

Fuel type No. of samples Coniferous 30 Deciduous 25 Mixed 2 Wood chips

SRC* 6 Bark 6 Sawdust 9 Shavings 1

Wood 6 Pellets Straw 3 Stones/shells 5 Olive cake 1 Grain 1 Energy grass 2 Peat 1 * Willow and Poplar

Samples of minimum 1 m3 volume were collected and divided into equal sub-samples (eight for wood chips, up to five for the remaining fuel types) using the coning and quartering method. For each fuel a moisture range was reflected by stepwise drying of the sub-samples using various drying techniques. The drying stage was e.g. monitored by a reference sample (a separate tray) which was weighed intermediately in order to be able to terminate the drying process at the desired MC-level. This conditioning procedure ensured that one of the samples was at 30 % MC (+/- 3 percentage-points), as this drying stage was consistently applied for the bulk density determination by all partners. The drying procedure was followed by an airtight storage in plastic bags for a minimum period of two weeks prior to measurement. This was to ensure that the fuel moisture was homogeneously distributed within the sample.


4.1 Reference test methods

The oven drying experiments for the determination of the total moisture content were performed according to CEN TC335 [1], the xylene distillation method [5] and a freeze drying method (see Deliverable II.1.D1 for further details). For the freeze-drying experiments a Modulyo 4K instrument (Edwards High Vacuum International, England) was used, where 100 g of sample were dried during 48 hours to constant weight.



4.1.1 Total moisture

Oven drying method

Oven drying at 80 oC vs. oven drying at 105 oC

0

10

20

30

40

50

60

Birch b

ark

Birch c

hips

Cork

Eucaly

ptus

Hard w

ood

Logg

ing re

sidue

s

Milled p

eat

Miscan

tus

Olive s

tones

Pine ba

rk

Pine ch

ips

Pitchy

woo

d

Rape-c

akes

Salix

Sawdu

st

Spruce

bark

Spruce

chips

Spruce

need

les

Tritica

le

Woo

d Pell

ets

Biofuel

Moi

stur

e (%

)

Figure 1: Moisture determination by the oven drying method at 80 oC (white bars) and 105 oC (grey bars),

respectively



Table 4: Results of the significance test for variances and mean values for comparisons of the oven drying in air at 80 oC and 105 oC, respectively. The significance tests were performed at the 95 % confidence level

Fuel Temp [°C]

Mean Value [%]

Standard Deviation

[%]

Relative std. Dev

[%] n F-test Fkrit t-test tkrit

Birch bark 80 45,94 0,427 0,93 3 105 46,49 0,146 0,31 10 8,51 4,26 -3,71 2,20 Birch chips 80 40,63 0,019 0,05 3 105 40,67 0,064 0,16 10 11,55 19,38 -1,09 2,20 Cork 80 3,52 0,038 1,09 3 105 4,08 0,083 2,03 10 4,65 19,38 -10,96 2,20 Eucalyptus 80 45,99 0,071 0,16 3 105 46,21 0,047 0,10 10 2,27 4,26 -6,32 2,20 Hard wood 80 14,37 0,082 0,57 5 105 14,48 0,086 0,59 16 1,11 5,86 -2,51 2,09 Logging residues 80 31,49 0,070 0,22 3 105 31,10 0,498 1,60 10 51,02 19,38 1,32 2,20 Milled peat 80 48,66 0,117 0,24 3 105 49,02 0,079 0,16 10 2,16 4,26 -6,22 2,20 Miscanthus 80 37,32 0,083 0,22 3 105 37,59 0,297 0,79 10 12,71 19,38 -1,52 2,20 Olive stones 80 8,90 0,096 1,08 3 105 9,35 0,021 0,22 10 21,89 4,26 -15,25 2,20 Pine bark 80 50,44 0,579 1,15 3 105 49,04 0,127 0,26 10 20,68 4,26 7,85 2,20 Pine chips 80 48,41 0,106 0,22 3 105 48,24 0,178 0,37 10 2,82 19,38 1,59 2,20 Pitchy wood 80 12,85 0,234 1,82 3 105 13,47 0,055 0,41 10 18,28 4,26 -8,40 2,20 Rape-cakes 80 7,65 0,129 1,69 3 105 9,50 0,034 0,36 10 14,12 4,26 -44,27 2,20 Salix 80 28,00 0,037 0,13 3 105 28,20 0,035 0,12 10 1,14 4,26 -8,43 2,20 Sawdust 80 51,78 0,099 0,19 3 105 51,66 0,063 0,12 10 2,48 4,26 2,46 2,20 Spruce bark 80 40,88 0,122 0,30 3 105 40,57 0,115 0,28 10 1,12 4,26 4,12 2,20 Spruce chips 80 20,04 0,147 0,74 3 105 20,40 0,038 0,19 10 15,16 4,26 -7,61 2,20 Spruce needles 80 49,86 0,145 0,29 5 105 50,27 0,072 0,14 16 3,98 3,06 -8,72 2,09 Triticale 80 11,98 0,295 2,46 3 105 12,90 0,018 0,14 3 257,52 19 -5,35 2,78 Wood Pellets 80 5,75 0,159 2,76 3 105 5,33 0,025 0,47 10 39,62 4,26 8,94 2,20

In Figure 1 and Table 3 the results from the comparison of oven drying in air at 80 oC and 105 oC are summarized. As shown in Table 4 the precision is acceptable for both temperatures with a relative standard deviation of 1 % or less in most cases. The F-tests, however, show that there is a significant difference in precision between the applied



temperature levels for about half of the materials (bold figures in F-test in Table 3) but no tendency was found that indicates one method as the better one.

Table 3 also summarises the results from the significance test of the mean values (t-test). However, this test demands that there is no significant difference between the variances of the compared methods. A significant difference in mean values is obtained for a majority of the biofuel materials (bold figures in t-test column of Table 3). Surprisingly for six materials a higher moisture content was obtained at 80 oC compared to 105 oC. As these differences are on a very low level they are most likely due to the imprecision of the method. Another reason could be that moisture loss or uptakes occurred during storage.

Oven drying at 130 oC vs. oven drying at 105 oC

0

10

20

30

40

50

60

Birch b

ark

Birch c

hips

Cork

Eucaly

ptus

Hard w

ood

Logg

ing re

sidue

s

Milled p

eat

Miscan

tus

Olive s

tones

Pine ba

rk

Pine ch

ips

Pitchy

woo

d

Rape-c

akes

Salix

Sawdu

st

Spruce

bark

Spruce

chips

Spruce

need

les

Tritica

le

Woo

d Pell

ets

Biofuel

Moi

stur

e (%

)

Figure 2: Moisture determination with the oven drying method at 105 oC (grey bars) and 130 oC (black bars), respectively



Table 3: Result of the significance test for variances and mean values for the comparison of the oven drying in air at 130 °C and 105 °C, respectively. The significance tests were performed at the 95 % confidence level

Fuel Temp [°C]

Mean Value [%]

Standard Deviation

[%]

Relative std. Dev

[%] n F-test Fkrit t-test tkrit

Birch bark 105 46,49 0,146 0,31 10 130 46,72 0,196 0,42 3 1,81 4,26 2,21 2,2 Birch chips 105 40,67 0,064 0,16 10 130 40,87 0,067 0,16 3 1,10 4,26 4,62 2,2 Cork 105 4,08 0,083 2,03 10 130 4,67 0,008 0,17 3 108,48 19,38 11,99 2,2 Eucalyptus 105 46,21 0,047 0,10 10 130 46,37 0,095 0,21 3 4,05 4,26 4,24 2,2 Hard wood 105 14,48 0,086 0,59 16 130 14,68 0,028 0,19 2 9,41 245,9 3,23 2,09 Logging residues 105 31,10 0,498 1,60 10 130 31,73 0,217 0,68 3 5,26 19,38 2,09 2,2 Milled peat 105 49,02 0,079 0,16 10 130 50,10 0,112 0,22 3 1,97 4,26 19,00 2,2 Miscanthus 105 37,59 0,297 0,79 10 130 37,79 0,175 0,46 3 2,90 19,38 1,09 2,2 Olive stones 105 9,35 0,021 0,22 10 130 9,99 0,105 1,05 3 26,07 4,26 19,95 2,2 Pine bark 105 49,04 0,127 0,26 10 130 50,98 0,493 0,97 3 15,00 4,26 12,30 2,2 Pine chips 105 48,24 0,178 0,37 10 130 48,59 0,321 0,66 3 3,27 4,26 2,51 2,2 Pitchy wood 105 13,47 0,055 0,41 10 130 14,08 0,021 0,15 3 6,87 19,38 18,70 2,2 Rape-cakes 105 9,50 0,034 0,36 10 130 9,82 0,007 0,07 3 22,87 19,38 15,62 2,2 Salix 105 28,20 0,035 0,12 10 130 28,54 0,062 0,22 3 3,18 4,26 12,54 2,2 Sawdust 105 51,66 0,063 0,12 10 130 51,88 0,044 0,09 3 2,01 19,38 5,61 2,2 Spruce bark 105 40,57 0,115 0,28 10 130 41,61 0,286 0,69 3 6,14 4,26 9,88 2,2 Spruce chips 105 20,40 0,038 0,19 10 130 20,59 0,020 0,10 3 3,60 19,38 8,38 2,2 Spruce needles 105 50,27 0,072 0,14 16 130 51,70 0,599 1,16 2 68,28 4,54 11,46 2,09 Triticale 105 12,90 0,018 0,14 3 130 13,37 0,136 1,02 3 54,60 19 5,98 2,78 Wood Pellets 105 5,33 0,025 0,47 10 130 6,24 0,029 0,47 3 1,35 4,26 53,45 2,2

In Figure 2 and Table 4 the results from the comparison of oven drying in air at 130 oC and 105 oC are summarized. As shown in Table 4 the relative standard deviations are 1 % or less in most cases at 130 oC. As in the former case the F-tests illustrate that there is a significant difference in precision between the methods for seven of the tested materials



(bold figures in F-test) but no tendency was found that indicates one method as the better one.

The results in Table 4 from the significance test of the mean values show that a significant difference of the mean values is obtained for all biofuel materials (bold figures in t-test) except logging residues and miscanthus. In this case, however, the difference in moisture content between the material meets the expectations, namely higher moisture at higher temperatures. Moreover the alteration of sample moisture occurring during storage, as mentioned earlier, was exclusively an increase in moisture content. This would enlarge the significant differences between the drying temperatures. Therefore these results must be interpreted with the same precaution as in the former case.

For the following materials, birch bark, birch chips, eucalyptus, cork, hard wood, olive stones, spruce needles and triticale, however, the results are reliable, since either the analyses were carried out within a limited period of time or, a re-analysis confirmed that the moisture contents had remained constant. The results from these materials clearly indicate almost exclusively a significant difference between the standard method at 105 °C compared to dryings at 80 °C and 130 °C, respectively.



Xylene distillation

Table 5: Result of the significance tests for variances and mean values for the comparison of the xylene distillation method and oven drying in air at 105 oC. The significance tests were performed at the 95 % confidence level

Fuel Method Mean Value [%]

Standard Deviation

[%]

Relative Std. Dev.

[%] N F-test Fcrit t-test Tcrit

Birch bark Xylene 46,20 0,599 1,30 8 105 46,49 0,146 0,31 10 16,79 3,29 -1,48 2,12 Birch chips Xylene 40,42 0,441 1,09 8 105 40,67 0,064 0,16 10 47,30 3,29 -1,84 2,12 Eucalyptus Xylene 46,22 0,058 0,12 6 105 46,21 0,047 0,10 10 1,47 3,48 0,37 2,15 Hard wood Xylene 14,25 0,173 1,21 10 105 14,48 0,086 0,59 16 4,03 2,59 -4,47 2,06 Logging residues Xylene 29,25 0,322 1,10 8 105 31,10 0,498 1,60 10 2,39 3,68 -9,09 2,12 Milled peat Xylene 49,41 0,548 1,11 8 105 49,02 0,079 0,16 10 47,57 3,29 2,25 2,12 Miscanthus Xylene 34,11 0,276 0,81 8 105 37,59 0,297 0,79 10 1,16 3,29 -25,44 2,12 Olive stones Xylene 9,92 0,267 2,69 4 105 9,35 0,021 0,22 10 168,36 3,86 7,06 2,18 Pine bark Xylene 49,47 0,367 0,74 8 105 49,04 0,127 0,26 10 8,32 3,29 3,47 2,12 Pine chips Xylene 48,13 0,383 0,79 8 105 48,24 0,178 0,37 10 4,64 3,29 -0,82 2,12 Pitchy wood Xylene 11,33 0,322 2,85 8 105 13,47 0,055 0,41 10 34,78 3,29 -20,80 2,12 Rape-cakes Xylene 9,54 0,219 2,30 4 105 9,50 0,034 0,36 10 40,46 3,86 0,53 2,18 Salix Xylene 27,84 0,206 0,74 8 105 28,20 0,035 0,12 10 35,27 3,29 -5,50 2,12 Sawdust Xylene 51,79 0,431 0,83 8 105 51,66 0,063 0,12 10 46,93 3,29 0,89 2,12 Spruce bark Xylene 40,67 0,462 1,14 8 105 40,57 0,115 0,28 10 16,01 3,29 0,65 2,12 Spruce chips Xylene 20,51 0,133 0,65 8 105 20,40 0,038 0,19 10 12,41 3,29 2,58 2,12 Spruce needles Xylene 50,37 0,280 0,56 12 105 50,27 0,072 0,14 16 14,94 2,51 1,26 2,06 Triticale Xylene 13,44 0,068 0,50 8 105 12,90 0,018 0,14 3 13,50 19,35 13,36 2,26 Wood Pellets Xylene 5,51 0,068 1,23 8 105 5,33 0,025 0,47 10 7,19 3,29 7,92 2,12

In Table 5 the result from the comparison of the xylene distillation and oven drying in air at 105 oC are shown. The F-tests show that the precision for the distillation method is almost exclusively lower compared to the oven drying method (bold figures in F-test). There are a number of reasons for that. First of all the sample sizes are smaller (30-200 g



compared to >300 g for the oven drying method), resulting in a decrease of representativity during sampling. Secondly, the readability of 0,1 ml at the moisture trap results in relative standard deviation of the order of 0,5 % which is larger than the weighing step (less than 0,05 %) in the oven method.

The t-tests show less significant differences of the standard method (bold figures in t-test) compared to the methods mentioned earlier (80 oC and 130 oC). The difference for hard wood and logging residues may be explained by larger particle sizes (see Table 1), which result in slower and probably incomplete diffusion of the water to the particle surface. The difference for pitchy wood is due to the very high content of volatile matter emitted during the drying process. Further, the discrepancies may be explained by the high temperature (the boiling point for xylene is 137-140 oC) applied during the distillation, especially for materials with long drying times (i.e. triticale and olive stones). The large difference for miscanthus is caused by a change in moisture content during storage between the analyses. Cork was not analysed since its low density and moisture content would have required about 1,5 litre xylene per analysis to obtain enough water for an accurate determination.

Freeze drying The results from the freeze drying experiments (Table 6) clearly show a significant difference in mean values compared to the standard oven drying method in air at 105 oC.

Table 6: Result of the t-test in the comparison between oven drying in air at 105 oC and the freeze drying method using a randomised paired comparison design. The significance test was performed at the 95 % confidence level

Fuel Oven 105oC Freeze drying O-F

51,69 51,24 0,446 52,06 51,53 0,524 Sawdust 51,91 51,29 0,620 39,00 37,72 1,276 38,06 36,88 1,171 Miscanthus 38,28 36,33 1,948 13,58 11,09 2,485 13,51 11,37 2,138 Pitchy wood 13,44 11,47 1,975 5,98 5,39 0,587 5,95 5,34 0,615 Wood pellets 5,97 5,37 0,597

Mean (O-F) 1,198 Stddev 0,747 t-value 5,55 tkrit 2,20

4.1.2 Moisture in the analysis sample

Comparison of oven drying in nitrogen and air (at 105 oC) Table 7 shows no significant difference between oven drying of the analysis sample in nitrogen at 105 oC compared to oven drying in air at 105 oC. This indicates that the drying



atmosphere is of minor importance during the determination of moisture content in the analysis sample. Nitrogen was used to prevent the sample from air oxidation [6].

Table 7: Result of the t-test in the comparison between oven drying of the analysis sample at 105 oC in nitrogen and air using a randomised paired comparison design. The significance test was performed at the 95 % confidence level

Fuel Mean (N2) Mean (105) (N2)–(105)

Birch bark 87,84 88,03 -0,194 Birch chips 95,92 95,86 0,065 Cork 98,15 98,14 0,015 Eucalyptus 89,70 89,78 -0,077 Hard wood 96,87 96,90 -0,026 Logging residues 89,32 89,30 0,024 Milled peat 88,50 88,55 -0,049 Miscanthus 93,79 93,74 0,055 Olive stones 96,16 96,13 0,028 Pine bark 90,94 90,85 0,091 Pine chips 95,84 95,79 0,059 Pitchy wood 95,53 95,31 0,225 Rape-cakes 95,90 95,84 0,059 Salix 94,20 94,17 0,030 Sawdust 79,11 79,00 0,111 Spruce bark 94,39 94,18 0,203 Spruce chips 95,74 95,68 0,058 Triticale 93,00 92,99 0,004 Wood pellets II 96,59 96,52 0,065 Mean 0,0394 Stddev 0,0927 t-value 1,85 tkrit 2,10

Comparison of oven drying of the analysis sample at 105 oC in vacuum and air, respectively Table 8 shows a small significant difference between the drying methods, which can be explained by the lower flow of air through the vacuum drying cabinet. Vacuum was used to prevent the sample from air oxidation [6]



Table 8: Result of the t-test in the comparison between vacuum drying of the analysis sample at 105 oC and oven drying in air at 105 oC using a randomised paired comparison design The significance test was performed at the 95 % confidence level

Fuel Mean (vac) Mean (105) (Vac)-(105)

Birch bark 87,99 88,03 -0,05 Birch chips 95,96 95,86 0,10 Cork 97,86 98,14 -0,28 Eucalyptus 89,85 89,78 0,06 Hard wood 96,70 96,90 -0,20 Logging residues 89,18 89,30 -0,12 Milled peat 88,16 88,55 -0,40 Miscantus 93,78 93,74 0,04 Olive stones 95,94 96,13 -0,19 Pine bark 90,71 90,85 -0,14 Pine chips 95,76 95,79 -0,03 Pitchy wood 95,47 95,31 0,16 Rape-cakes 95,62 95,84 -0,23 Salix 94,09 94,17 -0,08 Sawdust 79,13 79,00 0,13 Spruce bark 93,82 94,18 -0,37 Spruce chips 95,71 95,68 0,03 Triticale 92,80 92,99 -0,20 Wood pellets II 96,39 96,52 -0,13 Mean -0,10 Stddev 0,16 t-value -2,62 tkrit -2,10

4.1.3 Volatile compounds

The “GC-MS”-method was used for a screening examination of the amount of volatile compounds, which are released from a biofuel during oven drying in air at 105 oC. The non-moisture volatile matters (VM) that were released and trapped during the drying process have been determined for various biofuel materials. A schematic overview of the apparatus used in the “GC-MS”-experiments is shown in Figure 3.



Figure 3: A schematic overview of the apparatus used in the “GC-MS”-experiments

Many of the volatile compounds belong to various terpenes, i.e. α-pinene and β-pinene (Table 9). A relative bias of more than 1 % with respect to moisture content is obtained for four of the biofuel materials (cork, pitchy wood and two types of wood pellets), resulting in an overestimation of the moisture. The loss of volatile matter as percentage of dry matter on the other hand, causes a bias that is within the normal uncertainty for the calorific value determination, except for pitchy wood. One should remember, however, that most of the volatile compounds emitted in the drying oven have calorific values far greater than the value of the biofuel itself. α-pinene, for instance, which is the most common volatile in soft wood, has a calorific value of 45 MJ/kg. As a consequence the bias in fuel calorific is expected to be over-proportionally larger than the bias if mass changes.

The table also summarises the major volatile compounds of the fuels as a percentage of the total amount released during the drying step.



Table 9: Non-moisture volatile compounds emitted from selected biofuel materials during oven drying in air at 105 oC

Fuel Dry matter [g] Vmtot [g]

VMtot as[% of

moisture]

VMtot as [% of dry matter]

Major volatile compounds

Birch bark 24,00 0,071 0,35 0,30 2,6-dimethylbicyclohept-2-ene (35 %); 1,3,6,10-cyclotetradecatetraene (22 %)

Birch chips 26,03 0,011 0,06 0,04 Hexanal (12 %); 2-pentylfuran (8 %); 2,6-dimethylbicyclohept-2-ene (5 %)

Cork 34,71 0,018 1,13 0,05 Furfural (50 %); Phenol (22 %)

Eucalyptus 14,57 0,005 0,04 0,03 Hexanal (16 %); α-pinene (12 %); 2-pentylfuran (11 %)

Hard wood 51,75 0,002 0,03 0,004 Piperazine (58 %); 2-pentylfuran (42 %)

Logging residues 35,07 0,027 0,17 0,08 α-pinene (25 %); 3-carene (11 %); 1,4-methanoazulene (14 %)

Milled peat 15,15 0,031 0,27 0,20 2-cyclohexenecarboxanilide (21 %); Isoparvifuran (9 %)

Miscanthus 28,23 0,057 0,34 0,20 Naphtopyran derivative (40 %); 1,4-methanonaphtalene (20 %)

Olive stones 70,98 0,044 0,59 0,06 4-hydroxobutanoic acid (14 %); Pyrazole (10 %); Nonanal (10 %)

Pine bark 16,71 0,025 0,22 0,15 α-pinene (32 %); β-pinene (8 %); Camphor (5 %)

Pine chips 17,73 0,011 0,08 0,06 α-pinene (45 %); β-pinene (12 %); Methanoazulene (8 %)

Pitchy wood 6,85 0,119 11,33 1,74 Bicyclohept-3-ene derivative (24 %); β-pinene (17 %); α-pinene (11 %)

Rape cakes 81,54 0,034 0,51 0,04 2-pentylfuran (34 %); Hexanal (18 %); 1-pentanol (10 %)

Salix 24,42 0,006 0,06 0,02 Hexanal (53 %); 2-pentylfuran (30 %)

Sawdust 16,37 0,009 0,05 0,05 α-pinene (45 %); β-pinene (10 %); Bicyclohept-3-ene derivative (7 %)

Spruce bark 23,78 0,067 0,51 0,28 α-pinene (30 %); β-pinene (10 %); Naphtalene derivative (5 %)

Spruce chips 35,68 0,016 0,18 0,04 α-pinene (48 %); β-pinene (13 %); 3-carene (4 %)

Triticale 35,89 0,003 0,06 0,01 Hexanal (75 %); 2-pentylfuran (25 %)

Wood pellets 120,22 0,113 1,50 0,09 Hexanoic acid (7 %); Hexanal (6 %); Nonanal (4 %)

Wood pellets II 91,12 0,095 1,47 0,10 α-pinene (8 %); Hexanal (6 %); 3-carene (3 %)

4.2 Work plan for rapid on-site test methods

The rapid moisture test research was carried out applying the equipment as compiled in Table 10. Two sets of rapid-MC equipment were procured. 97 fuel samples were processed as given in Chapter 3. A description of the devices can be found in [7]. There were 63 wood samples among this sample set, which dried down to eight more or less equidistant moisture steps covering the full moisture range, from fresh comminuted chips down to about 10 % moisture content (w.b.). This was done in a round robin involving 11 European partners. The partners were divided into two groups’ each having one set of the rapid-MC sent around.



Table 10: Tested methods and instruments (work plan)

Rapid-MC equipment

Method Fabricate Measuring details

MC-range**

Sample volume in test

Oven drying* Various drying ovens 105°C/24hours 0-100 3 replications of

500 gram Thermo-gravimetric Halogen drying Mettler Toledo:

HB43 105°C 0-100 3 replications of max. 20 gram

Pandis: FMG 3000 0-55 3 replications of

60 litres ACO: MMS-0-1-2-0 0-80 15 replications

of ≈ 1 litre Farmcomp: Wile25 13-85 10 replications

of ≈ 10 litres Capacitive

Schaller: FS 2002

Radio frequency

0-40 optimised for 10-25

3 replications of 1700 gram

Electric (based on dielectrical properties)

Microwave hf sensor: Moist 100 2,45 GHz 0-80 20 replications

of ≈ 1 litre

Optical Infrared reflection Mesa: MM710 NIR spectrum 10-40*** 3 replications ≈

2 litre * reference method ** according to manufacturer *** for the tested set-up

A common set of guidelines for sample pre-treatment and measuring procedures were elaborated to secure equal test conditions for all partners. The rapid moisture determination followed the sequence outlined in Figure 4. After an initial homogenisation the sample was divided by into four sub-samples coning and quartering. In the following the numbers in brackets refer to Figure 4. A small sample was taken for the Mettler (1). Three of the sub-samples was unified in a plastic bag where the measurement with the Moist (2). This was performed as 20 replicate measurements on the surface of the mid-height of the filling volume in the bag. Subsequently 10 measurements were taken with the Wile (3) by pressing the probe into the top of the open bag and afterwards 3 measurements with the Pandis (4) was performed. Finally 3 samples for the reference measurements was selected and bulk density was measured on samples with 30 % ±3 MC. The fourth sub-sample was divided into three each used for one measurement with the Schaller (6), five measurements with the ACO (7) and five measurements with the Mesa (8).



Figure 4: Moisture determination with rapid testers and sampling of reference material followed a standard

sequence in order to obtain as uniform trial conditions as possible

4.3 Results of rapid moisture testing

Among the seven tested rapid-MC devices four were found to be particularly applicable for measuring moisture contents in the tested solid biofuels. These were the thermogravimetric Mettler-Toledo HB43 (Mettler), the capacitive devices Pandis FMG 3000 (Pandis) and the Schaller FS 2002-H (Schaller) and finally the optical MESA MM710 (MESA). The first three are on-site types, whereas the MESA is an on-line type. Although the devices are suitable for moisture determination they have some limitations depending e.g. on the sample size, moisture range and heterogeneity of samples which will be addressed in the following.

4.3.1 Fuel type

Key figures for the tested rapid-MC devices are listed in Table 11. To facilitate comparisons among the methods all statistical figures are related to three tests based on the following data set:

(1) coniferous, deciduous and SRC chips from all laboratories

(2) coniferous, deciduous and SRC chips from one laboratory

(3) same laboratory as in 2, but only coniferous chips



The most variable data set is number 1 and moving over 2 to3 it is evident that the data set gets increasingly well defined (i.e. the variance is reduced). This is reflected in the R2 values and standard error shown in Table 11.

All tests were performed on the moisture range recommended by the supplier.

Table 11: Key figures from the tested rapid-MC devices

Dataset 1 Dataset 2 Dataset 3

Method Device MC-range

[%] (w.b.)*Sample in

test R2 Stand. error R2 Stand.

error R2 Stand. error

Thermo-gravimetric

Mettler Toledo HB43

0-100 max. 20 g 0,96 7,96 0,99 3,02 0,997 0,45

Optical Mesa MM710 0-100 ≈ 2 litres 0,86 10,17 0,96 3,82 0,96 2,78

Pandis FMG3000 0-55 60 litres 0,83a 30,71 0,83a 38,08 0,98 5,07

Schaller FS2002-H 0-40 1700 g 0,82a 16,94 0,81a 23,84 0,94 6,14

ACO 0-80 ≈ 1 litre 0,76a 50,16 0,80a 8,64 0,67 71.50

RF

Wile 25 13-85 ≈ 10 litres 0,72a 59,75 0,79a 51,30 0,90 12,99

Di-electric

MW Moist100 0-80 ≈ 1 litre 0,61a 80,20 0,70a 74,53 0,84 22,96 * recommended by supplier a bulk density regarded in regression analysis RF Radio frequency MW Microwave

From Table 11 it is evident that the accuracy of the correlations increases by reducing the data set both in terms of laboratories and fuel types. The methods and devices are listed according to the overall performance measured as R2 values and standard error. It is seen that the most accurate equipment is, in descending order: MettlerToledo HB43, Mesa MM710, Pandis FMG3000 and Schaller FS2002-H. Among the remaining devices: ACO, Wile25 and Moist 100. The ACO is able to reach R2 values comparable to the best devices. However, the variation of the ACO results was relatively high among the laboratories. This can be seen by the large differences in R² between data treatment 1 and 2 and between data treatment 2 and 3. Apparently the trial set-up, the operator and the fuel type largely influence the results of this device. During the measurement the handheld sensor head was pressed towards the surface of the sample and it appears that the result is sensitive towards the inclination angle of the handheld. For practical applications it is therefore recommended to install the measuring head to a fixed platform where the inclination towards the sample material cannot be changed. In the following the results from the MettlerToledo HB43, the Mesa MM710 and the Pandis FMG3000 are further analysed and discussed.

4.3.2 Moisture level

Key statistical figures for the regression analyses of different moisture ranges are found in Table 12. All data are based on measurements with wood chips. The three columns present data on the default moisture range suggested by the supplier (column 1), the full available



moisture range (column 2) and a reduced moisture range with 25 % MC as upper value (column 3).

An expansion of the MC range from the level suggested by the supplier to the complete available MC range is only relevant for the Pandis, ACO and Schaller (column 1 to 2). In the case of the Pandis the R2 is unaffected whereas the R2 of the Schaller is reduced. In both cases, however, the standardised residuals of the models systematically show larger in-homogeneity. An expansion of the MC range should thus be avoided. A limitation of the MC range below the fibre saturation point (column 1 and 3) could be relevant for the Wile and ACO, however neither of the devices improved in R2 values by reducing the MC range.

Table 12: Influence from the moisture level on the performance of the tested equipment

Column 1 - MC range supplier 2 - Full MC range 3 - MC range

0 - 25 Method Default

MC-range [%]b R2c

Default R2d

0-100 MC R2e

0-25 MC Capacitive ACOa 0-80 0.603 0.761 0,516 Pandisa 0-55 0.832 0.835 Wilea 13-85 0.721 0.721 0,622 Schallera 0-40 0.815 0.685 Microwave Moista 0-80 0.628 0.628 0.450 Thermogravimetric Mettler 0-100 0.958 0.958 ** Optical Mesa 0-40 0.873 * **

a the bulk density of the dry matter is included in the calibration b wet based moisture content c data from fueltype 1 (coniferous chips), 2 (deciduous chips), 3 (short rotation coppice) and 16 (mixed chips)

d data from fueltype 1 (coniferous chips), 2 (deciduous chips), 3 (short rotation coppice) and 16 (mixed chips) tested at the full available moisture range

e data from fueltype 1 (coniferous chips), 2 (deciduous chips), 3 (short rotation coppice) and 16 (mixed chips) tested at below 30% MC (estimated fibre saturation point)

* the Mesa test setup is fitted specific for the 10-40 % moisture range ** test not relevant

4.3.3 Thermogravimetric method

The MettlerToledo HB43 is basically a very small drying oven. It has a sample capacity of approximately 15-20 gram (depending on the moisture content and particle density of the fuel) and dries the sample to constant weight by means of a halogen-drying unit. Using the thermogravimetric principle means that the device is able to measure moisture from 0 to 100 %. The regression is performed for the relationship between the reading on the first axis and the moisture content measured with the reference method at 105°C (Figure 5). The bulk density of the biofuel does not affect thermogravimetric methods. The upper and lower lines on Figure 5 represent the prediction intervals (p = 0,95) of the model i.e.



the upper and lower limits of the prediction made with one measurement with the device using the present model.

w = 1.32 + 0.97A

R2 = 0.96

s = 3.02

Reading (A)

Moi

stur

e co

nten

t (w

)

10

20

30

40

50

60

70

0

0 10 20 30 40 50 60 70

Figure 5: One-dimensional regression line for MettlerToledo HB43 with prediction intervals (upper and

lower line), coefficient of variation (R2) and standard error (s). Included data from chips (coniferous, deciduous and SRC), one laboratory and recommended moisture range (0-100%)

The Mettler measured the moisture content with a very high accuracy even when several fuel types were included in the regression analysis (Figure 5). However, the small sample size is a limitation of the device. On one hand the device potentially measures with a high precision on each individual sample, on the other hand the small sample size lowers the representativity of the sample. This means that the device is mostly suitable for moisture determinations on samples with a small nominal top size, e.g. sawdust.

4.3.4 Capacitive method

The capacitive Pandis, as the other capacitive devices, for details see [7], has an optimum measuring range from 0 % moisture up to the fibre saturation point for the tested material. Opposite to the thermogravimetric method the bulk density of the material influences the result. This means that calibrations on specific fuel types increase the power of the model. In Figure 6 regression on the relationship between the dimensionless reading of the Pandis and the moisture content measured with the reference method at 105 °C is shown. The regression analysis is based on data obtained from one fuel type at a single laboratory, which means that the influence of bulk density could be neglected here.



w = 3.54 + 0.0093A

R2 = 0.98

s = 5.07

Reading (A) (dimensionless)

Moi

stur

e co

nten

t (w

)

0 2000 4000 6000 8000

10

20

30

40

50

60

70

0

Figure 6: One-dimensional regression line for Pandis FMG3000 with prediction intervals (upper and lower

line), coefficient of variation (R2) and standard error (s). Included data from one fueltype (coniferous chips), one laboratory and the recommended moisture range (0-55%)

In Figure 7 the correlation on the relationship between the dimensionless reading of the Pandis and the moisture content measured with the reference method at 105 °C is given. The regression is based on data obtained from three fuel types (coniferous, deciduous and SRC chips) at a single laboratory. Therefore the influence of bulk density was considered by performing a multi-linear regression analysis (for the graphical presentation in Figure 7 a bulk density of 150 kg/m3 is used). From the figure it is clear that including several fuel types reduces the power of the calibration. This is shown here by the decrease in the R2 value and the increase of the standard error compared to the one-dimensional calibration in Figure 6. The large variation among the two calibrations speaks in favour of making single calibration functions for well-defined groups of fuels.



w = 19.08 - 0.0064p + 0.0054A

R2 = 0.83

s = 38.08

p = bulk density in kg/m3 dry matter

Reading (A) (dimensionless)

Moi

stur

e co

nten

t (w

)

p = 150 kg/m3

0 2000 4000 6000 8000

10

20

30

40

50

60

70

0

Figure 7: Two-dimensional regression line for Pandis FMG3000 with prediction intervals (upper and lower

line), coefficient of variation (R2) and standard error (s). Included data from three fuel types (coniferous, deciduous and SRC chips), one laboratory and the recommended moisture range (0-55%)

4.3.5 Optical method

The MESA MM710, which operates with infrared reflection, was tested within the 10-40% moisture range. Infrared reflection determines the moisture content by scanning the surface of the particles, this means that the method is unaffected by bulk density. Similar to the remaining devices the precision of the model is reduced when more fuel types are included in the calibration function (Fig. 8). Calibrating the device for a single fuel type leads to R2-values as high as 0,99. Thus the device has a great potential for on-line measuring of solid biofuels.



w = -2.96 + 1.41A

R2 = 0.95

s = 3.82

Reading (A)

Moi

stur

e co

nten

t (w

)

10

20

30

40

50

60

70

00 10 20 30 40 50 60 70

Figure 8: One-dimensional regression line for Mesa MM710 with confidence intervals (upper and lower line), coefficient of variation (R2) and standard error (s). Included data from three fueltypes (coniferous, deciduous and SRC chips), one laboratory and the recommended moisture range (10-40%)

Calibrating the device to a well-defined and narrow set of fuels can increase accuracy of the model. This is shown in Figure 9, where a regression of the dimensionless reading of the MESA towards the moisture content measured with the reference method at 105° C is performed. Again, the measurements were conducted within the recommended moisture range, here 10 - 40%. The difference between the regressions for only one and for several fuel types is however not as pronounced as for the capacitive devices, e.g. the Pandis.



w = -2.34 + 1.40A

R2 = 0.985

s = 1.44

Reading (A)

Moi

stur

e co

nten

t (w

)

10

20

30

40

50

60

70

00 10 20 30 40 50 60 70

Figure 9: One-dimensional regression line for Mesa MM710 with prediction intervals (upper and lower

line), coefficient of variation (R2) and standard error (s). Included data from one fueltype (chips coniferous), one laboratory and the recommended moisture range (10-40%).

In the present study only the moisture range from 10 to 40% moisture is considered. Nevertheless by changing the filter of the measuring radiation the method is in principle applicable to the full moisture range from 0 to 100 % (w.b.).

5 Scientific conclusion

5.1 Reference tests

The results from the moisture determination at different temperature levels indicate that there are significant differences between the standard method at 105 oC compared to the temperatures 80 oC and 130 oC, respectively.

The freeze drying method determined significantly lower moisture content values compared to the standard method, whereas the results from the xylene method were deviating only for some of the fuels.

The ‘GS-MS’ method revealed that the mass loss differences found among the different methods were primarily due to α- and β-pinenes.

5.2 Rapid-MC tests

The Mettler-Toledo HB45, Pandis FMG3000, Schaller FS2002-H and the MESA MM710 are suitable to measure moisture content of biofuels. However, the reduced sample size



(Mettler-Toledo) and the need of fuel specific calibrations (Pandis, Schaller and MESA) should be considered when selecting the device. The MESA was only tested in a reduced moisture range, 10 - 40 % MC, but the method is applicable to the full range of 0 to 100% MC.

The Moist100, Wile25 and ACO estimated the moisture content with a higher variation and can not be recommended for MC determinations in solid biofuels when the test procedures of the BioNorm project are applied.

Reducing the scope of fuels increases the power of the calibration functions and for the capacitive devices it reduces the significance of the bulk density influence.

6 Recommendations

At present the observations and knowledge obtained from the study of the reference method are incorporated in the CEN technical specification for determination of moisture content [1, 8, 9].

This means that a temperature of 105 ºC is applied and any mass loss of volatiles is ignored. In the case that the ‘GC-MS’ method is further developed it could be a method for moisture content determination with a very high accuracy.

The tested rapid-MC determination methods are always depending on a calibration by using a reference standard method (except rapid drying). A standardisation of such methods themselves is therefore questionable. However, the procedure of calibration should become standardised in order to reduce the uncertainties and the measuring bias in practise. The developed best practise guidelines from the Bionorm project can serve as a basis for preparing and performing calibrations and they could contribute to a European Technical Specification on how to calibrate a rapid-MC test device to be used for internal quality control and as quality assurance.



7 References

1 prCEN/TS 14774-1 Solid biofuels – Methods for the determination of moisture content – Oven drying method – Part 1: Total moisture – Reference method, 2004.

2 SS 18 71 70 (1997). Biofuels and peat – Determination of total moisture content

3 DIN 51718 (1995). Feste Brennstoffe - Bestimmung des Wassergehaltes und der Analysenfeuchtigkeit. Deutsches Institut für Normung e.V. (eds), Berlin, Beuth Verlag, 3 p.

4 T208 os-78 Moisture in wood, Pulp, Paper and Paperboard by Toluene Distillation. Tappi 1978.

5 Burvall, J. and Samuelsson, R.: Report on reference test methods for moisture content. EU-Project "BioNorm" (NNE5-2001-00158), Deliverable WPII.1, D1.

6 ISO 331-1983 Coal – Determination of moisture in the analysis sample – Direct gravimetric method

7 Daugbjerg Jensen, P., Hartmann, H., Böhm, T., Rabier, F., Temmerman, M. Report on rapid-MC test methods. EU-Prolect "BioNorm" (NNE5-2001-00158), Deliverable WPII.1, D2.

8 prCEN/TS 14774-2 Solid biofuels – Methods for the determination of moisture content – Oven drying method – Part 2: Total moisture - Simplified method, 2004.

9 prCEN/TS 14774-3 Solid biofuels – Methods for the determination of moisture content – Oven drying method – Part 3: Moisture in general analysis sample, 2004.



Part 2.2.2 Task II.1b - Bulk Density Determination

Final report prepared by: Hans Hartmann1), Thorsten Böhm1) Contributing co-authors: Peter Daugbjerg Jensen2), Michaël Temmerman3), Fabienne Rabier3), Raida Jirjis4), Jan Burvall4), Jean-Louis Hersener5), Josef Rathbauer6)

1) Technologie- und Förderzentrum für Nachwachsende Rohstoffe – TFZ, Straubing, Germany 2) Forest & Landscape, Vejle, Denmark 3) Departement Genie Rural – CRA, Gembloux, Belgium 4) Department of Bioenergy, Swedish University of Agricultural Sciences, Uppsala, Sweden 5) Ingenieurbüro HERSENER, Wiesendangen, in cooperation with Agroscope FAT Tänikon,

Switzerland 6) Bundesanstalt für Landtechnik (BLT), Wieselburg, Austria

1 Summary

Several test methods and procedures for bulk density determination were analysed by using several biofuels such as wood chips, bark, wood and straw pellets, sawdust, grain kernels, etc. Wood fuels were tested in 6 to 8 different drying stages. Bulk density was determined using four containers (one cube of 100 litres and three cylinders of 100, 50 and 15 litres). The determination was performed with and without a "shock impact" applied by dropping the containers from a defined height of 150 mm. A total of 8184 bulk density measurements were conducted using 341 biomass samples at 6 European laboratories.

The 50 and 100 litre cylinders produced highly comparable results, while a strongly reduced container volume of only 15 litres leads to an average underestimation of the bulk density by about -1,4 % for wood chips or -3,0 % for bark compared to the baseline method (100 l cylinder). The shape of the container was accountable for around 1,5 % differences in the measured bulk density of wood chips, where the 100-l-cube always produced lower values than the 100-l-cylinder.

Shock impact accounted for bulk density differences ranging between 6 % (wood pellets) and 18 % (chopped miscanthus). For wood chips an 11 % increase was measured, consistently for high and low bulk density fuels. For most biofuels improved repeatabilities were observed using a cylindrical container shape instead of a cube and by applying shock instead of no shock.

Fuel moisture content below the 25 % (w.b.) level (MC, as received) had a significant influence on measured bulk density (dry basis) due to swelling or shrinkage effects. A correction function was derived for wood fuels. It is suggested to apply a correction factor of 0,712 % for each 1 % moisture difference between the measured MC and the calculated MC for reporting.



2 Objectives

The research aimed at a comprehensive evaluation and identification of the best appropriate methods for the determination of bulk density of solid biofuels. The elaborated knowledge basis was used for deriving best practise guidelines. These guidelines and the fundamental findings from the research were also used as a basis for drafting Technical Specifications in the process of biofuel standardisation, which is currently ongoing within the CEN TC 335.


Each of the six participating European laboratories included at least one coniferous and one deciduous wood chips sample (see Table 1). In total 14 coniferous, 16 deciduous and 4 mixed wood samples were chosen, furthermore 3 bark samples, 2 sawdust samples, 2 wood pellet samples, 4 herbaceous pellet samples and 3 other fuels (chopped miscanthus, triticale kernels and peat) were tested. This adds up to a total number of 48 fuels, which were dried to several moisture steps as described in Chapter 4.1.3. For wood fuels the number of tested drying steps per sample was either 6 or 8, for the remaining fuels minimum 3 drying steps were achieved. In total a number of 341 samples were tested, each measured in all four bulk density containers applying both treatments and measuring three replications per sample. A total number of 8184 bulk density measurements were thus made.

Table 1: Tested fuels and relevant properties (to be continued next page)

Sample Id Fuel kind Number of MC-steps

Lowest MC (%)

Highest MC (%)

Bulk densitya

100-l-cylinder(kg/m³)

Low density wood chips (21) with bulk density (d.b.) < 180 kg/m³ LTW-01.1 Spruce, coarse 8 12,15 55,72 155,85 LTW-03.1 Poplar chips, SRC 8 7,19 60,84 136,06 CRA-01.4 Spruce -larch-douglas 8 8,98 54,80 160,04 CRA-01.1 Larch chips 8 10,74 53,03 175,38 CRA-01.2 Spruce chips 8 11,35 65,32 157,07 DFLRI-01.1 Picea abies 8 14,30 52,90 159,72 DFLRI-01.2 Pinus contorta 8 14,30 52,00 173,41 DFLRI-03.1 Willow 8 11,20 55,80 142,96 SLU-03.1 Willow chips, SRC 6 8,58 60,63 157,71 SLU-01.1 Forest residues 6 8,07 56,14 165,27 FAT-01.2 Spruce timber 8 8,54 50,85 158,70 FAT-01.3 Silver fir timber 8 8,50 48,56 169,22 FAT-01.4 Spruce branches 8 13,32 55,18 173,93 BLT-01.2 Spruce slab chips 8 15,10 59,69 152,25 BLT-01.3 Spruce slab chips 8 12,23 62,50 135,02 BLT-01.1 Spruce slab chips 8 16,14 60,50 182,96 BLT-16.1 Mixed chips, whole tree 8 11,70 45,10 177,98 BLT-03.1 Willow, SRC 8 12,50 52,40 140,00



Sample Id Fuel kind Number of MC-steps

Lowest MC (%)

Highest MC (%)

Bulk densitya

100-l-cylinder(kg/m³)

BLT-02.2 Poplar 8 10,38 51,30 180,26 BLT-02.3 Alder 8 4,47 50,24 170,87 BLT-02.4 Wild cherry 8 3,96 43,26 166,31 High density wood chips (13) with bulk density (d.b.) > 180 kg/m³ LTW-16.1 Power-Plant-Mix 8 14,31 54,73 208,79 LTW-02.1 Beech chips 8 6,25 40,69 217,50 CRA-01.5 Douglas fir, coarse 8 13,24 50,69 181,80 CRA-02.2 Oak chips 8 14,35 59,12 187,05 CRA-02.3 Beech chips 8 12,34 48,89 219,26 DFLRI-02.1 Fagus sylvatica 8 12,90 45,40 219,63 FAT-01.1 Silver fir branche 8 11,42 54,68 192,84 FAT-02.1 Maple timber 8 10,75 44,04 202,86 FAT-02.2 Ash branche 8 11,08 34,94 219,18 FAT-02.3 Ash timber 8 8,71 35,06 213,42 FAT-02.4 Beech branche 8 11,40 43,32 249,35 FAT-02.5 Beech timber 8 9,05 42,92 228,98 BLT-02.1 Hardwood slab chips 8 8,96 45,50 185,82 Bark (3) LTW-04.1 Bark-Mix 7 11,91 66,62 142,32 SLU-04.1 Bark 6 19,46 53,51 152,88 BLT-17.1 Bark-Mix 8 8,40 59,90 123,46 Sawdust (2) LTW-06.1 Sawdust 8 6,82 49,45 143,59 DFLRI-06.1 Sawdust 5 16,20 25,60 96,70 Wood pellets (2) LTW-08.1 Wood pellets 4 5,14 13,02 595,48 BLT-08.1 Wood pellets 3 3,25 11,51 612,52 Herbaceous pellets (4) LTW-09.1 Straw pellets 4 6,11 16,31 539,34 SLU-09.1 Energy grass pellets 3 5,10 9,50 635,92 FAT-09.1 Straw pellets 4 11,19 17,23 463,91 BLT-09.1 Hay pellets 3 4,66 16,10 491,08 Other fuels (3) SLU-15.1 Peat 6 25,84 59,97 143,47 LTW-12.1 Triticale kernels 4 9,41 21,53 643,19 LTW-14.1 Chopped miscanthus 8 6,58 63,79 80,23 Total: 48 341 a dry matter based, determination with shock impact




4.1 Tested methods and influences

The research focus was mainly set on three issues, the container size and shape, the effect of shock on the measurement and the influence of varying moisture content of the fuel on the measurement.

4.1.1 Container comparison

Four containers of different shape and size were tested in an international round robin. One cube and three cylinders were used (Table 2). The cube had a nominal volume of 100 litres following the Swedish wood chip standard SS 187179 [7]. It was compared to three cylinders of respectively 100, 50 and 15 litres nominal volume.

Table 2: Description of the tested measuring containers for bulk density determination

Description 15-l-cylinder 50-l-cylinder 100-l-cylinder 100-l-cube Volume (l)a 15,01 49,86 99,68 100,68 Weight (kg) 2,72 5,44 10,37 13,61 Wall area to volume ratio 19,70 13,18 10,40 10,69 Bottom area to volume ratio 3,01 2,04 1,67 2,14

a mean volume measured at each round robin partner by filling the container with water and weighing its maximum weight of water

4.1.2 Shock impact

All measurements were conducted by applying two different treatments: normal filling, levelling and weighing (treatment "without shock impact") followed by an additional weighing after having dropped the container three times from 150 mm height and having it refilled (treatment "with shock impact"). In this way the same sample material was tested in three replications for each treatment. For the containers below 100 l the sample had to be reduced as required, using the cone shaping and quartering method. The material was poured from a shovel into the container until a cone of maximum possible height was formed. The falling height was 100 to 200 mm above the upper rim of the container. Surplus material was removed using a scantling, which was shuffled over the edge in a sawing-like movement, in order to identify all particles that formed an obstacle. When the removal of larger particles had torn bigger holes into the levelled surface the cavities were



refilled and the removal procedure was repeated. The samples were weighed to the nearest 10 g on a platform balance.

After weighing the shock impact was exerted by positioning the container onto a 150 mm high wooden scantling, from where it was then horizontally moved aside and dropped onto a thin wooden plate placed on a concrete floor. This procedure was repeated three times. The container was refilled after the third dropping according to the described procedure and weighed. In this way, each measurement was repeated two more times. Finally, the moisture content of the tested fuel was determined in duplicate at 105°C according to prCEN/TS 14774-2 [5].

4.1.3 Moisture content influences

Several fresh wood chip samples were progressively dried in order to provide identical sub-samples at varying moisture content levels. After the completion of the drying process the samples were packed airtight and stored for at least 14 days before the bulk density measurements. During the drying, which was done either on a floor or by forced ventilation through a thin fuel layer, moisture content was estimated regularly by parallel weighing, in order to interrupt the process at a drying stage as desired. This was to ensure that the created moisture steps were as evenly as possible distributed over the usual fuel moisture range. The moisture content measurements were made as described in Chapter 4.1.2.

4.2 Work plan

Six European laboratories participated in the round robin trials. Two sets of the above described stainless steel containers were produced at TFZ. Each set was sent around within a group of three participating laboratories, which had prepared 5 to 10 test fuels as described in Chapter 4.1.3. Each partner followed a common guideline, which had been elaborated specifically for the trials.

4.3 Data Evaluation

All original bulk density data were initially calculated to dry matter basis (0 % moisture), assuming that the volume remains constant (shrinkage was ignored). These data were used for container and treatment evaluation. Statistical outliers were also eliminated from the data set applying a procedure described in [6]. The data were regarded as outliers when their difference to the mean value (or median) was higher than 4 times the standard deviation.

Further data processing was necessary to draw general conclusions upon the effect of the actual moisture content on the bulk densities (dry basis). In order to make use of the full range of the gained bulk densities from all wood fuels, a relative data basis was created. Therefore each of the bulk densities (at dry basis), which had been determined at different drying stages, was calculated relative to the bulk density (dry basis) at a uniform moisture content of 30 %. However, measurements at precisely 30 % moisture basis were not available. Therefore the required reference bulk density was calculated by linear interpolation between those two data spots (pairs of variants), where the measurements had been made closest to the 30 % moisture basis with one data spot above and one below (Figure 1). By this calculation the data from different fuels at different bulk densities could



be compiled for a broader database. The 30 % moisture level was preferred for this calculation because it is above the fibre saturation point of wood and a disturbance by particle shrinkage can be excluded.

150

155

160

165

170

175

180

185

0 5 10 15 20 25 30 35 40 45 50 55 60

Moisture content, M (wet basis)

Bul

k de

nsity

BD

, dry

basi

s-s

hrin

kage

igno

red

-

Area of fibresaturation

P30 (30 %; BD30): Interpolation for M = 30 % between two adjacent pairs of variates

P2 (M2; BD2)

Measured pair of variates

P1 (M1; BD1)

212

21230 BD

MM)M%(30)BD(BDBD +

−−⋅−

=

%

kg/m³

Figure 1: Interpolation.pptLinear interpolation for calculating the reference bulk density at a fixed basis of

30 % moisture

For identification of any influence on the measurement given by the bulk density of the fuel itself, all wood chips were allocated to either of two groups: high and low density fuels. The separation between the two groups was made at 180 kg/m³ (dry basis), because at this value a clear separation of all samples could easily be made. Chips consisting of spruce, larch, pine, willow, poplar, alder and wild cherry wood were defined as low density fuels, whereas beech, oak, maple and ash chips were classified as high density fuels. Samples from douglas and silver fir belong to both groups depending on their origin.

The repeatability of the measurements was tested by comparisons of the relative repeatability limits r. Following ISO 5725-1 [4] the repeatability limit r is "…the value less than or equal to which the absolute difference between two test results obtained under repeatability conditions may be expected to be with a probability of 95 %". At a significance level of 95 % the repeatability limit r is calculated by multiplying the square root of the repeatability variance sr² with a factor of 2,8 according to equation (1). The repeatability limit r allows to draw conclusions on the data spread under repeatable conditions and is a measure for the data scatter within a laboratory. Larger differences between results from two measurements than the repeatability limit r shall be considered as suspect and might be repeated. The relative repeatability limit is gained by dividing the absolute repeatability limit r by the mean value.



rs,r ⋅= 82 (1)

where

r Repeatability limit sr Repeatability standard deviation

4.4 Results and discussion

4.4.1 Influence of container shape and size

For wood chips the shape of the container was accountable for around 1,5 % differences in the measured bulk density, where the 100-l-cube always produced lower values than the 100-l-cylinder. This is demonstrated in Figure 2, where the 100-l-cylinder was chosen as reference container (zero line), to which all other measurements were calculated as relative deviation. An analysis of variance (ANOVA) confirmed that the differences between the two container shapes are significant at the 5 percent error level. The observed deviation was consistent for all tested fuels, but it was higher for bark (2,0 %) and lowest for high density fuels with a more homogeneous size distribution such as grain kernels (0,9 %) or wood pellets (0,7 %).

The container size, which was analysed in steps of 15, 50 and 100 litres also had an impact, although the results presented in Figure 2 require a differentiated interpretation. The 50-l-cylinder produces highly comparable results compared to the 100-l-reference-cylinder (deviations of only +0,4 to -0,5 %); here size effects were statistically insignificant, as confirmed by an ANOVA. However, the strongly reduced container size of only 15 litre volume leads to an average underestimation of the bulk density by about -1,4 % for wood chips or -3,0 % for bark. Here statistically significant differences were found by an ANOVA for both, wood chips and for fuels with a more homogeneous size distribution such as grain kernels or wood pellets (Figure 2). An explanation for this finding may be seen in the fact that smaller containers with a larger wall area to volume ratio (see Table 2) are associated with higher side effects at the walls. Additionally the lower total weight of the sample can lead to less compaction during the shock impact procedure, which had been repeated three times before the measurement.



480 312 63 21 42 39 24 12480 312 63 21 42 39 24 12

No. of Replica-tions

-7-6-5-4-3-2-10123

Rel

ativ

e de

viat

ion

to 1

00-l-

cylin

der

%

100-l

-cube

50-l-c

ylind

er

15-l-c

ylind

er

100-l

-cube

100-l

-cube

50-l-c

ylind

er

15-l-c

ylinde

r

100-l

-cube

50-l-c

ylind

er

15-l-c

ylinde

r

100-l

-cube

50-l-c

ylind

er

15-l-c

ylind

er

100-l

-cube

50-l-c

ylind

er

15-l-c

ylind

er

100-l

-cube

50-l-c

ylind

er

15-l-c

ylinde

r

100-l

-cube

50-l-c

ylinde

r

15-l-c

ylind

er

wood chips wood chips Bark pelletsHerbaceous

pellets dustLow density High density Saw- Chopped

miscanthusGrain

kernelsWood

-1,4

-0,2

-1,6

0,4

-1,3

-2,0

-0,5

-3,0

-0,7-0,3

-1,6 -1,8

-0,5

-2,6

-1,4

0,2

-2,1-1,6

0,1

-1,5

-0,9-0,3

-1,4

15-l-c

ylinde

r

50-l-c

ylinde

r

Mean value{{

Standard deviation

Standard deviation

-1,4

a b c

Elimination of: a10 outliers b1 outlier c3 outliers Figure 2: Deviation of bulk densities (dry basis), measured in three of the tested containers compared to the

reference container (100-l-cylinder) as given by the zero-line. All trials were conducted using shock impact and the full data scope including drying stages. The boundary value for high or low bulk density grouping was 180 kg/m³.

4.4.2 Influence of shock impact

The effect of shock impacts on the sample accounted for an increase in bulk density, which ranged between 6 % (wood pellets) and 18 % (chopped miscanthus). For wood chips an 11 % increase was measured; it was consistent for high and low bulk density fuels (Figure 3). Obviously the observed compaction was lowest for all fuels with a low tendency towards bridging. Theses findings comply with an earlier research where both, the magnitude and the fuel ranking for these differences were similar [2].



No. of replica-tions

02468

101214161820

24

Rel

ativ

e de

viat

ion

to

non-

shoc

k ap

plic

atio

n

%

Wood pellets

Grain ke

rnels

Herbac

eous p

ellets

High densit

y wood ch

ips

Low densit

y wood ch

ips

Sawdust

Peat

Bark

Chopped m

iscan

thus21 12 42 311 487 39 18 63 24

6,27,1

9,510,9 11,2

12,413,2

16,2

18,3

21 12 42 311 487 39 18 63 24

Minimum

25th Percentil

Median

75th PercentilMaximum

Figure 3: Effect of shock impact compared to non shock application, here given for the 50-l-container, which was dropped three times before refilling, surface levelling and weighing. The boundary value for high or low bulk density grouping was 180 kg/m³.

By shock application an attempt is made to account for effects of vibrations during transportation or for compacting effects during shipment and storage. A further benefit is given by an increase of the measurement reproducibility. This is indicated by a consistent decrease of the mean relative repeatability limits r for almost all tested fuels.

4.4.3 Effects on repeatability

For most biofuels improved repeatabilities were observed using a cylindrical container shape instead of a cube of similar volume. This is expressed by mean relative repeatability limits r in Table 3, this observation is valid for both treatments, shock and no shock. A cylindrical container shape should therefore be preferred.

Regarding the influence of the container size the repeatability limits r increase with smaller container sizes. This is due to smaller sample sizes and greater rim effects caused by the larger wall area to volume ratio in the smaller container (e.g. 19,7 in the 15-l-cylinder; 10,4 in the 100-l-cylinder, see also Table 2). Nevertheless, for the 100-l and 50-l-cylinder applying shock impact, the repeatability limits are on a low level anyway (all below 2,6 %). If the coefficients of variation (CV) are calculated, they are all below 1 %. Therefore a container size of 50 l leads to acceptable repeatabilities.

Shock impacts also improved the repeatability of bulk density determinations. This can be read from the absolute differences between shock and no shock application in Table 3. For



almost all fuels the absolute differences have a negative value, which means that lower repeatability limits were obtained by shock applications compared to no shock.



Table 3: Mean relative repeatability limits r for determinations of bulk densities in various measuring containers applying shock impact or no shock

Mean relative repeatability limits r (%) 100-l-cube 100-l-cylinder 50-l-cylinder 15-l-cylinder Fuel Shock impact Shock impact Shock impact Shock impact

yes no Absolute

differencea yes no Absolute



differencea

Number of

samples Low density wood chipsb 2,19 3,49 -1,30 1,45 1,97 -0,53 2,53 3,25 -0,72 3,21 4,05 -0,84 164 High density wood chipsc 2,98 4,64 -1,65 1,96 2,48 -0,52 2,22 3,12 -0,90 2,91 4,22 -1,31 104 Bark 2,29 2,90 -0,61 2,21 2,35 -0,14 2,45 2,79 -0,33 5,67 7,77 -2,10 21 Wood pellets 1,17 1,81 -0,64 0,55 1,05 -0,50 0,85 1,13 -0,28 1,24 1,65 -0,41 7 Herbaceous pellets 2,10 2,21 -0,11 1,24 2,05 -0,80 1,67 2,06 -0,39 3,13 3,46 -0,32 14 Sawdust 2,13 2,25 -0,12 1,70 1,48 0,22 1,97 1,96 0,01 3,03 3,35 -0,32 13 Chopped miscanthus 1,33 3,48 -2,15 1,87 2,47 -0,61 1,38 2,48 -1,10 2,65 5,02 -2,37 8 Grain kernels 0,87 0,82 0,05 0,58 0,35 0,22 0,47 0,61 -0,14 0,81 0,85 -0,03 4 Peat 1,59 2,14 -0,55 1,67 2,29 -0,63 2,46 2,29 0,18 3,11 3,27 -0,16 6

a Absolute difference between shock and no shock application b Low density wood chips with bulk density (d.b.) < 180 kg/m³ c High density wood chips with bulk density (d.b.) > 180 kg/m³



4.4.4 Influence of moisture content

From each test series with different drying stages the bulk density data were plotted as a function of the actual moisture content (MC) during the measurement (see example in Figure 4). In contrast to the wet matter based function, which is also plotted in Figure 4, an obvious decrease in bulk density (dry basis) was observed with increasing moisture content. For all tested wood fuels the decrease was largest in the lower moisture range, and it was consistent for all tested containers.

In order to draw general conclusions upon this effect all measured bulk density data were based relative to an interpolated reference bulk density at a fixed moisture level of 30 % (see Figure 4). The results of this evaluation, which were calculated relative to this basis (the zero line), are given in Figure 5.

For the lower MC range a clear correlation between moisture and the measured bulk density (BD) is given. This observation can be explained by the fact that any shrinkage of wood particles is known to take place below the fibre saturation point. This point is usually between 18 and 26 % moisture (wet basis, see [8]). Beyond this boundary no further correlation was found.

195

200

205

210

215

220

225

230

0 5 10 15 20 25 30 35 40 50Moisture content (wet basis)

Bulk

den

sity

(dry

bas

is)

X

240

250

260

270

280

290

300

310

320

330

340

350

360

380

Bulk

den

sity

(wet

bas

is)

xx

Bulk density,(wet basis)

(kg/m³)(kg/m³)

%

15-l-cylinder50-l-cylinder100-l-cylinder100-l-cube

Bulk density(dry basis):

Figure 4: Example of a bulk density gradient (dry basis) for wood chips in various drying stages. Three

replications were measured per data point applying shock impact.

From the correlation function in Figure 5 a need for data adjustment can be derived, if bulk density values shall be related to a common moisture basis. Below 25 % moisture content (w.b.) it is suggested to apply a correction factor of 0,712 % (from Figure 5) for each 1 % moisture difference between the measured MC and the calculated MC for reporting.



-20

-15

-10-5

0

5

10

15

2025

30

35

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70Moisture content (wet basis)

Rel

ativ

e de

viat

ion

to re

fere

nce

BD

%

% y = -0,712 x + 17,0R2 = 0,58

-20

-15

-10-5

0

5

10

15

2025

30

35

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70Moisture content (wet basis)

Rel

ativ

e de

viat

ion

to re

fere

nce

BD

%

% y = -0,712 x + 17,0R2 = 0,58

Figure 5: Correction function for bulk density measurements of wood chips (dry basis) as a function of the

actual moisture content as received. The function was derived from all data "with shock impact" using the results from the three larger containers (100-l-cube, 100-l-cylinder, 50-l-cylinder).


For best practise in bulk density determination of solid biofuels the following conclusions are made.

• A measuring container size of 50 l is acceptable for all tested solid biofuels. For practical reasons a cylindrical container shape should be preferred because of a higher stability and an easier manageability.

• A standardised shock impact on the filled container significantly increases the measured bulk density of fuel chips by about 10 to 12 percent, while there was only found a minor improvement for the mean relative repeatability limits r.

• The fuel's moisture content during the measurement is of high importance, and it has to be recorded. The comparability of bulk density data is only given if any inconsistency in moisture content between the samples is accounted for by the use of a correction factor, as given in Chapter 4.4.

• Moisture content effects are largely restricted to the MC range up to 25 %. Beyond this point possible effects can be neglected.

6 Recommendations

The guidelines for the conduction of the here presented test series with 6 European partners have already served as the major basis for the standardisation of the bulk density determination method as elaborated by CEN TC 335 "Solid Biofuels", Working Group 4 (Physical and mechanical test methods). In this draft standard the conclusions, which were compiled in Chapter 5, have also largely been considered. For example this applies for the chosen container size, the container shape, the shock impact and for the calculation procedure based on the here given results. This direct consideration of BioNorm



recommendations had been possible due to the direct involvement of several project partners in the above CEN Working Group, which has always closely been observing the progress in BioNorm in order to make best and instantaneous use of the research findings.

7 References

1 Böhm, T. Hartmann, H. (2004): Guidelines for bulk density determination. Deliverable D4, Part 2, BioNorm-Project, Task II.1

2 CEN/TC 335 WG4: Solid Biofuels – Method for the determination of bulk density. 3rd draft, 18.09.2003

3 Hartmann, H.; Böhm, T.; Bock, M.: Measuring Bulk Density of Solid Biofuels. In: Proceedings of the 12th European Conference and Exhibition on Biomass for Energy, Industry and Climate Protection, Amsterdam, 17-21 June 2002, published by ETA-Florence and WIP-Munich, pp. 211-214

4 ISO 5725-1: Accuracy (trueness and precision) of measurement methods and results – Part 1: General principles and definitions. 1994

5 prCEN/TS 14774-2: Solid biofuels – Determination of moisture content – Oven dry method, Part 2: Total moisture – Simplified procedure (Final draft 2003)

6 Sachs, L.: Angewandte Statistik - Anwendung statistischer Methoden. 8th edition, Springer-Verlag, Berlin, Germany, 1997 (Applied statistics – Application of statistical methods, in German Language)

7 SS 187179: Peat – Determination of raw bulk density and calculation of dry raw bulk density (in Swedish). Swedish Standards Institute, Stockholm, Sweden 1990; 3 p.

8 Trendelenburg, R.: Das Holz als Rohstoff - Seine Entstehung, stoffliche Beschaffenheit und chemische Verwertung. J. F. Lehmanns Verlag, München/Berlin, 1939, 435 p.



Part 2.2.3 Task II.2 - Ash Melting Behaviour

Final report prepared by: Klaus Hjuler1), Jenny Larfeld2) Contributing co-authors: Hermann Hofbauer3) et al.4)5)

1) FORCE Technology, Denmark 2) Termiska Processer AB, Sweden 3) Vienna University of Technology - Institute of Chemical Engineering, Austria 4) National Technical University of Athens - Department of Chemical Engineering, Greece 5) University of Oulu – Department of Chemistry, Finland

1 Abstract

This report summarises the technical discussions and results on ash melting behaviour that was obtained in Task 2, WP II. It includes a comparison of the methods and what can be expected when using them. The work performed aims at a comprehensive evaluation and identification of the best appropriate methods for the determination of ash melting behaviour. In this report the performance of suggested new methods is tested and evaluated for fuels known to be problematic with respect to their ash. The tested fuels are wood/bark, straw, olive stones and lucerne.

The determination of ash melting behaviour of biofuels is very important for all thermal conversion processes. The sintering, softening and melting of ashes from biofuels can be very different depending mainly on the composition of the ash, but also on various other parameters. Today it is not possible to forecast the ash melting behaviour by calculations. Therefore it is necessary to establish a suitable standard method for determination of the ash melting behaviour of various biofuels.

Today standards developed for coals and coke are available and used also for biofuels. These relatively identical standards are especially DIN 51 730, ISO 540 and ASTM D 1857. A small test specimen is pressed from laboratory prepared ash and heated at a defined heating rate. The samples are observed and the changes in shape are used to more or less subjective identification of different points of the sintering and melting process. However, these coal standards are in general not suitable for biomass ashes that have completely different compositions and thus melting behaviour.

The partners and their contributions FORCE Technology (former dk-TEKNIK Energy & Environment) runs the MAF method that provides a semi-quantitative measure of the ash melt fraction as function of temperature. The method has potential as being reproducible and repeatable.

The Bioresource Technology Unit at National Technical University of Athens have run tests with TGA/SDTA showing a potential of confirming various phenomena involved in ash melting behaviour although the technique can not be standardised at present.



Department of Chemistry at University of Oulu have performed tests using SEM-EDS, supplying valuable information on the ash compositions and species present in the ash samples. Also tests using the XRD method have shown that the amount of amorphous phase can be evaluated. However the relation between amorphous phase and melt phase need to be clarified before the method can be made useful for standardisation purposes.

Umeå University has performed tests with controlled fluidised bed agglomeration, in which the ash melting behaviour was characterised. It is however concluded that the scale of the test method is too large to become standardised, but it is relevant as a reference method.

The Technical University of Vienna has investigated ash melting behaviour using a computer based improvement of the existing DIN standard in order to eliminate the subjective element of the test. Among the parameters investigated were the ash preparation temperature, size and shape of the test piece, heating rate and atmosphere (oxidizing vs. reducing)

TPS Termiska Processer AB has studied another fluidised bed method, the continuously fed fluidised bed method. TPS also coordinated the work in Task 2, WP II.

2 Introduction

The work aims at a comprehensive evaluation and identification of the best appropriate methods for the determination of ash melting behaviour. The performance of existing laboratory methods as well as suggested new methods is tested and evaluated for fuels known to be problematic with respect to their ash. Advantages and disadvantages of the methods are identified and the scope of the sources of error is determined. Any interactions and correlation between the results from each method are identified. From the elaborated comprehensive knowledge basis, best practice procedures should be derived and standard methods are developed. They are required for the definition of solid biofuel requirements as demanded by CEN TC 335.

3 Methods included

The methods used by the involved partners are listed in Table 1 and described below. Also listed in the table is the sample size needed to operate each method. It is evident that the sample size varies for the methods. MAF, TGA/DTA, SEM-EDS and XRF are all laboratory methods that require small amounts of sample. DIN use a sample that is about ten times bigger. While the mentioned methods require that the ash is prepared before analysis the sintering tests use the original fuel, about 1 to 10 kg. Thus the bed sintering methods differ from the other methods in both scale and sample preparation.



Table 1: Methods included in BioNorm WPII Task 2

Method Sample size in grams Partner DIN 1 TUV MAF 0.1 Force TGA/DTA 0.1 NTUA Continuously fed bed sintering 10000 TPS CFBA 1000 Umeå SEM-EDS 0.1 Oulu XRF 0.1 Oulu

3.1 Improved DIN method

The investigations were carried out according to the DIN 51730:1998 standard. Differences between DIN, ASTM and ISO are documented in [1]. The heating microscope consists of three main parts, a light source, an electrically heated chamber and a video camera (CCD-camera), see Figure 1, Figure 2 and Figure 3. A software (EMI.EXE) was used for the automatic control of the procedure, the data acquisition, and the determination of the different characteristic temperatures. The CCD-camera takes continuously pictures from the test piece during the heat up period. These pictures are transferred to the computer and digitalised. With the aid of the software pictures are analysed at adjustable times. The contours of the test piece are important for the determination of the characteristic temperatures (see below). From these contours the edge angle, base angle, test piece area, shape factor und height as well as broadness are determined and stored. The test piece area is determined as the area of the test piece's shadow image. During a measurement, the change compared to the original value is stored, so calibrating the image scale is unnecessary. After the finishing the measurement the results can manually be improved. All results (pictures, determined values, etc.) are stored and can be printed in tables or time dependent diagrams.

Five characteristic temperatures were determined, sintering temperature (S), deformation temperature (A), sphere temperature (B), hemisphere temperature (H) and flow temperature (D).

The first characteristic point, S, can be recognized by a decreasing thermal expansion and the software computes such points and marks them in a diagram showing the test piece area. These marks are considered as suggestions and thus the determination is semiautomatic since you may choose one of the marks or any other point of the curve as start of sintering. In this work the points was determined according to DIN 51006.

The second characteristic point, A, is determined as the first temperature at which the shape factor has changed by 1.5% and the tracked corner's angle has grown by 10% since starting the measurement.



Point B is the first temperature, at which the upper corners are completely rounded and the test piece’s height equals its width1. The software evaluates this point from the height to width ratio, which should have been between 0,9 and 1 at least one time, but not less than 0,85, the corner angles, which must be equal or greater than 150° and the shape factor, which has to be at least 0,8.

The hemisphere point, C, is determined from the shape factor of the test piece, which should be greater than or equal to 0.985, and the test piece's height which should be less than or equal to half the length of the base line. Eventually the flow point, D, is evaluated either according to DIN 51730 1998-04 / ISO 540 1995-03-15 when the height is one third of the height at hemisphere temperature or according to DIN 51730 1984 when the height is one third of the initial test piece height.

Preparation of test pieces involves sieving of the ash into a sample of minimum 2 grams with particles smaller than 63 µm. The ash is mixed with some water (or other wetting means) and pressed into the shape-holder. According to the desired shape different shape-holders are available. Different shapes of test pieces were used for the investigation, cylinders, cone and cube or cuboid. All pieces had a height of 3 to 4 mm.

Figure 1: Heating microscope with the heating chamber, thermocouple and test piece holder.

1 When determining the sphere temperature, two different cases have to be accounted for: Tall test pieces

(higher than wide) must fulfil the condition ‘height equals width’ at least one time, since their height to width ratio continuously decreases when melting. Test pieces with quadratic shadow image on the other hand must maintain their height to width ratio of about 1 until the corners are rounded, which is practically never the case. The method used to determine the sphere temperature therefore has to tolerate a deviation to the ‘height equals width’ condition to some extent.



Figure 2: Schematic drawing of the parts shown in Figure 1.

Figure 3: Light source, heating chamber and video camera.

Figure 4: Shape-holder and press for the production of test pieces.

3.2 MAF

Since 1995 FORCE Technology (former dk-TEKNIK ENERGY & ENVIRONMENT) has developed the “Melt Area Fraction” (MAF) - method for examination of the melting behaviour of biomass ashes. This method is based on analysis of images that are acquired from a high temperature light microscope during the melting process. The homogenised and hand-milled ash sample is placed randomly on a 7 mm diameter sapphire specimen



glass that is inserted in a gas tight heating stage flushed with nitrogen or air, see Figure 5. The sample is heated at a rate of 100°C/min to 550°C, and, after this, at 10°C/min to the desired temperature. In the range from 550°C to the final temperature, digital images are acquired every 5th second of a fixed, random part of the sample (at approximately 100 times magnification using transmitted light). Each image is analysed in real-time and a ‘solid area’, A, is calculated. The analysis result is reported as (1-A/A550) x 100% as function of temperature, where A550 is the ’solid area’ of the image that was acquired first. The result is a continuous curve indicating the relationship between temperature and “melt area fraction” [2]. It should be mentioned that A is not only reduced as transparent melt is produced from the solid but also as a result of shrinking and sintering. Thus, a change in A is not always related to melt formation.

The heating stage used so far was a Leitz 1350 that enables observations using transmitted light. However, due to a limited maximum working temperature (about 1200°C), a new Linkam stage has recently been purchased. The Linkam TS 1500 can operate at up to 1500°C with variable atmospheres and has a robust design. The method needs to be evaluated concerning; the repeatability and ruggedness depending on sample type, quantity, particle size, magnification, the reproducibility, the method of quantification and the possible quantification of condensing volatile species on the cover glass during the heating stage.

The temperature in °C at which the melt area fraction is 10 % is referred to as t10. At is the calculated area of the “ash silhouette” at the temperature t. Correspondingly, at t50 the area fraction is 50 %. It has been found that t10 and t50 are suitable levels of “melting” from an analytical point of view. Some sintering/melting has to take place in order to be detectable and t10 has been found to be appropriate. The temperature, at which 50 % of the ash material is melted, t50, is normally reached below the maximum operating temperature of the method, which is currently about 1200°C. The maximum temperature is set by “false” light emitting from the heating stage itself due to radiation. Thus t50 is found to be a suitable for many samples and motivated in order to minimize the influence from radiation on the temperature. Comparing t10 versus the temperature difference t50 - t10 may give useful information to discriminate between fuel ashes. While t10 determines the initiation of melting the t50 - t10 describes the propagation of melting as temperature is increased. Thus the ashes can be classified according to their melting behaviour during thermal conversion of biofuels into little, moderate and severe sintering/slagging, respectively.

Glass cover

Specimen disc

Heating wire

Thermocouple

Figure 5: Outline of the microscope heating stage.



3.3 TG/DTG

Thermogravimetric Analysis (TGA) and Simultaneous Differential Thermal Analysis (SDTA) analyses are performed by NTUA in a Mettler TGA/SDTA851 apparatus. The sample is placed in a furnace and subjected to a desired heating programme, typically heating from 25 to 1200°C with 10°C/min. The atmosphere in the furnace can be regulated and nitrogen gas was flushed at 50 ml/minute in the reported tests. The analysis implies continuous measurement of the sample weight and the sample temperature during heating. The quantity of sample is typically between 8 and 15 mg.

The weight measurement reveals any mass changes occurring in the sample and by comparing the sample temperature to the temperature of for instance an inert reference material any heat producing or consuming chemical or physical processes taking place in the sample is detected. Melting is detected as an endothermic process involving no change in mass. For a combustion ash, the melting will result in several peaks overlapping each other, corresponding to the melting of the different chemical species in the ash, at different temperatures and involving different melting enthalpies. Previous work has shown that a melting curve (i.e. the %-melt formed, as function of temperature) can be evaluated from the measured data [3c].

The weight loss and melting detected by the method is illustrated in Figure 6 and Figure 7, testing two samples of pure salt.



Figure 6: TG/DTG (top) and SDTA (bottom) analysis of pure CaCO3 salt.



Figure 7: TG/DTG (top) and SDTA (bottom) analysis of pure KCl salt.

3.4 Bed sintering methods

A batch wise controlled fluidised bed agglomeration method (CFBA) is used by UMU [4]. The bench-scale reactor (5 kW), is constructed from stainless steel, being 2 m high, 100 mm and 200 mm in bed and freeboard diameters, respectively Figure 8. To obtain isothermal conditions in the bed, and to minimize the significant influence of cold walls in such a small-scale unit, the reactor is equipped with electrical wall heating elements, equalizing the wall and bed temperatures. The agglomeration tests were initiated by loading of the bed with a certain ash to bed material ratio, under normal FBC conditions. The excess oxygen concentration was controlled to 6 %dry. A fluidization velocity of four times the minimum fluidization velocity was used, and the bed temperature was maintained at 760°C for all fuels except Lucerne and Straw. To avoid agglomeration



during the ashing procedure, which normally occurs in the temperature range of 650-750°C for Lucerne fuels and 750-850 for straw fuels, bed temperatures of 600°C and 680°C, respectively, were used. At an ash amount corresponding to a theoretical value of about 20 wt-% ash in the bed, the fuel feeding was stopped and the operation was switched to external heating. Previous studies runs have shown that only 1.5 % of ash in the bed is sufficient for agglomeration to occur. The bed was then heated up, at a rate of 3 °C/min, to the point where it agglomerated. To maintain a combustion atmosphere in the bed during the external heating phase, propane was mixed with the primary air in a chamber prior to the air distributor. The onset of bed agglomeration was determined by monitoring differential pressures and temperatures in the bed. The detection of initial bed particle cohesion was facilitated by PCA [5] by considering all bed-related variables simultaneously.

The bed sintering tests performed by TPS differ from the CFBA method since fuel is fed continuously to the fluidised bed and the temperature is kept constant in the bed during the test. The time of operation, with increasing ash content, until the bed agglomerates is determined. The reactor is 4,5 m high with an internal diameter of 100 mm, see Figure 9. The lower part is insulated with ceramics and above 1 m it is electrically heated. Initially the bed material is heated with propane until the desired bed temperature is reached. The choice of bed temperature depends on the fuel and for unknown fuels tests at various temperatures are preformed. Fuel is fed to the bed at about 1 to 2 kg/h and the bed is kept at a constant temperature (with propane if necessary). The bed pressure drop and temperature are recorded together with analysis of the flue gases. The elapsed time until bed agglomeration occurs is evaluated from the temperature data and the bed pressure drop.

Prim.Air

PropaneSec.Air

Pre-heater

Wall heater

F1

F2 F3

Propane burner

Fuel

Pump

CO

CO2

O2

NO

THC

Cyclone

Ventilation

Condenser

T6

T7

T8

T5

T4

T3

P4

T2

P3

T1P2

P1

T/P Signals

Data Acquisition System with On-Line PCA

F4

.

.

.

.

. .

.

..

.

DP

.

View window

.

x

Figure 8: Experimental set-up controlled fluidised bed agglomeration test facility.



Air

Fuel

Height 4.5 m Diameter 100 mm10 kW

Fluidised bed

Figure 9: Experimental reactor for continuous feeding bed sintering test

3.5 SEM-EDS and XRF

The University of Oulu has performed tests using a Jeol JSM-6400 scanning electron microscope combined with a Link ISIS energy dispersive x-ray analyser (SEM/EDS). The analyses reveal the distribution of chemical species in individual particles from a sample. The analysis of several particles (up to thousands of particles) is performed with help of sophisticated image analysis and the extensive amount of information is then presented as quasi-ternary diagrams. From each domain Al, Ca, Cl, Fe, K, Mg, Mn, Na, P, S, Si and Ti contents were determined. Making extensive number of area analysis from the microscope images results in elemental analysis in each particle. Both fuel and ash samples have been mounted with epoxy resin (Struers Epofix), cross-sectioned by grinding with 240, 600 and 1200 mesh sandpaper and coated with a thin carbon layer to eliminate the electrostatic effects.

By using XRD the crystalline phases present in ash and the content of amorphous material can be determined. A typical XRD diagram is shown in Figure 10. Comparing this, to the average composition determined from bulk elemental analysis and the compositional distribution determined by SEM-EDS results in detailed information about the ash samples. The crystalline components can be identified from their characteristic diffraction peaks. An analysis using XRD results in a diagram, see Figure 10, where each crystalline material in the test sample give rise to one or several peaks at specific angle. Melting of crystalline results in the formation of an amorphous phase that is seen as an amorphous halo (increase in the background, see Figure 10 and 11).



1 0 2 0 3 0 4 0 5 0 6 0

2 0

6 0

1 0 0

1 4 0

1 8 0

2 Θ (d e g .)

A b

SiO2 35%

Al2O3 30%

Fe2O3 35%

t

ba A

AR =

Figure 10: The amount of amorphous material corresponds to shadow area below the XRD spectrum [6].

SiO2

0

2000

4000

6000

8000

10000

12000

5 10 15 20 25 30 35 40 45 50 55 60

2-Theta

Inte

nsity

SiO2Glass

0

1000

2000

3000

4000

5000

6000

5 10 15 20 25 30 35 40 45 50 55 60

2-Theta

Inte

nsity

Glass

Figure 11: Wavelength spectra of SiO2 and glass for illustration of highly crystalline (SiO2) and highly

amorphous (glass) materials

In Figure 10 the amorphous phase is showed as the shadow area and the amount of amorphous phase, Ra, is proportioned as the normalised area, divided with the total area below the diagram. Figure 11 illustrates diffraction diagrams of highly crystalline (SiO2) and highly amorphous (silica glass) materials.

Mixing of known crystalline salts and known amounts of amorphous material have been studied in order to calibrate the area measure to the weight fraction of amorphous material and a linear relation was found [6]. Preparation of ash samples often involves grinding in order to achieve homogeneous samples. However, grinding crystalline matter results in smaller particles that are detected as an amorphous phase by the XRD.

3.6 Fuels

Four fuels: olive stones, lucerne, straw and wood/bark, were chosen for the project. Their elementary analysis and composition of ash minerals are surveyed in Figure 12, Table 3 and Table 4. From combustion experience straw and particularly lucerne is known to have very sticky ash causing sintering and slagging in boilers, Table 2. Wood/bark on the other hand is a relatively unproblematic fuel concerning the ash. The behavior of olive stone ash is not well known.



Ash samples were prepared by different partners at 550 °C according to working document CEN/TC335/WG4/N46, and in some cases at 710°C and 815°C according to DIN 51719 and ISO 1171-1997.

0.00 0.25 0.50 0.75 1.00

0.00

0.25

0.50

0.75

1.00 0.0

0.2

0.4

0.6

0.8

1.0

LucerneOlive

StrawWood/bark

Al+S

i Na+K

Mg+Ca Figure 12: Molar relations of average composition of ash forming species in the fuels tested.

Table 2: Fuels used in BioNorm WPII Task 2

Fuel Known characteristics Olive stones ? Lucerne Very sticky ash Straw (BioNorm reference fuel) Sticky ash Wood/Bark (BioNorm reference fuel) Relatively unprobl. ash

Table 3: Fuel analysis of Olive stones, Lucerne, Wood/bark and Straw

Olive stones Lucerne Wood/bark Straw

Moisture content [% of sample] 10.9 8.7 9.4 8.3 C [% of dry sample] 50.2 46.1 50.4 46.0 H [% of dry sample] 6.2 5.8 6.1 5.7 N [% of dry sample] 0.2 3.1 0.2 0.8 O [% of dry sample] 41.2 35.2 42.6 40.9 S [% of dry sample] 0.03 0.08 0.01 0.13 Cl [% of dry sample] 0.02 0.21 < 0.01 0.14 Ash [% of dry sample] 2.2 9.5 0.7 6.5 Higher heating value [MJ/kg dry] 19.59 18.96 20.16 18.14 Lower heating value [MJ/kg dry] 18.24 17.67 18.83 16.90



Table 4: Chemical analysis of ash performed by a Swedish national laboratory

Olive stones Lucerne Wood/bark Straw

Amount of ash (550°C) in dry fuel [%] 2,56 9,5 0.6 6.5 Ash composition SiO2 [%] dry ash 12.65 6.77 5.34 53.38 Al2O3 [%] dry ash 1.79 0.76 0.87 0.62 CaO [%] dry ash 17.19 27.9 20.8 8.46 Fe2O3 [%] dry ash 1.55 0.39 2.17 0.48 K2O [%] dry ash 23.28 25.5 8.24 17.85 MgO [%] dry ash 5.08 3.52 3.13 2.17 MnO2 [%] dry ash 0.06 0.044 3.49 0.07 Na2O [%] dry ash 0.98 0.89 0.22 0.30 P2O5 [%] dry ash 2.13 5.99 1.70 2.26 TiO2 [%] dry ash 0.09 0.036 0.08 0.05

The compositional distributions of wood ash, straw ash, olive-stone ash, and lucerne ash, prepared at 550°C, are shown in Figure 13. It can be seen that olive-stone ash resembles that of wood ash, and the compositional distributions of straw and lucerne ash are quite similar

The differences in the phase compositions of ash from different fuels can be deduced by XRD. The diffraction diagram of wood ash indicates the presence of two crystalline phases: silicon dioxide (SiO2) and calcium carbonate [Ca(CO)3]. No crystalline phases can be identified from straw ash. The crystalline phases in the olive-stone ash are silicon dioxide (SiO2), calcium carbonate [Ca(CO)3], and potassium calcium carbonate [K2Ca(CO3)2]. This is consistent with the SEM-EDS results. The crystalline phases in the lucerne ash are silicon dioxide (SiO2), calcium carbonate [Ca(CO)3], potassium calcium carbonate [K2Ca(CO3)2], and potassium chloride (KCl). The presence of a substantial amount of potassium chloride explains leaning of the distribution maximum of the lucerne ash towards the alkali metal corner in the quasi-ternary diagram.



Figure 13: The compositional distributions of wood, straw, olive-stone, and lucerne ash, prepared at 550°C.

The content of amorphous material in the four biomass ashes is shown in Table 5. Straw ash has the highest content of amorphous material, whereas lucerne and olive-stone ash exhibit the smallest content. Also studied was the relationship between the ash preparation temperature and the content of amorphous material in the ash. For wood the ash preparation temperature has only a small effect on the content of amorphous material, whereas for olive stones the temperature has a more significant influence.

Table 5: Calculated amounts of amorphous material in different biomass ashes

Sample Temperature [°C]

Amorphus [%]

550 67 710 67 Wood 815 74 550 89 710 86 Straw 815 97 550 29 710 52 Olive stones 815 52

Lucerne 550 33

3.7 Synthetic ash

Synthetic ash samples were introduced in the project by grinding so called Orton Cones, used by potters for temperature detection, see Figure 14. Orton Cones are available for kiln



temperatures ranging from 590 to 1330 °C and they are designed for a heating rate of 30 °C/min. Eight samples, A to H, were prepared from Orton Cones (of unknown chemical composition). The samples were then tested using the laboratory methods; DIN, MAF, TGA/DTG and XRD.

Figure 14: Pyrolytic self-supporting Orton Cone of approximately 60 mm height (left). The degree of bending of the cones is used by potters to detect when firings have finalised (right).

4 Results

4.1 Synthetic samples

Table 6 is a survey of the characteristic temperatures found using the improved DIN method for the eight samples. The evaluated temperatures are visualised in Figure 15 from which can be seen that material A has the lowest melting point, followed by sample C, H, E, B, F, G and D. This order of melting behaviour is also found using the MAF method. The melting curves found for the eight samples are shown in Figure 16. The temperature of 10 % melt is evaluated and is shown in Table 7.

Tests with TGA/DTA are shown as weight loss of sample A, C and H in Figure 17, their release of heat in Figure 18. Weight and heat release of sample B, D, E, F and G is shown in Figure 19 and Figure 20. The latter group of samples shows a weight loss during heating up to 820 °C. The first group of samples shows change in mass up to 540 °C and then, at increasing temperature, a constant weight.

Thus all methods managed to show that A, C and H were the low melting samples. However, only the DIN and the MAF method are able to quantify melting temperatures.



Table 6: Characteristic temperatures found using DIN method for the Orton Cones A to H.

A B C D E F G H Sintering temperature °C 405 777 580 1076 770 1006 885 676 Deformation temperature A °C 581 1007 682 1281 916 1134 1239 805 Sphere temperature B °C 649 1084 725 1391 969 1190 1314 844 Hemisphere temperature C °C 786 1219 830 >1430 1107 1313 1373 930 Flow temperature D °C 868 1344 890 >1430 1203 1414 >1430 1088

S A B C D

400

600

800

1000

1200

1400Orton Cone sampleDGFBEHCA

Characteristic form

Tem

per

atu

re [

o C]

Figure 15: Characteristic temperatures determined for DIN method for the eight Orton Cone materials (data

from Table 6).

0

10

20

30

40

50

60

70

80

90

100

550 650 750 850 950 1050 1150 1250

Temperature (°C)

Are

a fra

ctio

n, d

eter

min

ed a

s (1

-A/A

550)

x 1

00%

Cone 022/590°C

Cone 019/695°C

Cone 016/796°C

Cone 010/915°C

Cone 06/1013°C

Cone 03/1104°C

Cone 5/1207°C

Cone 9/1280°CA

HAC

A

D

G

FB

E

Figure 16: Amount of melt versus temperature evaluated during heating at 10 °C/min in nitrogen atmosphere of the eight samples (A to H) using MAF method.



Table 7: Temperatures of 10% melt found using MAF method for the Orton Cones A to H.

A B C D E F G H

t10 °C 623 699 753 848 878 1002 1093 1143

Figure 17: Weight loss and rate of weight loss versus sample temperature in tests with Orton Cone A, C and

H in oxidizing atmosphere.

Figure 18: Release of heat versus sample temperature in tests with Orton Cones A, C and H in oxidizing

atmosphere.



Figure 19: Weight loss and rate of weight loss versus sample temperature in tests with Orton Cone B, D, E, F

and G in oxidizing atmosphere.

Figure 20: Release of heat versus sample temperature in tests with Orton Cones B, D, E, F and G in

oxidizing atmosphere.



StaA

0

1000

2000

3000

4000

5000

6000

5 10 15 20 25 30 35 40 45 50 55 60

2-Theta

Inte

nsity

AStaB

0

1000

2000

3000

4000

5000

6000

5 10 15 20 25 30 35 40 45 50 55 60

2-Theta

Inte

nsity

B

StaC

0

1000

2000

3000

4000

5000

6000

5 10 15 20 25 30 35 40 45 50 55 60

2-Theta

Inte

nsity

CStaD

0100020003000400050006000

5 10 15 20 25 30 35 40 45 50 55 60

2-Theta

Inte

nsity

D

StaE

0

1000

2000

3000

4000

5000

6000

5 10 15 20 25 30 35 40 45 50 55 60

2-Theta

Inte

nsity

E

StaF

0

1000

2000

3000

4000

5000

6000

5 10 15 20 25 30 35 40 45 50 55 60

2-Theta

Inte

nsity

F

StaG

0

1000

2000

3000

4000

5000

6000

5 10 15 20 25 30 35 40 45 50 55 60

2-Theta

Inte

nsity

G

StaH

0

1000

2000

3000

4000

5000

6000

5 10 15 20 25 30 35 40 45 50 55 60

2-Theta

Inte

nsity

H

Figure 21: Wavelength spectra resulting from XRD analysis of the Orton Cones A to H.

XRD spectra of the samples, Figure 21, were compared with XRD spectra of pure silica sand and glass, Figure 11. Sample A and C were seen to be amorphous while sample H coincide with the pure silica sample. Thus, the group A, C and H found by TGA/DTG above is concluded to be pure material according to the XRD spectra. In sample B, D, E, F



and G it is concluded that the main crystalline phase is silica. Other possible phases are Fe2O3 and (MgO)0.77(FeO)0.23.

4.2 Ash samples

Results from the study of the ash samples are briefly surveyed here and in detail in the deliverables produced within the project [1, 7].

4.2.1 DIN

Figure 22 shows the characteristic temperatures determined for the four ash samples considered in the project. Four samples of each ash is shown of which two tested in oxidising atmosphere (red and green bars to the left) and two in reducing atmosphere (light blue and blue bars to the right). All tested samples show an initial sintering between 600 and 800 °C. The characteristic temperature A was below 800 °C for the straw and Lucerne samples and also for olive stone ash tested in reducing atmosphere. For wood all characteristic temperatures was above 1400 °C, except characteristic temperature B that could not be determined. Only ash from straw showed such melting behaviour that all 5 characteristic temperatures could be evaluated. For Lucerne ash temperature B could be determined at reducing atmosphere, but not at oxidising atmosphere, temperature C was different depending on the gas atmosphere and temperature D is high for all samples. As mentioned the characteristic temperature A of olive stone ash varied considerably with the testing atmosphere and characteristic temperature B to D is above 1600 °C if they can be evaluated.

It is also reported, from tests not shown here, that the size and shape of the test piece (cylinder, cube or cone) has no significant influence on the determination of ash melting behaviour. Nor has small amounts of additives (water, methanol or dextrin) to improve the stability of the test piece when necessary or varying heating rate (from 3 K/min to 10 K/min).



a)

S A B C D0

200400600800

1000120014001600

Te

mpe

ratu

re [o C

]

Characteristic temperature [oC]

Sample 1 (ox.) Sample 3 (red.) Sample 2 (ox.) Sample 4 (red.)

b)

S A B C D0

200400600800

1000120014001600

Tem

pera

ture

[o C]



c)

S A B C D0

200400600800

1000120014001600

Tem

pera

ture

[o C]



d)

S A B C D0

200400600800

1000120014001600

Tem

pera

ture

[o C]



Figure 22: Characteristic temperatures determined with DIN method for four samples each of a) Wood/Bark, b) straw, c) Lucerne and d) olive stones. Two samples were tested in oxidising atmosphere and two in reducing atmosphere [1].

4.2.2 MAF

Examples of the resulting MAF curves of the 4 biomass samples are shown in Figure 23 as the area fraction versus the inverse abs. temperature. The curves of wood/bark, olive stone and straw are approximately linear, i.e. the melting behaviour can be approximated by an Arrhenius function, whereas the behaviour of lucerne can be said to be non-Arrhenius. In this case, non-Arrhenius behaviour is caused by a high content of potassium calcium carbonate and potassium chloride, according to the SEM/EDS analysis. These components volatilize and are detected as a condensate at 650-700°C on the top glass of the MAF heating stage. No changes are detected after this initial volatilization, until the remaining silicates start melting. Such behaviour results in the mirror S-shaped curve in Figure 23. t10 and t50 can be read from the MAF curves or calculated using the Arrhenius fit.



1

10

100

0.00065 0.0007 0.00075 0.0008 0.00085 0.0009 0.00095 0.001 0.00105

1/Abs. temperature (K-1)

Are

a Fr

actio

n, d

eter

min

ed a

s [(1

-A/A

550)

x 1

00%

]

Wood/bark

Straw

Olive stone

Lucerne

MAF = 10 %

MAF = 50 %

Figure 23: The melt area fraction versus inverse sample temperature of wood/bark, straw, lucerne and olive

ash prepared at 550°C. This plot type enhances the most relevant region of initial melt formation.

4.2.3 TGA/DTG

Tests with ash samples show that endothermic processes occur sometimes related to weight loss of the sample. The weight loss is referred to decomposition of Ca(OH)2, KCl, CaCO3 and MgCO3 for instance, or, more likely, mixtures of such species in the ash sample. The temperature at which weight loss occur varies for the ash samples analyzed, reflecting the varying compositions of the ash material. However, it is difficult to evaluate the melting behavior of the tested samples in order to rank them.

4.2.4 Bed sintering methods

The results obtained from the fluidized bed agglomeration methods are shown in Table 8. There is some disagreement between the CFBA and the continuously fed bed results that probably has to do with the behaviour of volatile alkali species in the biomasses. Work is continued to improve understanding why the results in some cases differ considerably.



Table 8: Agglomeration temperatures found in tests with bed sintering methods (CFBA and continuously fed bed method)

Fuel CFBA (Quartz sand)

CFBA (GR-Granule) Continuously fed method (GR-Granule)

Bark/Wood 900-930 °C Not tested No agglomeration at 1000 °C

Straw 970 °C 970 °C < 9 minutes at 900 °C < 7 minutes at 950 °C

Olive stones 820 °C 870 °C No agglomeration between 850 and 1000 °C

Lucerne 660 °C 640-660 °C 50 minutes to agglomeration at 650 to 700 °C.

26 minutes to agglomeration at 700 °C. No agglomeration (> 7 hours operation 900 °C).

5 Discussion and comparison

The biomass samples wood/bark, straw, olive stone and lucerne were studied by the methods listed in Table 1. However, only the improved DIN, MAF and CFBA methods produce temperature information that can be compared directly. DIN and MAF are ash testing methods, suitable for standardisation, whereas CFBA is a fuel testing method that may be used as reference for agglomeration/sintering.

Figure 24 compares the improved DIN results (mean values of two replicates) with the CFBA results. Except for straw, it is clearly seen that the deformation, hemisphere and flow temperatures are far higher than the CFBA temperatures. The sintering temperature compares relatively well with the CFBA temperature for wood/bark and lucerne, but for wood/bark ash in air only. The sphere temperature (for both atmospheres) is comparable with the CFBA temperature for straw, but was not detected, or is far too high, for the other samples. No clear picture is obtained about the influence of the atmosphere. Overall the atmosphere has probably no effect, which is in agreement with MAF results, but for some temperatures large differences are seen. In Figure 25 the MAF and CFBA results are compared.



600

700

800

900

1000

1100

1200

1300

1400

1500

1600

1700

Wood/bark Straw Olive stone Lucerne

Tem

pera

ture

(deg

. C)

CFBASinteringDeformationSphereHemisphereFlow

Red Ox

Figure 24: Results of the improved DIN test. The data to the left were obtained using CO/CO2 as atmosphere,

and data to the right, using air. The CFBA results (with quartz sand) are shown as reference.

600

700

800

900

1000

1100

1200

1300

1400

1500

1600

1700

Wood/bark Straw Olive stone Lucerne

Tem

pera

ture

(deg

. C)

CFBAMAF T10MAF T50

Figure 25: Mean values of t10 and t50 obtained by MAF with the CFBA results (with quartz sand) shown as

reference.



6 Conclusions

Various methods for the characterization of biomass ash melting behavior were investigated. The purpose was to identify methods that are suitable for standardization and/or reference. Some of the methods (MAF, TGA/SDTA, SEM-EDS and XRF) are all laboratory methods that require very small amounts of ash sample. Except from MAF, they are well-known techniques that are adopted for this specific use. MAF (Melt Area Fraction) is based on computer analysis of images from a microscope, acquired during heating of the sample. In principle, shape - or phase - changes of individual particles are observed. DIN, a method adopted from coal analysis, needs a sample quantity that is about ten times bigger. In bench scale, two fluidized bed tests were studied – the continuous feed method and the controlled agglomeration (CFBA) method. Both use samples of the original fuel, about 1 to 10 kg. They were designed to simulate the combustion environments in a fluidized bed boiler and could be suitable as reference methods for the sintering/agglomeration behavior of biomass ash.

The essential questions about the tests are: What information is really needed to forecast problems related to ash melting? What analysis cost is acceptable? And then, what type of test(s) is/are appropriate? It is generally accepted that it is very important to identify the temperature of initial melt formation. And also the rate of melt forming as the temperature increases is valuable information. The cost should be as low as possible – this eliminates the fluidized bed methods and methods that produce „hard to evaluate“ results that have to be evaluated by specialists, as TGA/SDTA.

The coal ash melting standards DIN 51730, ISO 540 and the like provide melt information in a relatively simple, low-cost manner. The “Improved DIN” method is an attempt to get around with the subjective element in the standards. However, there certainly are some problems about the identification of the characteristic temperatures for biomass ashes. Another potential source of error (not studied) is variations in the density of the test specimen that is produced by compression.

An alternative candidate for a standard method is the recently developed MAF method. The idea behind it is to avoid the identification of standard shapes of a test piece by observing the behaviour of individual ash particles during heating. It is simple and there is no subjective evaluation. The major causes of analytical error have been identified: one is the melt viscosity and the other is particle agglomeration (cold state). High viscosity melts tend to produce “shadows” that are identified as solid matter. Particle agglomeration makes it difficult to dispense the particles in an even layer. Solutions to these problems are available, but the sources of error have not been completely eliminated so far.

It has to be recognized that ash melting is a complex physical/chemical process, where material properties (e.g. density, particle size distribution, melt viscosity, heterogeneity) play a significant role. Therefore, it is unrealistic to expect perfect repeatability from any ash melting analysis.



7 References

1 BioNorm, WPII Task 2, Deliverable 2.

2 Hjuler K., “Prediction of fouling and sintering in thermal conversion of biomass – determination of ash fingerprints”, 1st World Conference and Exhibition on Biomass for Energy and Industry, 5-9 June 2000.

3 Hansen L. A., Frandsen F. J., Dam-Johansen K. and Sørensen H. S., “Quantification of fusion in ashes from solid fuel combustion”, Thermochimica Acta, Vol. 326, p 105-117, 1999.

4 Nordin, Öhman et.al., “CFBA”, Energy & Fuels, 12:1, p 90-94

5 Wold, S. Technometrics 1978, 20, 397

6 Tiainen M., Daavitsainen J. and Laitinen R. S., “The role of amorphous material in ash on the agglomeration problems in FB boilers. A powder XRD and SEM-EDS study”, Energy & Fuels, 16, p. 871, 2002.

7 BioNorm, WPII Task 2, Deliverable 1.



Part 2.2.4 Task II.3 - Particle Size Distribution and Dimension

Final report prepared by: Hans Hartmann1), Thorsten Böhm1) Contributing co-authors: Peter Daugbjerg Jensen 2), Michaël Temmerman3), Fabienne Rabier3), Michael Golser4), Paul Herzog4)

1) Technologie- und Förderzentrum für Nachwachsende Rohstoffe – TFZ, Straubing, Germany 2) Forest & Landscape, Vejle, Denmark 3) Departement Genie Rural – CRA, Gembloux, Belgium 4) Holzforschung Austria – HFA, Vienna, Austria

1 Summary

Methods for size classification were analysed in an international round robin using 13 different conventional wood fuels and two specially prepared standard samples, one wood chip and one hog fuel sample. The true size distribution of these special samples (length, width, height) had been determined by hand measurings with a calliper and by weighing each of the about 7000 particles (reference method). The application of four horizontal, one vertical vibrating and three rotary screen equipments were tested using 5 different screen hole diameters. These systems are compared to a commercially available image analysis equipment. This image analysis system had the highest conformity to the reference measurement, whereas for all horizontal and vertical vibrating screening machines the median value of the size distribution (according to particle length) was only between one third to half of the reference median value. This is attributed to the high particle misplacement, which was particularly found in the larger fractions. The median particle length from a rotary screening is between the measurements by an image analysis and the horizontal screening device.

Furthermore, several influencing factors on the results of screening operations were syste-matically analysed. For horizontal screening a critical shaking frequency was identified, while the chosen initial sample volume was found to be less important. A larger effect was observed for the screening duration, where a fixed minimum time requirement was found useful. For rotary screening the influencing factors are mainly the rotation speed and the inclination angle of the rotating drum. Also the feeding rate and the moisture content of the sample play an important role, as reflected by the measured differences in the calculated median particle size. The influences of mechanical wear due to the sieving process are usually low for wood fuels, but they should not be neglected in the standardisation. Generally, the sample pre-conditioning should also be considered carefully, as any pre-drying requirements can lead to an inhomogeneous moisture distribution within the sample, which would then cause an over- or underestimation of the total particle share in individual classes. However, the fixation of a tolerable moisture range for the samples is required if results from different fuels and laboratories shall be compatible.



2 Objectives

The research aimed at a comprehensive evaluation and identification of the best appropriate methods for the determination of the particle size distribution of wood fuels. The elaborated knowledge basis was created to be used for deriving best practise guidelines [10]. These guidelines and the fundamental findings from the research were also used as a basis for drafting Technical Specifications in the process of biofuel standardisation, which is currently ongoing within the CEN TC 335.


Two standard fuel samples and 13 conventional wood chip samples were produced. The main research findings were elaborated by the use of the standard samples. For these standard samples (SF1 and SF2) the size parameters were determined by hand measurement of 6036 particles of a wood chip sample (volume 9,3 l) and 7534 particles of a hog fuel sample (volume 7,2 l). For each measured particle the expansion of all three dimensions (maximum length, width and thickness) and its weight were determined (Figure 1). The dimensions were measured by using a digital calliper gauge having an electronic transmission of the readings to an excel data sheet. This comprehensive information about each single particle permitted the calculation of the "true" size distribution for both standard samples while differentiating between either particle length, particle width or the calculated equivalent diameter. In contrast to the hog fuel sample with longish and thin shaped particles (mean length-width-ratio: 3,6) the standard chip sample consists of more square shaped particles with a mean length-width-ratio of 1,9.

In order to draw conclusions on the influence of the particle's shape both samples were composed artificially following a consistent length distribution. Before sending the standard fuel samples to the round robin laboratories the particles where hand sorted into different size classes (separated by their length) and coloured by stain colour before subsequent drying. Each colour represents one of five size classes following the same size sequence as applied for the screening devices: 3,15 to 8 mm (1), 8 to 16 mm (2), 16 to 45 mm (3), 45 to 63 mm (4) and above 63 mm (5), thus permitting to select easily and quantify all wrong allocations from any sieving trays.

Figure 1: Producing a standard wood chip sample by manual size determination with a digital calliper



Figure 2: Example of particles in the manually prepared and coloured standard wood chip sample (SF1, left hand) and hog fuel sample (SF2, right hand)

Additionally to the standard samples 13 conventional practise wood fuel samples were produced using four different chipper types: disc (7), drum (2), spiral chipper (2) and shredder (2). Debarked round wood with homogeneous stem diameters were comminuted (spruce and beech). Furthermore, a chipped short rotation coppice sample (poplar), three samples from logging residues and a heterogeneous spruce sample with a broad range of particle sizes was procured (see Table 3).

Table 1: Characteristics of conventional wood chip fuels tested in round robin trials

Comminution Bark Species Source of raw material Chipper Sample

code Dimensions provided by fine medium coarse yes no

Spruce Debarked logs Spiral CF1 P63 TFZ x x Spruce Debarked logs Disc CF2 P16 TFZ x x Spruce Debarked logs Drum CF3 P16 TFZ x x Spruce Debarked logs Shredder CF4 P16 TFZ x x Beech Debarked logs Spiral CF5 P45 TFZ x x Beech Debarked logs Disc CF6 P16 TFZ x x Beech Debarked logs Drum CF7 P16 TFZ x x Beech Debarked logs Shredder CF8 P16 TFZ x x Poplar SRC, whole tree Disc CF9 P16 TFZ x x Spruce Logging residues Disc CF10 P45 DFLRI x x

Spruce Logging res. high needle content Disc CF11 P45 DFLRI x x

Beech Wood thinnings Disc CF12 P45 DFLRI x x Spruce Debarked logs Disca) CF13 P45 HFA x x x x a)a sample with a broad range of particle sizes was created deliberately


4.1 Tested equipment

The round robin was conducted using different size classification methods. These methods were represented by four horizontal screens, one vertically vibrating screen (a modification



of a horizontal screen), three rotary screens and an image analysis system. The applied measuring principles are presented in the following. More information on the equipment is given in a specific project report [6]. All round robin partners followed common guidelines, which had jointly been elaborated to exclude any effects due to inconsistent handling.

Horizontal screens Rotating screens (cylindrical rings) Image analysis

Figure 3: Functioning principles of screening and image analysis methods for wood chips classification

4.1.1 Horizontal screening

In all screening processes five screen hole diameters were applied: 3,15 mm, 8 mm, 16 mm, 45 mm and 63 mm. The screens were manufactured according to ISO Standard 3310-2 [5], using round holes. In all trials the largest particle was measured by hand.

The four tested horizontal screen classifiers applied different screen trays with round shaped or rectangular screen surfaces between 1257 and 2600 cm²; they were horizontally shaking either in a one-dimensional to and fro movement or in round, two-dimensional operations at frequencies between 160 and 300 min-1, or in a three dimensional shaking mode (shaking frequency: 3000 min-1). The shaking amplitude varied between 30 and 120 mm for one- and two-dimensional operations and was smallest for the three-dimensional screening device (2,0 mm). During the shut down of the shaking operation the sample was fed onto the upper of five screening trays whereby the tray with the largest hole diameter was situated on top. Finer particles had to pass through this or the subsequent trays or they were collected in the final collecting pan after having passed the last screen with the smallest hole diameter (see Figure 3).

From preliminary tests the duration of the screening operations was set between 11 and 16 minutes. Any further screening would lead to insignificant mass changes of below 0,3 % per minute, determined between any two consecutive sieves. In all trials the sample size was 8 litres.

4.1.2 Vertical vibration screening

In contrast to horizontal screening, the here tested vibrating table (by Stiletto Inc.) operates at a one-dimensional vertical movement at a very high frequency (3000 min-1) and at a small amplitude (1,3 mm). The screens are also horizontally orientated. Sample handling and screening procedure follow the same rules as applied for horizontal screenings. The



screens were clamped into wooden sieve frames, having a sieve surface of 2500 cm² (500 × 500 mm).

4.1.3 Rotary screening

Three rotary screen classifiers were tested. They were equipped with 5 subsequent screening cylinders, each of 400 mm height (360 mm effective screening length when excluding the blind part at the stiffened ring connection, see Figure 3). The cylinders form a rotating drum of 500 mm diameter and 2230 mm total length. The inclination of the drum towards the horizontal ground (α) was 3,0 degrees; the drum rotated at a speed of 20 min-1. The sample was added by hand over a period of 8 minutes in order to generate a continuous mass flow (1 litre/min). For more details on this sieving machine see [1].

4.1.4 Image analysis

For the photo-optical analysis a commercially available classifier was used (Haver-CPA 4 RT Band by Haver&Boecker, Germany). It represents a completely redesigned edition of a technique, which had previously been tested and compared to screening methods [1].

The sample was fed into a container, which serves as a feeder hopper discharging the particles to a vibrating feed canal. At the end of the canal the particles are dropped onto a conveyer belt, which singularises the particles by transporting them at approximately 22 times higher speed than on the vibrating feed canal (belt speed: 0,9 m/s). Before the particles are falling off the end of the conveyer belt, they pass a linear light source whose light is continuously registered by a digital CCD horizontal line camera on the opposite side of the light source (see Figure 3). This camera records 4096 pixels over a width of 400 mm, thus the resolution per pixel is 98 µm. The camera itself processes 40 million pixels per second, whereby the matrix-equivalent resolution is consistent with 24 megapixels. At the moment when a particle passes the projection plane of the camera, the incoming light is disturbed at a width, which is proportional to the particle's momentary horizontal expansion. From the retention time within the camera's scope and the recordings for the varying horizontal expansion the size of each particle's shadow is calculated and recorded by a computer. This calculation assumes a constant particle velocity at the measuring plane. Overlapping particles are identified as one particle of respectively larger dimension. Therefore the singularization as well as the horizontal spreading of the particles on the conveyer belt needs to be controlled carefully. The processing duration for each 8-l-sample varied between 5 (coarse chips) and 20 (fine chips) minutes. All trials were conducted using the classification according to the maximum length. The maximum expansion of a particle is calculated from the identified longest distance between two pixels of a particle area.

4.2 Round robin trials (equipment comparison)

In the following a general overview on the major research findings for the equipment evaluation is given. More information is available in the detailed scientific report [9].

The results show that the available measuring devices are largely incompatible. A horizontal screening operation generally tends to overestimate the share of small particles



compared to all other methods. This is indicated in Figure 4 and Figure 5, where the 100-%-line represents the "true value" for either the particle length or the particle width, as determined by stereometric measurement.

For the length distribution, which is the usually used fractionating criteria, the highest conformity to the reference values is given by the measurements with the image analysis system. This applies particularly for the wood chip sample, while there is a 13 percent overestimation of the particle length for hog fuels (Figure 4). That is probably due to a less effective singularization for this fuel, which is obviously more susceptible to overlapping positions on the conveyor, while it is being transported through the measuring plane of the camera.

Reference length value

Imag

e ana

lysis

(Max

. leng

th)0

102030405060708090

100110120

M

easu

red

med

ian

valu

ere

lativ

e to

refe

renc

e va

lue

(%)

Vertic

al vib

ration

Horizo

ntal 1

(1D)

Horizo

ntal 2

(1D)

Horizo

ntal 3

(2D)

Horizo

ntal 4

(3D)

Rotary

1

Rotary

2

Rotary

3

Horizontal screening

5 5 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

102

113

55

37

58

39

53

39

52

39

52

37

90

51

89

50

87

50


Rotary screening

Wood chips Hog fuel

Figure 4: Measured median values for the standard wood chips (SF1) and hog fuel (SF2) sample relative to

the reference median length distribution as determined by stereometric method



Imag

e ana

lysis

(Max

. widt

h)

020406080

100120140160180200

M

easu

red

med

ian

valu

ere

lativ

e to

refe

renc

e va

lue

(%)

Vertica

l vibra

tion

Horizo

ntal 1

(1D)

Horizo

ntal 2

(1D)

Horizo

ntal 3

(2D)

Horizo

ntal 4

(3D)

Rotary

1

Rotary

2

Rotary

3


5 5 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

109104 101

121

105

126

96

126

95

125

95

121

163165 161163 158161


Rotary screening

Reference width value

Wood chips Hog fuel

Figure 5: Measured median values for the standard wood chip (SF1) and hog fuel (SF2) sample relative to the reference median width distribution as determined by stereometric method

For the rotary classifying system, of which three different machines were tested, a relatively high conformity to the reference median particle length was given when wood chips were tested. However, this technique has some drawbacks when using hog fuels (Figure 4). Nevertheless, the results from rotary screening come closer to the reference values than any of the conventional horizontal screening devices. Additionally, the reproducibility of both screening principles is quite high, as indicated by the uniformity of the measurements.

The horizontally operating screening machines produce quite consistent results if, as it was done here, the screening operation was terminated after significant mass changes did not occur between the sieves anymore. However, there is a strong underestimation of the actual particle length. This underestimation is higher for hog fuels than for wood chips. The measured median particle size can be as low as half of the reference value for the wood chips length or it is even only around one third for hog fuel samples (Figure 4). The also tested vertically vibrating screens (horizontally orientated) behave similar to the shaking screens. The underestimation of particle lengths can also be read from Figure 6, where the position of the respective size distribution curve from the screening methods was offset to the left hand side compared to the reference curve ("true length distribution").

Obviously these horizontal systems seem to be more influenced by other characteristics than the particle length. In fact there is a higher compliance to the reference median particle width (Figure 5) or to reference median equivalent diameter [2] (results not shown). This can be explained by the fact that long and thin particles can pass the screen holes vertically with the width being a more decisive dimension for the size allocation.



3,15 638 16 45

0

10

20

30

40

50

60

70

80

90

100

1 10 100Hole or particle diameter

Cum

ulat

ive

parti

cle

shar

e X

True length distribution


Rotary screening

Image analysis (max. length)

Median value

mm

%

Figure 6: Cumulative particle size distribution, examples for different testing methods applied on the standard wood chips sample SF1

This error is reduced by applying a rotary screening equipment, where the passage through each of the screen holes is limited by the retention time of a hole below a particle during each rotation. Consequently, long and thin particles are partly detained from slipping through the aperture. Therefore the rotary screen can be characterised as a measuring system for a deliberate incomplete screening.

The degree to which oversized particles pass the screen holes vertically can be read from Table 2. For larger size classes the misplacement is higher than for smaller ones, and above 45 mm almost all particles are longer than the hole diameter they pass. For image analysis the trend is opposite, but the degree of misplacement is generally rather low. Vibrating and shaking screening operations behave similar in this concern. As expected, the total level of particle misplacement is largely higher for hog fuels (data not given here) than for wood chips.

Table 2: Share of misplaced particles from the wood chip standard sample SF1 (for horizontal and rotary screens the mean values of the participating laboratories are given)

Shares of misplaced materiala (%) Fraction

(length in mm) Share of mass

(%) Image analysis (Max. length)

Vertical vibration

Horizontal screens Rotary screens

> 3,15 – 8 0,4 39,4 2,8 3,5 2,2 > 8 – 16 6,2 11,4 65,5 61,3 24,6

> 16 – 45 77,4 3,7 53,6 56,0 21,9 > 45 – 63 11,4 19,7 100,0 99,6 96,5

> 63 4,7 6,1 100,0 100,0 100,0 Average misplacementb (%) 26,7 61,6 63,2 34,2

a Total of shares above and below cut size referring to weight of sample fraction b Total of shares referring to total sample weight



The conventional round robin fuels (mainly wood chips and hog fuels as differentiated by [4]) behaved similar to the standard fuels (Table 3) when screened. Hog fuel samples, such as shredded spruce or the standard hog fuel sample (SF2), had a high relative deviation of median values to image analysis. Generally, deviations were high when comminution was made with blunt tools. Those fuels would usually be characterised as "hog fuels", even if a "chipper" was used for comminution.

Deviations were also higher for spruce than for beech wood. This reflects the fact that spruce wood is mostly longer fibred and therefore any crushing action after the cutting in the chipper (e.g. in the calibration screen) will easier lead to breakings over the length axis rather than across the fibre line. The length-width-ratio of spruce particles is therefore higher than for hard wood.

For the fine fuels, which would be classified as P16 according to [4], the deviations to the reference image analysis results are higher than for medium or coarse wood chips (P45 or P63). This can be read from the average results in Table 3. This observation reflects the different moments of inertia, which the particles are exposed to before they pass a screen hole. Smaller particles are easier to be brought into a favourable orientation towards the respective screen holes as the required torqueses are lower.



Table 3: Conventional round robin fuels: deviations of median values of size distribution to the image analysis as reference measuring method

Relative deviation of median values to image analysis, max. length mode (reference) Tested fuel

Image analysis, max. length

(reference) (mm) Horizontal screensa

Vertical vibrationd

Rotary screensb

Fine fuels, P16 Spruce, shredderc 29,73 -80,7% -77,4% -66,8% Spruce, drum chipper 27,84 -72,9% -72,5% -60,1% Beech, shredderc 29,30 -66,7% -65,1% -55,8% Spruce, standard hog fuel (SF2)c 33,22 -66,1% -67,1% -56,3% Beech, drum chipper 26,06 -58,8% -59,0% -50,3% Poplar (SRC), disc chipper 20,83 -53,5% -51,3% -42,7% Spruce, disc chipper 18,30 -49,8% -48,6% -37,1% Beech, disc chipper 16,36 -49,6% -51,1% -35,6% Average: - -62,3% -61,5% -50,6% Medium medium, P45 Spruce, logging residues, disc chipper 38,87 -51,7% -49,3% -32,1% Beech thinnings, disc chipper 25,83 -50,1% -49,7% -41,7%

Spruce, logging residues, high needle content, disc chipper 40,06 -47,3% -46,7% -32,0%

Spruce, broad range blend, disc chip. 38,80 -44,8% -43,3% -27,9% Spruce, standard wood chips (SF1) 30,28 -40,7% -45,5% -14,0% Beech, spiral chipper 44,27 -35,9% -32,8% -29,5% Average: - -45,1% -44,6% -29,5% Coarse fuels, P63 Spruce, spiral chipper 67,44 -46,1% -44,5% -33,6% Overall average: -54,3% -53,6% -41,0% a Mean values for four horizontal screening devices b Mean values for three rotary screening devices c Hog fuel sample d using flat screens in horizontal orientation

4.3 Influencing factors

In the following a general overview on the major research findings for the influencing factors is given. More information are available in the detailed scientific reports [7, 8].

4.3.1 Influence of shaking frequency (horizontal screening)

Shaking frequency influences were tested using a debarked air-dry beech chip sample of two litres, which was screened in three replications in a horizontal screening device (Retsch AS 400 control). The screening duration was set to 15 minutes applying the screen set as mentioned in Chapter 4.1. For minimising abrasions during the tests the sample had been pre-screened. The shaking frequency was varied in steps of 10 rpm between 150 rpm (2,5 Hz) and 320 rpm (5,33 Hz). Details of the test procedure are given in [7].

The results show, that mean and median values decrease with increasing frequencies until the shaking movement is strong enough (approx. at 200 rpm for the applied screening device) to ensure a constant and repeatable material flow. This can be read from the results in Figure 7. Apparently the probability of a particle passing a hole is greater when a certain



momentum is imposed on the particle. Additionally, there seems to exist a critical frequency (at 190 rpm, see Figure 7, right hand side) with an extremely poor reproducibility. Obviously at this critical point the shaking movement is still too slow to induce a constant amount of particles to fall through the respective holes. The critical shaking frequency may vary for different devices and fuels; it is likely, that such a critical operation point can occur with other horizontal screening devices, too. For newly applied equipment an initial investigation of the machine in combination with a typical fuel sample is therefore advisable.

10

12

14

16

18

20

22

24

26

28

140 160 180 200 220 240 260 280 300 320 340Shaking frequency (rpm)

Mea

n pa

rtic

le s

ize

/ M

edia

n va

lue

(mm

)

Mean particle size

Median value

0

2

4

6

8

10

12

14

140 160 180 200 220 240 260 280 300 320 340Shaking frequency (rpm)

Coe

ffici

ent o

f var

iatio

n (%

)

Mean particle size

Median value

Figure 7: Size parameters of fine beech chips for various shaking frequencies (amplitude 30 mm) in a

horizontal screening device, each frequency was tested with three replications (except 190 rpm: six replications)

4.3.2 Influence of sample volume (horizontal screening)

In horizontal screening procedures the sample material is fed in a single bulk onto the upper screening tray before starting the shaking operation. In order to test the impact of variable sample volumes (machine loadings), seven different sample volumes between one and four litres were tested in 0,5-l-steps by horizontal screening (sieve area: 1257 cm²) using a shaking frequency of 210 rpm and a screening duration of 15 minutes. Four litres of a debarked air-dry beech chip sample were applied. The details of the test procedure are given in [7].

No influence significant of the sample volume on the screening result was found within a range of filling heights between 8 and 32 mm. This is demonstrated in Figure 8, where mean particle sizes, median values and their respective coefficients of variation for various filling heights are compiled. Even a thick sample layer is not connected with an increase in mean particle size parameters, although the generation of sieve undersizes (mainly smaller particles) was assumed to be prevented by bigger particles blocking the sieve holes. Apparently the momentum, which the shaking movement exerts, is strong enough to induce smaller particles to fall through their respective sieve holes. Also the reproducibility is not affected by larger sample volumes, which is illustrated in Figure 8 (right graph).

For practical applications a filling index could be established, identifying a sample volume to sieve surface ratio. According to the here presented results, sample volumes below 3,2 litres per 1000 cm2 sieve surface are still feasible. Although there is no evidence, that



this variable should be carefully restricted in a standard procedure an overloading of the screening device is still thinkable and should be avoided to prevent any blocking sample layers.

10,0

10,5

11,0

11,5

12,0

12,5

13,0

13,5

5 10 15 20 25 30 35Filling height (mm)

Mea

n pa

rticl

e si

ze/

Med

ian

valu

e (m

m)

Mean particle size

Median value

0,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

1,6

1,8

5 10 15 20 25 30 35Filling height (mm)

Coe

ffici

ent o

f var

iatio

n (%

) Mean particle size

Median value

Figure 8: Particle size parameters (left) and coefficients of variation (right) for particle size determinations of beech chips at various filling heights in a horizontal screening device (400 mm sieve diameter, 210 rpm, 15 min screening duration). All measurements were made in three replications.

4.3.3 Influence of the screening duration (horizontal screening)

Too long screening durations can cause abrasion by mechanical wear and increase the share of fines and smaller particles artificially while a too short screening can lead to an incomplete screening process with smaller particles remaining on screen trays with larger holes. The trials to this effect were made in using a debarked beech chip sample. Details of the test procedure are given in [7].

The results in Figure 9 show that longer screening durations lead to lower mean particle sizes and median values. Median value differences between the one-minute-intervals are larger at a short screening durations and they decrease with a further prolongation of the screening process. Within longer screening durations the probability of a particle passing a hole increases while a higher number of shaking movements can bring the particles into a suitable position for separation through a screening hole.



9,2

9,4

9,6

9,8

10,0

10,2

10,4

10,6

10,8

4 5 6 7 8 9 10 11 12 13 14 15 16 17Screening duration (min)

Mea

n pa

rticl

e si

ze/

med

ian

valu

e (m

m)

Mean particle sizeMedian value

Minimum screening duration

Figure 9: Particle size parameters of beech chips (debarked) for various screening durations determined by

a horizontal screening device with Retsch AS 400 control (210 rpm shaking frequency, 3 mm amplitude) applying a sample volume of 2,4 l / 1000 cm2 sieve area

In combination with the here described screening time trials further investigations were made on a procedure to determine a minimum screening duration for each device individually. However, this procedure, which is described in [7], proved to be highly unreliable.

Generally it can be concluded, that the screening duration has a high influence on the results and it shall always be reported. Especially short screening durations lead to an incomplete screening process. A fixed screening duration for all screening devices is preferable for solid biofuels, except for those, which are susceptible to abrasion (e.g. pellets). According to the current state of knowledge a fixed screening duration of 15 minutes could ensure a completed screening process.

4.3.4 Influence of samples moisture content

The moisture content (MC) of the sample is supposed to have an influence on size classification results due to several effects. For the tests on this effect both, rotary and horizontal screening devices were applied. For rotary screening eight debarked wood chip samples each of ten litres volume were produced at the highest naturally available moisture content. The samples were dried down step-wise to equilibrium moisture by floor drying on a plastic sheet before conducting the rotary screening tests, which were done parallel by 3 different partners. The horizontal screenings were made using 3 different wood type samples. All details of the procedure are given in [7].

For the rotary screenings examples of the gained results for the median values and mean particle sizes are given in Figure 10. Generally there was a trend towards higher mean particle sizes and median values when the fuel moisture content increased. This trend can be explained by a swelling of the particles at higher moisture, but this usually takes place only below the fibre saturation point (approx. between 19 and 25 % w.b.). Adherences and sticking effects or an increased roughness of the particle surfaces are therefore also likely to be responsible for the increase, too, as they prevent particles from being separated from



each other. Additionally, dryer samples can become brittle, thus facilitating abrasion by screening, which would lead to an increased share of fines.

Medium beech chips (CRA)

Slope = 0,031R2 = 0,55

Slope = 0,045R2 = 0,52

22,022,523,023,524,024,525,025,526,026,527,0

0 5 10 15 20 25 30 35 40 45 50Moisture content (%)

Mea

n pa

rticl

e si

ze/

Med

ian

valu

e (m

m)

Mean particle size

Median value

Coarse beech chips (CRA)

Slope = 0,130R2 = 0,48

Slope = 0,147R2 = 0,43

38394041424344454647484950


Mea

n pa

rticl

e si

ze/

Med

ian

valu

e (m

m)

Mean particle size

Median value

Fine spruce chips (TFZ)

Slope = 0,025R2 = 0,78

Slope = 0,011R2 = 0,31

10,5

11,0

11,5

12,0

12,5

13,0

13,5

14,0

14,5


Mea

n pa

rticl

e si

ze/

Med

ian

valu

e (m

m)

Mean particle size

Median value

Coarse spruce chips (TFZ)

Slope = 0,092R2 = 0,57

Slope = 0,163R2 = 0,70

38394041424344454647484950

0 5 10 15 20 25 30 35 40Moisture content (%)

Mea

n pa

rticl

e si

ze/

Med

ian

valu

e (m

m)

Mean particle size

Median value

Figure 10: Particle size and linear regression parameters of wood chip samples at various moisture contents determined by rotary screening (examples of trials from two partners)

However, for rotary screening the influence of the sample moisture on the size classification results are relatively low, compared to horizontal screenings. This can be read from the results given in Figure 11. Apparently the median particle size rises sharply when moisture contents are above 10 to 15 % and it reaches a maximum for the moisture range of around 30 %. Obviously the applied mechanical impact is not strong enough to separate all particles that stick together.



68

10121416182022242628

0 5 10 15 20 25 30 35 40 45 50 55 60Moisture content (%)

Med

ian

valu

e (m

m)

Medium beech chipsFine beech chips

Figure 11: Median values for size distributions of fine and medium beech chips (disc chipper) at different

moisture contents determined in a horizontal screen (Retsch AS 400 control, 15 min screening duration, 210 rpm shaking frequency)

From the results it can be clearly concluded, that in a standard measuring procedure a fixed sample moisture should be demanded allowing only little tolerances for variations. This is particularly true for horizontal screenings. Such a target MC-value should be in the range of an air dried sample and the moisture should be homogeneously distributed within the sample (see also Chapter 4.3.8). Overdried samples shall be avoided. Also re-wettings of the sample are not feasible. A determination of the moisture content parallel to the screening is useful in order to provide a documentation for conformity with these requirements.

4.3.5 Influence of inclination angle and rotation speed (rotary screening)

The inclination angle was believed to have an influence on the screening result as it determines the retention time of the sample within the area of each screen cylinder. Two samples (fine spruce and medium beech chips) were screened at four different inclination angles (2 / 2,5 / 3 / 3,5 degrees) and at three different rotation speeds (12 / 16 / 20 rpm) in four replications for each combination. The raw material of both samples was debarked before chipping; it had a moisture content below 20 %. More details on the test procedure are given in [7].

For both fuels a clear tendency to increasing median values was observed when a higher rotation speeds of the drum is applied. The differences are more distinctive for the coarser sample (see Figure 12, beech chips). In an ANOVA it was found, that the differences between the rotation speeds were statistically significant at the 95 % confidence level. Apparently lower rotation speeds increase the probability of a particle falling through a hole as the retention time above the hole is longer.

A steeper inclination angle leads to a higher median value, at least this is true for the medium beech chip sample (Figure 12, left). Regarding the coefficients of variation (CV) neither inclination angle nor rotation speed seem to affect the repeatability of the results. Even for high rotation speeds combined with steep slopes the CV was below 0,9 %.



Figure 12: Median values of medium beech chips (left) and fine spruce chips (right) determined at different rotation speeds and inclination angles of the sieving drum in a rotary classifier (4 replications per combination)

4.3.6 Influence of feeding rate (rotary screening)

The feeding rate was also supposed to influence the result of a size classification by rotary screening. A high material flow can lead less opportunities for finer particles to pass through a screen hole because they may be blocked by larger particles.

In the here conducted trials an eight-litre-sample of air-dried debarked spruce and a medium debarked beech chip sample were applied in a rotary screen (inclination angle: 2,5 degrees, rotation speed: 16 rpm) at 13 feeding rates between 0,2 and 16,0 l/min at three replications. This was achieved by using a conveyor belt feeder at variable speeds. The details of the test procedure are given in [7].

The results show that higher feeding rates lead to an obvious increase in median values and mean particle sizes (see example in Figure 13). With the rise of mean particle sizes and median values consequently the shares of fines and particles below 8 mm decrease.

9,5

10,0

10,5

11,0

11,5

12,0

12,5

13,0

13,5

14,0

0 1 10 100Feeding rate (l/min)

Mea

n pa

rticl

e si

ze/

Med

ian

valu

e (m

m)

Mean particle size

Median value0,0

0,10,2

0,3

0,40,5

0,6

0,7

0,80,9

1,0

0 1 10 100Feeding rate (l/min)

Coe

ffici

ent o

f var

iatio

n (%

)

Mean particle sizeMedian value

Figure 13: Particle size parameters and coefficients of variation for fine spruce chips (debarked, disc chipper) determined at various feeding rates by a rotary screen (rotation speed: 16 rpm, inclination angle: 2,5 °). Each measuring point represents the mean value of three replications.

Medium beech chips

14

15

16

17

18

19

20

21

22

1,5 2,0 2,5 3,0 3,5 4,0

Inclination angle (°)

Med

ian

valu

e (m

m)

20 rpm16 rpm12 rpm

Fine spruce chips

9,8

10,0

10,2

10,4

10,6

10,8

11,0

1,5 2,0 2,5 3,0 3,5 4,0Inclination angle (°)

Med

ian

valu

e (m

m)

20 rpm16 rpm12 rpm



For the reproducibility (CV, Figure 13) no clear tendency was found, although for finer spruce chips a slightly smaller variation occurs at lower feeding rates. However, even for high material flows the coefficients of variation show an extremely good repeatability; all values are below 1 %. The results underline the importance of a uniform and fixed feeding rate. This parameter should therefore be defined in a standard procedure.

4.3.7 Influence of mechanical wear (rotary screening)

Solid biofuels are natural materials, which often have “un-solid” properties due to their origin or conditioning. Therefore abrasion of particles and comminution processes can be caused by mechanical wear during the screening process. For tests on quantitative effects of such mechanical impacts three air-dry wood chip samples were screened in 10 to 15 replications using a rotary screen. All details of the procedure are given in [7].

In all trials a reduced median particle size value was observed with an increasing number of successive replications (see example in Figure 14). The linear regression lines had a negative slope exclusively, which was usually higher for coarser than for finer fuels.

Slope = -0,015R2 = 0,84

Slope = -0,018R2 = 0,87

10,0

10,5

11,0

11,5

12,0

12,5

13,0

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16Number of replication

Med

ian

valu

e (m

m)

Poplar chips (SRC, disc chipper)

Fine spruce chips (drum chipper)

Slope = -0,159R2 = 0,81

Slope = -0,075R2 = 0,59

20

22

24

26

28

30

32

34

36

0 1 2 3 4 5 6 7 8 9 10 11Number of replication

Med

ian

valu

e (m

m)

Coarse spruce chips (spiral chipper)

Larch bark

Figure 14: Median values of different wood fuels determined in a variable number of successive replications in a rotary screen (12 rpm; 2,2 degrees)

Particles abraded by mechanical wear were mostly found in the fine fractions. Here the increase in the share of particles was highest. From this observation it should be concluded, that the number of replications for a measurement should be kept low. Due to the high reproducibility of the measurements there is also no need for demanding more than two replications per measurement. Nevertheless, abrasion can also appear during the sample preparation, therefore the sample handling shall always be conducted carefully.

4.3.8 Influence of sample drying

The drying time and the drying method can cause inhomogeneous moisture distributions in the different size fractions and can thus lead to over- or underestimations of their total share. Therefore different sample pre-treatments (drying methods and climates) were tested for such effects. An example of sample drying in a dry climate is given in Figure 15. It shows the MC of the particle size fractions at different analysis steps (drying stages). The



reference line (step 1) represents the data of the start samples. Due to a high variation, the values of the fractions >45 mm are here not considered. All details of the procedure are given in [8].

Drying in dry climate (20°C/35% r.h.) Drying by forced ventilation (20°C/65% r.h.)

0

10

20

30

40

50

60

70

0 - 3,

15

>3,15

- 8

>8 - 1

6

>16 -

45

>45 -

63 >63

Fraction (mm)

Moi

stur

e co

nten

t w.b

. (%

) Step 1 Step 2Step 3 Step 4Step 5 Step 6

0

10

20

30

40

50

60

70

0 - 3,

15

>3,15

- 8

>8 - 1

6

>16 -

45

>45 -

63 >63

Fraction (mm)M

oist

ure

cont

ent w

.b. (

%) Step 1 Step 2 Step 3

Step 4 Step 5 Step 6Step 7

Figure 15: Moisture content of size fractions of spruce chips (debarked, disc chipper) for various drying stages (steps) in dry atmosphere (left graph, 20°C/35% r.h.) and by forced ventilation (right graph at 20°C/65% r.h.). Determinations by horizontal screening

The results indicate, that the drying of the smaller particles happens faster than for larger particles. In the course of the drying the moisture differences within the sample are reduced. Regarding the also tested moisture distribution of fresh wood chips which were wrapped in synthetic fabric or paper, no clear trend was observed after 14 days with respect to a reduction of the MC differences. Further results are reported in [8].


Particle size analysis of biofuels is a difficult task, which is associated with high measuring uncertainties. This is basically due to the fact, that the three major measuring principles (horizontal and rotary screening, image analysis) produce largely incompatible results. In horizontal screening a severe underestimation of the particle lengths is given, while some improvements are found for the rotary screening method. The highest conformity to the reference values is given for an image analysis system.

Comparable measurements must therefore consistently be made using only one of the three principles, while for the same principle modifications of the equipment type (e.g. one-, two- or three dimensional shaking operations) are usually acceptable, this is verified by the high reproducibility observed. However, this flexibility is only given, when all relevant measuring variables and the influencing factors (moisture content, frequency, screening duration, feeding rate, rotation speed, inclination angle, mechanical wear, sample pre-treatment) are carefully considered and standardised according to the results and conclusions from this research.



6 Recommendations

The elaborated laboratory guidelines for the conduction of the here presented tests with 4 European partners and the intermediate research findings have already served as the major basis for the standardisation of methods for determining particle size distribution of solid biofuels as elaborated by CEN TC 335 "Solid Biofuels", Working Group 4 ("Mechanical Properties"). In the draft technical specification for horizontal and rotary screening operations the here elaborated requirements have largely been considered. However, the suggested methods in the technical specification still represent a compromise in respect of conformity with the reference values. A better information about true particle dimensions can only be provided by an image analysis system. But this technology is still relatively costly, therefore its major focus of application is presumably given when large sample volumes are processed or a frequent sampling is required. As long as the widespread horizontal screening methods are still common in practice, the image analysis can preferably be applied when conformity to such other measurements is not required. This can for example be given in an internal quality assurance system for large biofuel suppliers or purchasers. Here the additional benefits of this technology, such as combinations with automated sampling processes or the additional measuring features (e.g. the mean particles' sphericity or the mean length-width-ratio) could also be utilised. Nevertheless it would be useful to launch a European standardisation process in order to include the image analysis method to the scope of applicable standard laboratory principles for biofuels. This would be a fist step to overcome the disadvantages of the screening methods.

7 Acknowledgements

The authors would like to thank the Swedish research company TPS (Termiska Processer AB) for the helpful exchange of information on the subject of size classification, particularly by aeroclassification, and for fruitful discussions during their participation in several BioNorm project meetings.



8 References

1 Hartmann, H.; Böhm, T.; Bock, M.: Measuring Size Distribution of Wood Chips. In: Proc. 12th Europ. Conf. on Biomass for Energy, Amsterdam, 17-21 June 2002, publ. by ETA-Florence and WIP-Munich, pp. 215-218

2 Hartmann, H.; Böhm, T.; Daugbjerg Jensen, P. Temmerman, M. Golser, P. Herzog: Methods for Size Classification of Wood Fuels. Proceedings 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection in Rome, May 10-14; 2004 (in press)

3 CEN TC335 WG4: Draft Technical Specification: Solid Biofuels –Methods for the determination of particle size distribution (May 2004).

4 prCEN/TC 14961: Solid Biofuels – Fuel Specifications and Classes. Draft Technical Specification, Final draft, July 2004.

5 ISO 3310-2: Test sieves, Technical requirements and testing; Part 2: test sieves of perforated metal plate (3rd ed. 1990-11-01)

6 Böhm, T.; Hartmann, H., Herzog, P.; Rabier, F., Daugbjerg Jensen, P.: Report on equipment for size classification and determination of particle dimensions. EU-Project "BioNorm" (NNE5-2001-00158), Deliverable WPII.3, D1&D3

7 Böhm, T.; Hartmann, H.: Report on sensitivity of size distribution testing to influencing factors, Part 1: Influence of device adjustments, procedures of determination and sample properties. EU-Project "BioNorm" (NNE5-2001-00158), Deliverable WPII.3, D2 (a).

8 Herzog, P.; Golser, M.: Report on sensitivity of size distribution testing to influencing factors, Part 2: Influence of general moisture content and different pre-treatment impacts. EU-Project "BioNorm" (NNE5-2001-00158), Deliverable WPII.3, D2 (b).

9 Böhm, T.; Hartmann, H.: Report on interactions and fundamental findings for size distribution. EU-Project "BioNorm" (NNE5-2001-00158), Deliverable WPII.3, D4.

10 Böhm, T.; Hartmann, H., Daugbjerg Jensen, P., Temmerman, M., Rabier, F., Golser, M.: Best practice guidelines for size classification. EU-Project "BioNorm" (NNE5-2001-00158), Deliverable WPII.3, D5.



Part 2.2.5 Task II.4 - Durability and Raw Density of Pellets and Briquettes

Final report prepared by: Fabienne Rabier1), Michael Temmerman1) Contributing co-authors: Peter Daugbjerg Jensen2), Hans Hartmann3), Thorsten Böhm3), Josef Rathbauer4), Juan Carrasco5), Miguel Fernadez5)

1) Departement Genie Rural – CRA, Belgium 2) Forest & Landscape, Denmark 3) Technologie- und Förderzentrum für Nachwachsende Rohstoffe – TFZ, Germany 4) Bundesanstalt für Landtechnik – BLT, Austria 5) Centro de investigaciones energéticas, Medioambientales y tecnologicas – CIEMAT, Spain

1 Summary

Several test methods, devices and different procedures for particle density and durability of pellets and briquette determination, were tested and evaluated. Round robin tests were organized involving 5 partners for 15 pellet and 5 briquette types representing as much as possible the actual or potential European market.

For particle density determination, the tested methods were: stereometric method and methods based on liquid displacement, hydrostatic and buoyancy method with different procedures. Concerning the pellets durability, two main methods were tested and compared: the first one based on the American standard (tumbling device) and the second one based on the Austrian standard (pneumatic device). The briquettes durability was evaluated by two different briquette testers using different duration of treatment.

The results were analysed following the ISO 5725.2 standard about the repeatability and the reproducibility.

Regarding the different results for the particle density determination, it is clear that the methods based on liquid displacement give lower repeatability and reproducibility than stereometric methods.

For briquettes, the three methods based on liquid displacement lead to similar results. In the case of pellets, the buoyancy method using non coated samples and wetting agent mixed with water gives the lowest values in repeatability and reproducibility and also needs less replications to fulfil a given accuracy level. Both for pellet and briquette, the variability of all tested methods is influenced by the fuel type itself.

The most repeatable and reproducible method for the estimation of the briquettes durability is to tumble the briquettes for 105 rotations with the briquette tester. Nevertheless, the briquettes durability testing leads to highly variable results. Concerning the pellets durability, the tumbling device gives better results compare to the pneumatic one. It must



be added that there is no accurate relation between the results of pellets durability obtained by both devices.

For briquettes, as well as for pellet, the fuel itself influences the variability of the durability results.

The selected methods for the determination of the particle density and durability were used to compare both parameters for the tested fuels. Nevertheless, no relation was found either for briquettes and pellets.

2 Objectives

Task II.4 research aims at evaluating and identifying in a comprehensive way the best appropriate methods for the determination of particle density and durability of pellets and briquettes. The performance of the existing laboratory methods and equipment have been tested in comparative trials and evaluated under technical and work-efficiency aspects. Different fuels, representative of the European market have been tested. Advantages and disadvantages of the method have been identified and the scope of the sources of error has been studied. Moreover the accuracy, the repeatability and reproducibility of selected methods have been calculated in order to provide objective data for method selection. Additionally, any interactions and correlations between the results from each tested method have been calculated in order to calculate conversion factor if applicable. From the elaborated comprehensive knowledge basis, best practice procedures have been derived. These best practice guidelines are essential for drafting both, the standard methods and the fuel quality standards as demanded by CEN TC 335. Standards are seen as major measure to develop the markets for solid biofuels to achieve the goals of the commission in preserving the climate and the environment in developing rural areas.


The briquettes selected for the round robin test are commercially available. The briquette selection includes two briquette types produced by piston press (B4, B5), two by extruder press (B1, B2) and one is produced by a chamber press (B3). The raw material and the moisture content of the different briquette types are listed in Table 1.



Table 1: Characteristics of the different briquette types selected for testing (MC: Moisture content, StD: standard deviation, wb: wet basis)

Briquette Code Press Cross Section Drawing Raw Material

MC Average (% wb)

MC StD

B1 Extruder Hexagonal Mixed wood 7,74 0,76

B2 Extruder Square

Hardwood 5,93 0,48

B3 Chamber Rectangular

Mixed wood 8,39 0,34

B4 Piston Circular

Mixed wood 7,68 0,35

B5 Piston Circular

Softwood 8,59 0,39

The wood pellets selected for the round robin test are commercially available. The selection includes wood pellets of a diameter of 6 mm (8 different types) and 8 mm (3 different types). Additionally, pellets made of agricultural residues have been included in the selection. The raw material and the moisture content of the different pellet types are listed in Table 2.

Table 2: Characteristics of the different pellet types selected for testing (MC: Moisture content, StD: standard deviation, wb: wet basis)

Pellets code Diameter Raw material MC Average (% wb) MC StD

P1 6 mm Mixed wood 6,0 0,8 P2 6 mm Softwood 6,7 0,4 P3 6 mm Hardwood 8,1 0,5 P4 6 mm Hardwood 8,3 1,7 P5 8 mm Mixed wood 6,8 0,8 P6 8 mm Mixed wood 7,4 0,6 P7 8 mm Mixed wood 7,0 0,7 P8 6 mm Mixed wood 8,0 0,4 P9 6 mm Mixed wood 8,7 0,9 P10 9 mm Straw 8,3 0,6 P11 6 mm Miscanthus 8,1 0,6 P12 6 mm Softwood 8,8 0,5 P13 6 mm Softwood 8,9 0,5 P14 8 mm Hay 9,8 3,1 P15 9 mm Straw 8,9 0,7




The hereunder-listed deliverables have led to the results presented in 4.1 for particle density and 4.2 for durability.

• Report on equipment for durability determination (pellets and briquettes) August 2003 II.4.D1

• Report on equipment for particle density determination (pellets and briquettes) November 2003 II.4.D2

• Report on interaction and fundamental findings in testing durability and particle density of pellets and briquettes December 2004 II.4.D3

• Best practise guidelines for testing durability and particle density of pellets and briquettes December 2004 II.4.D4

• Sensitivity analysis January 2002

• Report on extra round robin test (ASAE/ÖNORM) October 2003

4.1 Particle density

In some national standards [1-3], particle density of pellets and briquettes is considered as a quality indicator of densified fuel. Moreover, this property has been reported as having an influence on bulk density and on combustion properties of pellets and briquettes (heat conductivity and rate of degasification) [4, 5].

Nevertheless, the measuring of this parameter (ratio between the mass and the volume of a sample) faces some difficulties. If the mass measurement is not a problem, the determination volume of briquettes or pellets sample is not obvious.

Indeed, the now and then irregular shape and the roughness of their surface, increase the variability of the measures resulting from using given methods (e.g. stereometrics). Moreover, the hygroscopic characters of these materials (which seem to be linked to the raw material of what the fuel is made [6]) induce some incertitude for the estimation of the volume with liquid displacement methods. Finally, the hygroscopic characters of pellets and briquettes act not only on the mass of the sample but it also influences its volume [5].

This research aims to compare different methods for the estimation of the particle density of pellets and briquettes. It focuses on stereometric methods and methods using liquid displacement. These methods have been tested for 5 briquette and 15 pellet types, by 5 different laboratories.

4.1.1 Tested Methods

For the following methods, the particle density is calculated as the ratio of the mass on the sample volume. The increase of the volume, due to the increase of the moisture content, is neglected and the particle density is expressed for moisture content “as received”.



Method description

Liquid displacement methods Liquid displacement methods are based on the Archimedes principle. The volume of a sample is estimated through the mass of the volume that is displaced while the sample is submerged. Those methods have, nevertheless, two major disadvantages: as the samples are hygroscopic, a part of the liquid is absorbed by pellets and briquettes, and air bubbles may appear on the surface of the sample while submerged. To avoid the impact of these inconvenients, samples may be coated with paraffin and wetting agent can be added in the liquid.

The here presented tests on liquid displacement methods were performed in pure water, on coated (paraffin) or non-coated samples and in a mix water – wetting agent for non-coated samples. The paraffin used for the tests is characterised by a melting point of 52 – 54 °C. The used wetting agent is t-Octylphenoxypolyethoxyethanol; Polyethylene glycol tert-octylphenyl ether (CAS number: 9002-93-1; trade mark Triton X-100), at a concentration in water of 1,5 g/l.

The calculation of the volume is taking into account the liquid density (as a temperature function) and wetting agent concentration.

The hydrostatic method. For this method, a beaker glass filled with liquid (water or mix of water and wetting agent) is placed on the weighing plate of the balance, a cage intend to receive the sample is submerged in the liquid. The cage is hanged up on a tripod, which has no physical contact with the beaker glass – balance system. The mass difference between the empty system and the system with a sample placed in the cage allows the volume determination of the sample.

The buoyancy method. The difference between buoyancy and hydrostatic method is that, in this case, the beaker glass filled with water is not weighed. Here the volume is determined by the difference between the mass of a sample in the air and the mass of the same sample submerged in the liquid.

This kind of weighing implies the use of a balance allowing to take the measurement by the bottom side or the use of a particle density buoyancy kit. This kit allows to bridge the beaker glass in order to avoid contact with the weighing plate.

Paraffin coating. The paraffin coatings have been performed by immerging the sample in a bath of liquid paraffin (temperature: 100°C). Samples have been weighed before and after the paraffin coating, in order to evaluate the mass of applied paraffin. By knowing the paraffin density (900 g/l), it is possible to calculate the applied volume of paraffin on the sample.

Stereometric methods Stereometric methods consist of measuring the dimensions (diameter, length, width, height…) of a sample with the help of length measuring instruments (calliper rule, palmer). Afterwards, the volume of the sample is approached by calculating the volume of the nearest regular geometrical shape (cylinder, parallelepiped, cube…).



Stereometric methods applied to pellets. A single pellet sample is weighed to the nearest 0,0001g and its shape is considered as a cylinder. The length of the pellets is measured twice at the nearest 0,1 mm (measures are taken at the right angle of each other). The diameter is measured twice (measures are taken at the right angle of each other) at each extremity and in the middle of the pellet.

Stereometric methods applied to briquettes. Two different methods are applied; both need to measure the length of the sample twice (measures are taken at the right angle of each other, at the nearest 0,1 mm).

The first method (St 1) consists of measuring all the dimension of the briquettes to the nearest 0,1 mm (if necessary: diameter, length, width, height, diagonal length…) and to estimate its volume by using mathematical formulas.

The second method (St 2) aims to estimate the surface of the cross section of the briquettes using a sheet of paper of known dimensions. The sheet is weighed and the outline of the briquette is drawn on the paper. Afterwards, the outline is precisely cut and weighed. The mass of the cut piece of paper and the one of the original sheet of paper allow determining the surface of the cross section of the briquette. The latest is multiplied by the length of the sample in order to estimate its volume.

Method tested on briquettes Two stereometric methods, the hydrostatic method using paraffin coating and two buoyancy methods (with and without paraffin coating) have been tested for the particle density estimation of briquettes. The number of laboratories involved in the tests of the different methods is shown in Table 3. Each participating laboratory had to perform 15 replications for each of the five fuels and each tested method.

Table 3: Particle density of briquettes, tested methods, number of laboratories involved and number of replication for each method

Tested method Abbreviation Laboratories involved n

Stereometric measurements 1 St1 5 15 Stereometric measurements 2 St2 3 15 Hydrostatic and paraffin coating HP 4 15 Buoyancy and paraffin coating BP 4 15 Buoyancy without paraffin coating BW 2 15

Method tested on pellets Five methods for estimating particle density of pellets have been tested by a round robin: two were using the buoyancy principle and two were using hydrostatic measurements. The effect of paraffin coating has also been tested by four laboratories and this for 15 different fuels. In addition, all the fuels were tested by stereometric methods by four laboratories. Each fuel was tested for each method and 15 replications. The tested methods are listed in Table 4.



Table 4: Particle density of pellets, tested methods, number of laboratories involved and number of replication for each method

Tested method Abbreviation Laboratories involved n

Hydrostatic and paraffin coating HP 4 15 Hydrostatic and wetting agent HWA 3 15 Buoyancy and paraffin coating BP 4 15 Buoyancy and wetting agent BWA 3 15 Stereometrics measurements St 4 15

4.1.2 Analysed parameters

In order to evaluate the influence of each tested method and each selected fuel on the particle density results, mean values and standard deviations have been calculated (through the 15 replications and the participating laboratories).

Moreover, the repeatability (by laboratory and mean value) and the reproducibility for each tested method have been calculated following ISO 5725 2 standard [8].

Finally, the number of replications needed to secure a given accuracy (0,5; 1; 3 or 5%) has been calculated [4, 9]. This parameter has been evaluated for each pellet and briquette type, for each tested method. Other parameters have also been studied, for details see deliverable II.4.D3 [14].

4.1.3 Results and discussion

Briquettes

Mean values and standard deviation Figure 1 shows the results (mean value for all laboratories and standard deviation) of the particle density estimation for the 5 briquettes tested by the round robin test.

It is obvious that the briquette type tested influences the variability of the results. For example, briquettes B5 and B1 show a higher variability with all the tested methods. This could be due to intrinsic physical properties of the fuel as surface roughness and the presence of a central hole. It can also be seen that stereometric methods lead to more variable results compared to immersion methods.

A one-way analysis of variance (ANOVA) shows that there is no method for which the given results are all significantly equivalent for the different laboratories. Testing the mean equality allows to conclude that the two stereometric methods are equivalent for four of the tested briquettes and the methods are significantly different for the last one (B1 briquette type). It also permits to highlight that, except twice for one laboratory, the two methods using samples coated with paraffin (hydrostatic and buoyancy) are equivalent.



Particle density of 5 briquette types, estimated by 5 different methods (mean of involved laboratories)

0,95

1,00

1,05

1,10

1,15

1,20

1,25

1,30

briquette types

Parti

cle

dens

ity

St 1St 2HPBPBW

B3 B4 B5 B1 B2

Figure 1: Particle density of 5 briquette types estimated by 5 different methods (mean value and standard deviation of involved laboratories)

Repeatability, reproducibility and number of replications Table 5 shows the repeatability (for each laboratory involved in the round robin, and the mean value), the reproducibility and the number of repetitions needed in order to gain a given accuracy level (0,5%; 1%, 3% and 5%).

It can be concluded, from the repeatability and reproducibility results for briquettes, that both stereometric methods lead to higher values (higher variability) than the ones gained by liquid displacement methods.

It has to be added that the briquette type clearly influences the repeatability and the reproducibility calculated for each method and each single briquette type.

This table clearly shows that the buoyancy method-BW (without paraffin coating of the sample) needs the less replications to gain a given accuracy level. For example, to secure an accuracy level of 3%, only 3 replications are needed, while 8 are necessary for the stereometric 2 method.

It also appears that it seems difficult to gain an accuracy level lower than 1%, for the particle density estimation of briquettes. Indeed, even using the buoyancy pure water method, 26 replications are needed. Nevertheless, the here presented data are the ones calculated for the most variable briquette types, but if other briquettes are considered, the needed number of replications is far smaller (e.g. the minimum number of replications to obtain a precision of 0,5% is 4, when using buoyancy pure water method on briquette type B3).



Table 5: Briquettes particle density testing: repeatability, reproducibility and number of replications needed for a given accuracy level for the tested methods (NP: not performed)

r r%

Laboratories Laboratories N (for a given accuracy level)

Met

hod

1 2 3 4 5 Mea

n

1 2 3 4 5 Mea

n R R%

0,5% 1% 3% 5%

St1 0,08 0,09 ,08 0,06 0,07 0,08 6,82 7,68 7,28 5,40 6,26 6,75 0,11 10,17 257 64 7 3

St2 0,08 0,07 0,10 NP NP 0,08 7,29 6,21 8,66 NP NP 7,46 0,11 9,56 300 75 8 3

HP 0,05 0,04 0,05 0,03 NP 0,05 4,47 3,43 4,62 2,91 NP 3,92 0,063 5,39 165 41 5 1

BP 0,06 0,04 0,05 0,04 NP 0,05 4,69 3,69 3,80 3,61 NP 3,99 0,069 5,91 141 35 4 1

BW NP 0,05 NP 0,05 NP 0,05 NP 4,59 NP 4,16 NP 4,37 0,061 5,24 105 26 3 1

Pellets

Mean values and standard deviation Figure 2 shows the results (mean value for all laboratories and standard deviation) of the particle density estimation for the 15 pellet types tested by the round robin test.

It is obvious, as for briquette testing, that the tested pellet type influences the variability of the results. But, in case of pellet, the variability seems to be influenced by the factor level: it seems that the lowest the particle density of the pellet is, the higher the variability is.

It may also be noted that stereometric method leads to more variable results compared to liquid displacement methods. Additionally, wetting agent methods give higher particle density results compared to paraffin methods.

A one way analysis of variance (ANOVA) shows that there is no method for which the given results are all significantly equivalent for the different laboratories. On the 15 tested pellet types, an analysis of variance has shown that the hydrostatic with paraffin method (HP) gives equivalent results to the buoyancy with paraffin method (BP). For the two methods using wetting agent (BWA and HWA) the means are significantly equal for 11 pellet types.



Particle density of 15 pellets types estimated by 5 different methods (mean of involved laboratories)

0,9

1

1,1

1,2

1,3

1,4

Pellet types

Parti

cle

dens

ity

HPBPStBHWABWA

P15 P3 P10 P7 P9 P12 P6 P1

P5P8 P13 P2P11P14 P4

Figure 2: Particle density of 15 briquette types estimated by 5 different methods (mean value and standard

deviation of involved laboratories)

Repeatability, reproducibility and number of replications Table 6 shows the repeatability (for each laboratory involved in the round robin, and the mean value), the reproducibility and the number of repetitions needed to gain a given accuracy level (0.5 %; 1 %, 3 % and 5 %). Those parameters are calculated for wood pellets, for agricultural pellets and for all pellets together.

It may be noted that, for wood pellets, both stereometric and hydrostatic with paraffin methods lead to higher repeatability and reproducibility values (higher variability) than the ones gained by wetting agent methods (both hydrostatic and buoyancy).

It has to be added that the pellet type influences the repeatability and the reproducibility calculated for each method and each group of pellet types. Pellets made of agricultural residues clearly have higher values for those parameters. This table clearly shows that the buoyancy method-BWA (without paraffin coating of the sample and wetting agent added to water) needs less replications to gain a given accuracy level. For example, to secure an accuracy level of 3 %, only 2 replications are needed for wood pellets, while 19 are necessary for the stereometric method.

It also appears that it seems difficult to gain an accuracy level lower than 1 %, for the particle density estimation of pellets. Indeed, even using the buoyancy wetting agent method, 32 replications are needed, considering all the pellet types.

Nevertheless, the here presented data are the ones calculated for the most variable pellet types, but for other pellets the number of replication needed is far smaller (e.g. the minimum number of replications to obtain a precision of 0.5 % is 2, when using buoyancy with wetting agent method on pellet type P1).



Table 6: Pellets particle density testing: repeatability, reproducibility and number of replications needed for a given accuracy level, for the tested methods (NP: Not Performed)

r r%

Laboratories Laboratories

N (for a given accuracy level)

Met

hod

1 2 3 4 Mea

n

1 2 3 4 Mea

n R R%

0,5 1 3 5

Wood pellets

St 0,10 0,11 0,11 0,06 0,09 7,90 8,56 8,78 4,66 7,66 0,12 9,36 1137 284 32 11

HP 0,07 0,07 0,11 0,05 0,08 5,65 6,05 9,18 4,15 6,52 0,11 9,59 285 71 8 3

HWA NP 0,04 0,06 0,04 0,04 NP 3,08 4,43 2,82 3,52 0,05 4,13 240 60 7 2

BP 0,05 0,06 0,08 0,07 0,07 4,14 5,54 6,66 5,75 5,60 0,08 6,75 124 31 3 1

BWA NP 0,04 0,02 0,05 0,04 NP 2,88 1,41 3,88 2,91 0,04 3,47 127 32 4 1

Agricultural residues pellets

St 0,16 0,01 0,15 0,08 0,13 14,05 11,08 12,96 6,75 11,55 0,16 13,85 644 161 18 6

HP 0,09 0,10 0,11 0,09 0,10 8,08 9,25 9,80 8,31 8,89 0,12 11,44 417 104 12 4

HWA NP 0,10 0,10 0,06 0,09 NP 8,29 8,58 5,20 7,51 0,16 13,59 391 98 11 4

BP 0,08 0,09 0,11 0,08 0,09 7,82 8,98 10,8 8,02 9,00 0,12 11,64 367 92 10 4

BWA NP 0,06 0,03 0,06 0,05 NP 5,12 2,67 5,38 4,56 0,07 6,33 79 20 2 1

All pellets

St 0,12 0,11 0,12 0,06 0,11 9,75 9,23 9,85 5,26 8,73 0,13 10,63 1137 284 32 11

HP 0,07 0,08 0,11 0,06 0,08 6,33 6,80 9,33 5,41 7,12 0,12 10,04 417 104 12 4

HWA NP 0,06 0,07 0,04 0,06 NP 5,21 5,69 3,53 4,90 0,10 7,80 391 98 11 4

BP 0,06 0,07 0,07 0,07 0,07 5,17 6,27 6,59 6,28 6,10 0,09 7,75 367 92 10 4

BWA NP 0,05 0,02 0,05 0,04 NP 3,77 1,79 4,29 3,46 0,05 4,40 127 32 4 1

4.2 Durability

The physical quality of densified fuels like pellets and briquettes is mainly described by its durability and particle density. Durability is the resistance measurement of fuels towards shocks or/and friction. Therefore, it is an important quality parameter with regard to the handling and transportation processes of briquettes and pellets. Particles with a higher durability guarantee that a bigger part will arrive in one piece, without falling apart into smaller pieces or dust. The dust influences the combustion process and emissions negatively. The presented research aims at the identification and the evaluation of the best appropriate methods for briquettes and pellets durability determination.



4.2.1 Briquettes durability test

The test sample (2 liters) is weighed and subjected to controlled shocks by collision of briquettes against each other and against the walls of a defined rotating test chamber (see Figure 3). The durability is calculated from the mass of sample remaining after separation of abraded and fine broken particles and expressed as a percentage of the initial mass. The remaining briquettes are sieved for 2.5 minutes by a sieve having an aperture size as close as possible to 2/3 of the briquettes diameter.

Figure 3: Principle of briquette durability tester

The number of rotations is tested in order to define the best appropriate rotating period (105, 210, 315, 420 and 630 rotations). Each rotating period is considered as a durability value.

Moreover, a durability index is calculated. After each rotating period, all the material is replaced in the drum for another period till 630 rotations are reached. It is possible to draw a durability curve with all the values obtained at the given rotating periods. The durability index is defined as the ratio between the area under the durability curve and the area related to a non-abraded material.

An international round robin test has been organized with five laboratories and for five different briquette types (see Table 1).

Each rotating period and the durability index were considered as different methods and compared together following the ISO 5725 standard [8].

4.2.2 Pellets durability test

Two main standards have been described as giving useful results for pellets durability tests: the ASAE S269.4 DEC 96 [3] and the ÖNORM M 7135 [2]. These methods have been compared to the briquette durability drum (described at section 4.2.1) and to a Spanish method, which mainly differs to the briquette tester by the number of rotations and by the volume of the sample.



The ASAE Standard refers to a tumbling device made of boxes in aluminum or stainless steel with inner dimensions of 300 × 300 × 125 mm. In order to enforce the tumbling effect each box is equipped with a 230 mm long baffle, which extends 50 mm into the box. The baffle is affixed symmetrically to a diagonal of one side of the box. Rivets and screws are kept to a minimum and well rounded. The box rotates on an axis, which is centered perpendicular to the sides of the box. The rotation speed is fixed to 50 rpm. For the tests described here, a 500 g sample is tumbled for 500 rotations before being sieved manually on a 3.15 mm round holes sieve [4], the durability is expressed by the percentage in mass of the pellets remaining on the sieve.

The Ö-Norm Standard refers to a pneumatic device (commercial name: Ligno-Tester LT II of Messrs. Borregaard Lignotech), in which the test sample is subjected to controlled shocks by collision of pellets against each other and the perforated walls of a test chamber, being swirled by a defined air stream. The test chamber has a four-sided (2 mm round holes sieve) pyramid form, which is orientated with the tip downwards. The inside dimensions of this pyramid are 230 ± 5 mm and 126 ± 10 mm high. For the tests described here, the fines are removed, before testing, by sieving manually with a 3.15 mm round holes sieve [4] and a 100 ± 0.1 g sample is placed in the pellet chamber and blown for 60 seconds at a pressure of 70 mbar. After treatment, the remaining pellets in the tester chamber are weighed. The durability is expressed by the percentage in mass of the pellets remaining on the sieve.

The procedure and the briquette durability tester are described at section 4.2.1, but in this case, fines are removed by using a 3.15 mm round hole sieve. The results presented here correspond to a 105 rotations of the drum.

The Spanish method use the same principle as described in section 4.2.1. The procedure of this method use a sample of 2.5 kg which is tumbled for 15 minutes at 20.4 rpm. The remaining pellets are sieved manually or mechanically for 2.5 minutes by a sieve having an aperture size as close as possible to 2/3 of the pellets diameter.

An international round robin test has been organized with five participants and for 15 different pellets types which are representative of the market of the countries involved in the trials: Austria, Belgium, Denmark, Germany and Spain. The tested pellets are presented in Table 2.

Methods have been compared together following the ISO 5725 [8].

4.2.3 Results and discussion

Briquettes durability Figure 4 shows the results (mean value and standard deviation) for one laboratory of the durability estimation for the 5 briquette types tested in the round robin test. As expected, the values of the durability decrease as the number of rotations increase. But the variability of the durability measurements increases when the number of rotations of the drum is higher.

This has been the case for all the tested briquette types and for all laboratories included in the round robin test on briquette durability testing. The variability of the durability index



is close to the variability obtained with 210 rotations. It is also to be noted that the tested briquette type influences the variability of the results.

Influence of the number of rotations on the durability result

0

20

40

60

80

100

B 1 B 2 B 3 B 4 B 5Briquette types

Dur

abilit

y %

105 rot.

210 rot.

315 rot.

420 rot.

630 rot.

Figure 4: Influence of the number of drum rotations on the durability result standard deviation

(Laboratory 2)

The relation between the global mean of durability (through all the laboratories) and the single laboratory measurements shows that the variability of durability results, between laboratories, decreases as the durability of the briquettes increases (see Figure 5). For low quality, the results are more variable than for high quality briquettes.

Influence of the briquettes durability level on the variability of the results (105 rotations)

20

40

60

80

100

40 50 60 70 80 90 100General mean of durability for the 5 briquettes

Labo

rato

ry

mea

sure

men

ts

Lab 1Lab 2Lab 3Lab 4Lab 5

Figure 5: Briquettes durability level influence on the result variability (105 rotations)

Table 7 shows, for each tested method, the repeatability (for each laboratory involved in the round robin), and the mean value, the reproducibility and the number of repetitions needed in order to achieve a given accuracy level (2 %, 3 %, 5 % and 10 %).

It can be concluded, from the repeatability and reproducibility results for briquettes, that both parameters are really high for all the tested methods. It has to be added that the briquette type clearly influences the repeatability and the reproducibility calculated for each method and each single briquette type.

This table indicates that the method using 105 rotations needs less replications to achieve a given accuracy level. For example, to secure an accuracy level of 10 %, 35 replications are



needed, while 1919 should be necessary for the 420 rotations method. Nevertheless the accuracy level, that can be reached, are large and the number of replications needed is high.

It also appears that it seems difficult to achieve an accuracy level lower than 10 % for the durability estimation of briquettes. Indeed, even with the 105 rotations method, the time necessary to perform the test is really long (105 rotations test needs 5 minutes to be achieved).

Table 7: Briquettes durability testing: repeatability, reproducibility and number of replications needed for a given accuracy level for the tested method (r: rotations)

r r%

Laboratories Laboratories N (for a given accuracy

level)

Met

hod

1 2 3 4 5 Mea

n

1 2 3 4 5 Mea

n R R%

2% 3% 5% 10%

105 r 16,97 9,13 8,80 8,51 15,13 12,25 20,78 11,18 10,77 10,43 18,52 15,00 33,8 49,3 879 390 141 35

210 r 21,25 11,72 16,62 13,54 11,62 15,38 30,23 16,68 23,65 19,27 16,54 21,89 33,4 40,9 10625 4722 1700 425

315 r 20,33 9,25 15,17 20,02 11,43 15,88 31,46 23,48 30,98 17,69 24,58 43,9 62,6 18279 8124 2925 731

420 r 20,59 11,88 15,74 21,81 11,37 16,84 34,08 19,67 26,05 36,10 18,83 27,87 40,5 62,7 47987 21327 7678 1919

630 r 20,50 14,65 13,73 26,58 11,81 18,27 37,70 26,95 25,26 48,89 21,72 33,61 38,4 63,5 24324 10811 3892 973

Index 15,96 8,50 11,92 16,31 9,65 12,87 23,26 12,39 17,38 23,78 14,07 18,76 37,1 68,2 1198 532 192 48

The results are much more variable for the less durable briquettes. The difference is pronounced between the briquettes B4 (mean value of durability through all partners of 46,4+/-14,8), B1 (mean value of durability through all partners of 71,5+/-19,9) and the group made of B5 (mean value of durability through all partners of 95+/-2,1), B2 (mean value of durability through all partners of 98,5+/-0,5), B3 (mean value of durability through all partners of 96,9+/-1,7).

The briquettes were separated in two groups regarding their durability: durability < 90 and durability ≥ 90. The repeatability are given in the following Table 8 et seqq.:

Table 8: Repeatability for the briquettes with a durability ≥90 (Briquette tester 105 rotations)

Lab 1 Lab 2 Lab 2 Lab 4 Lab 5

r 6,25 3,33 3,47 1,89 3,33 r% 6,45 3,44 3,58 1,95 3,44

Table 9: Repeatability for the briquettes with a durability <90 (Briquette tester 105 rotations)

Lab 1 Lab 2 Lab 2 Lab 4 Lab 5

r 25,72 13,85 13,24 13,26 23,56 r% 43,63 23,49 22,46 22,49 39,97



The reproducibility was also calculated for the two groups of briquettes.

Table 10: Reproducibility for the briquettes (Briquette tester 105 rotations) for both classes

Durability <90 Durability ≥90

R 52,58 4,56 R% 89,20 4,71

Regarding those repeatability and reproducibility values, it is obvious that the expected values should be very different between both defined groups.

Pellets durability The selected pellets had a durability range from 88.3 to 99.4 %, for all tested methods and all the laboratories (see Figure 6). Figure 6 shows the mean results of pellets durability, for all laboratories included in the round robin test, the 4 tested methods and the 15 selected pellet types.

The tested pellet type influences the variability of the results given by the different tested methods. But, in the case of pellets, the variability also seems to be influenced by the factor level: the lower the pellets durability is, the higher the variability is.

Durability of the selected pellets estimated by 4 different methods

88

90

92

94

96

98

100

P8 P11 P2 P6 P15 P14 P10 P9Pellets type

Dur

abilit

y%

Briquette tester

Spanish method

ÖNORM

ASAE

Figure 6: Durability of the selected pellets estimated by the tested methods (mean value of involved

laboratories)

As for briquettes, the pellets quality influences the variability of the durability measured by the different laboratories (see Figure 7), especially for durability below 96 %. The lower the durability, the higher is the variation between laboratories.



Influence of the level of durability on the variation of the result between laboratories, using ÖNORM

88

90

92

94

96

98

100

88 90 92 94 96 98 100

General mean of durability

Dur

abilit

y %

at o

ne

labo

rato

ry Lab 1Lab 2Lab 3Lab 4Lab 5

Figure 7: Influence of the level of durability on the variation of the result between laboratories (15 Pellets

types with ÖNORM)

Table 11 shows, for ASAE and ÖNORM methods, the repeatability (for each laboratory involved in the round robin), and the mean value, the reproducibility and the number of replication needed to achieve a given accuracy level (0.1 %, 0.5 %, 1 %, 3 %). These parameters are calculated for 6 mm wood pellets, 8 mm wood pellets, agricultural residues pellets and for all pellets together. Additionally, these values have been calculated for pellets having a durability over 97.5 % (which corresponds to the best durability class in CEN technical specifications [11]).

It may be noted that, for wood pellets, ÖNORM standard lead to higher repeatability and reproducibility values (higher variability) than the ones obtained by following ASAE standard.

It has to be added that the pellet type influences the repeatability and the reproducibility calculated for each method and each group of pellet types. Pellets made of agricultural residues clearly have higher values for these parameters. The lowest values are obtained for pellets having a durability higher than 97.5 %. It also appears that 8 mm pellets have better repeatability and reproducibility than pellets of 6 mm.

Table 11 shows that the tumbling device needs less replications to achieve a given accuracy level. For example, to secure an accuracy level of 1 %, only 6 replications are needed for the whole pellets group, while 22 are necessary for the pneumatic device.

It also appears that it seems difficult to achieve an accuracy level lower than 0.5 %, for the durability estimation of pellets. Indeed, even using the tumbling device, 24 replications are needed, considering all the pellet types.

Nevertheless, the data presented here are the ones calculated for the most variable pellet types, but for other pellets, the number of replications needed is far smaller (e.g. the minimum number of replications to obtain a precision of 0.1 % is 2, when pneumatic device is used on pellets having a durability over 97.5 %).



Table 11: Pellets durability testing: repeatability, reproducibility and number of replications needed for a given accuracy level, for the tested methods (NP: not performed)

r r %

Laboratories Laboratories N (for a given accuracy level)

Met

hod

1 2 3 4 5 Mea

n

1 2 3 4 5 Mea

n R R%

0,1% 0,5% 1% 3%

All pellets

ASAE NP NP 1,07 0,57 NP 0,86 NP NP 1,11 0,59 NP 0,89 1,41 1,46 595 24 6 1

Önorm 2,15 1,74 2,28 1,61 2,70 2,13 2,23 1,81 2,37 1,67 2,80 2,21 3,84 3,99 2063 83 22 2

6 mm diameter


Önorm 1,92 1,63 2,48 1,84 2,63 2,13 2,00 1,69 2,58 1,91 2,73 2,22 3,79 3,94 2063 83 22 2

8 mm diameter

ASAE NP NP 0,21 0,28 NP NP NP 0,21 0,28 NP 0,25 1,36 1,39 6 1 1 1

Önorm 0,69 0,34 0,86 0,49 0,58 0,62 0,70 0,35 0,88 0,50 0,59 0,63 2,89 2,99 82 3 1 1

Agricultural residues pellets


Önorm 3,10 2,45 2,58 1,67 3,62 2,76 3,26 2,58 2,72 1,76 3,82 2,91 7,12 7,50 1218 49 12 1

Wood pellets DU97,5


Önorm 0,28 0,31 0,28 0,17 0,37 0,29 0,28 0,31 0,28 0,17 0,37 0,29 0,99 1,00 9 2 1 1

Those results have been confirmed by the extra round robin test organized in October 2003 by CRA with 4 laboratories and 11 pellets type [15]. The results are summarized in table 12.

Table 12: Repeatability and reproducibility for 11 pellets obtained during the round robin test of October 2003 with 4 partners.

ASAE Standard Mean square Pellet 1 Pellet 2 Pellet 3 Pellet 4 Pellet 5 Pellet 6 Pellet 7 Pellet 8 Pellet 9 Pellet 10 Pellet 11Repeatability 0,437 0,366 0,133 0,229 0,090 0,142 0,313 0,061 0,217 0,189 0,180

r % 13,232 8,676 6,207 10,382 10,310 10,475 12,974 9,601 17,616 17,307 9,566

Reproducibility 0,846 0,366 0,287 0,356 0,254 0,284 0,511 0,147 0,307 0,373 0,342

R % 25,579 8,676 13,382 16,152 29,192 20,930 21,198 23,078 24,875 34,181 18,180

Ligno-tester Repeatability 0,818 3,447 0,293 0,584 0,196 0,651 0,884 0,280 0,435 0,542 0,259

r % 25,407 49,702 15,222 27,002 15,596 38,474 33,552 26,391 24,811 35,847 19,207

Reproducibility 2,149 5,064 1,700 2,285 1,062 2,062 2,266 1,407 1,696 1,820 1,127

R % 66,741 73,021 88,408 105,609 84,390 121,919 85,953 132,775 96,802 120,272 83,721



Relation between results obtained with the tumbling device and the pneumatic device

Relation between ASAE and ÖNORM for durability of pellets

y = 0,4984x + 49,351R2 = 0,6852

95

96

97

98

99

100

91 92 93 94 95 96 97 98 99 100

Durability % with ligno-tester

Dur

abilit

y %

follo

win

g As

ae

stan

dard

Figure 8: Relation between the two methods for pellets durability testing [15]

There is no clear relation between the results of the durability of pellets measured by the tumbling device and those given by the pneumatic device. It is thus difficult to extrapolate the results from a method to another.

4.3 Relation between durability and particle density

The relation between durability and particle density has been studied using the more accurate methods identified during the tests.

4.3.1 Briquette

The following figure presents the relation between durability and particle density for briquettes. Durability has been measured using the drum described in 4.2.1 with a treatment of 105 rotations and the particle density has been estimated with the buoyancy method with water.

Relation between particle density and durability correlation coefficient 0,19

2030405060708090

100

1 1,05 1,1 1,15 1,2 1,25 1,3

Particle density briquettes - BW method

Dur

abilit

y br

ique

ttes

- BD

5

min

Figure 9: Relation between particle density and durability for briquettes.



It clearly appears that no relation between the particle density and durability has been found. The coefficient of correlation between the two parameters is equal to 0,19. It will not be possible to predict the value of durability using the particle density, or the other way round, the particle density using the durability.

4.3.2 Pellet

Figure 10 presents the relation between durability and particle density for pellets. Durability presented in the graph was measured using the tumbling device (described in 4.2.2) and the particle density was estimated with the buoyancy method with wetting agent.

Relation between density and durability correlation coefficient 0,57

92

93

94

95

96

97

98

99

100

1,1 1,15 1,2 1,25 1,3 1,35

Particle density pellets - BWA method

Dur

abilit

y pe

llets

- M

AT m

etho

d

Figure 10: Relation between particle density and durability for pellets

For pellets, the relation between particle density and durability is characterized by a coefficient of correlation of 0,57.


5.1 Particle density

In order to estimate the particle density of briquettes, five methods were tested in an international round robin test, involving 5 laboratories with five briquette types commercially available.

Regarding the different results obtained during the project, it is clear that the methods based on liquid displacement: hydrostatic and buoyancy methods (HP, BP and BW) gave better results than stereometric methods.

For each analysed parameter the two methods based on physical measurements (ST1 and ST2 methods) led to higher variability, higher bias between labs, higher values of repeatability and reproducibility and needed more replications to reach a given accuracy.



The liquid displacement methods showed similar standard deviation. The values of repeatability and reproducibility were similar for the 3 methods based on liquid displacement.

For all tested methods, it clearly appears during this analysis that the briquette type influences the variability of the results.

Regarding the results, the selection of a method inside the group of liquid displacement methods as a standard can be made considering other aspects like practical facilities. The choice of the method used may be based on the material available at the lab, hydrostatic with paraffin coating method needs to have a balance which can weight between 10 and 15 kg with an accuracy of 0,1g and the buoyancy with paraffin coating method needs to have a balance which allows to weight by the bottom.

The buoyancy without paraffin coating method decreases the time of sample preparation (no coating with paraffin) but the water has to be often changed (sometimes after each replication) because briquettes start to disintegrate very fast and fall apart in the water.

The following table summarizes the expected values of repeatability and reproducibility with the liquid displacement methods for briquettes particle density testing (based on 15 replications).

Table 13: Repeatability, reproducibility and accuracy reached with a number of replications of 15 (liquid displacement methods) for briquette.

Methods r r% R R% Accuracy reached with n=15

HP 0,05 3,92 0,06 5,39 1,66% BP 0,05 3,99 0,06 5,91 1,53% BW 0,05 4,37 0,06 5,24 1,32%

For pellets particle density testing, 5 different methods were also tested in an international round robin test involving 5 laboratories. 15 pellet types (either wood or agricultural residues pellets) representing the actual or potential pellets market were tested. The buoyancy method using non-coated samples and wetting agent mixed with water gave the lowest values in repeatability and reproducibility (lowest variability). This method also needs the less replications to fulfil a given accuracy level. The stereometric method used for particle density determination led to higher variability compared to the liquid displacement methods.

But in practice, it seems illusive trying to reach an accuracy higher than 1 %, for the tested methods. Indeed, over this limit, the number of repetitions needed is far too high for field measurement. The fuel itself influences the variability of the method and, for given fuel types, the number of replications needed to reach a given accuracy level is smaller than for the others. It could be suggested to determine on which other parameters the number of replications (or the expected accuracy) could be based.

The following Table 14 summarizes the expected values of repeatability and reproducibility with the buoyancy with wetting agent methods for pellets particle density testing (based on 15 replications).



Table 14: Repeatability, reproducibility and accuracy reached with a number of replications of 15 (BWA method) for the different groups of pellet.

Pellets r r% R R% Accuracy reached with n=15

All pellets 0,043 3,46 0,054 4,40 1,45% Agricultural residues pellets 0,053 4,56 0,074 6,33 1,15%

Wood pellets 0,067 5,81 0,043 3,47 1,45%

5.2 Durability

To estimate the durability of briquettes, six methods were tested in an international round robin test involving 5 laboratories with five briquette types commercially available.

For the pellets durability testing, 4 different methods were also tested in an international round robin test involving 5 laboratories. 15 pellet types (either wood or agricultural residues pellets) representing the actual or potential pellets market have been tested

The most repeatable and reproducible method for the estimation of the briquettes durability is to tumble the briquettes for 105 rotations (or 5 min treatment). Nevertheless, briquettes durability testing leads to highly variable results and it seems illusive to try to reach better accuracy than 10%. But, as the fuel itself influences the variability of the method, and as for given fuel types, the number of replications needed to reach a given accuracy level is far smaller than for the others, it could be suggested to determine on which other parameters the number of replications (or the expected accuracy) could be based.

The maximum values of repeatability and reproducibility obtained during the round robin test by the 5 laboratories with the briquettes tester (105 rotations, 5 replications) are given in Table 15.

Table 15: Repeatability, reproducibility and accuracy reached with a number of replications of 5 (Briquette tester 105 rotations) for both groups of briquettes.

Durability r r% R R% Accuracy reached with n=5

Durability <90 25,72 43,63 52,58 89,20 27% Durability ≥90 6,25 6,45 4,56 4,71 3%

Concerning the pellets durability determination, the tumbling device gives the most repeatable and reproducible results for the measurement of the pellets durability. Moreover, this method needs the less replications to reach a given accuracy level, while the pneumatic device needs the higher number of replications. In practice, an accuracy level of 1 %, could be reached by the method based on ASAE standard. Indeed, over this limit, the number of repetitions needed is far too high for field measurement.

There is no accurate relation between the results of the durability of pellets measured by the tumbling device and those given by the pneumatic device. It is thus difficult to extrapolate the results from a method to the other.

The variation of durability results is highly influenced by the fuel itself, either between laboratories or inside the same laboratory. It clearly appears that the level of durability



influences the variability of results: the lower the pellet durability is, the higher the variability is.

The following Table 16 summarizes the expected values of repeatability and reproducibility obtained with the tumbling device (based on 3 replications).

Table 16: Repeatability, reproducibility and accuracy reached with a number of replications of 5 (ASAE) for the different groups of pellet.

Pellets R r% R R% Accuracy reached with n=3 All pellets 1,07 1,11 1,41 1,46 1,41% Agricultural residues pellets 1,01 1,06 2,08 2,18 0,75%

Wood pellets DU97.5 0,12 0,12 0,43 0,43 0,08%

5.3 Relation between particle density and durability

Comparing particle density and durability by using the most accurate methods for both parameters, no relation has been found between those two parameters. In consequence the particle density cannot be used to estimate the durability.

6 Recommendations

The recommendations for pellets particle density testing is to use a method based on liquid displacement: the buoyancy method with non coated sample in water with addition of wetting agent.

For the briquettes, the determination of the particle density can be made with a method based on the Archimedes principle: hydrostatic or buoyancy with paraffin coating or buoyancy without paraffin coating.

Concerning the durability of pellets, the recommended method uses a tumbling device (50 rpm, 10 minutes).

Regarding the briquettes durability the best appropriate method among all the tested methods is the briquette tester with 105 rotations.

7 Acknowledgements

The research was conducted within the European project "Pre-normative work on sampling and testing of solid biofuels for the development of quality assurance systems (BioNorm) ENK6-CT-2001-00556.



8 References

1 DIN 51731, 1996, Testing of solid fuels – Compressed untreated wood – Requirements and testing. Deutsches Institut für Normung (DIN), Beuth Verlag GmbH, Berlin.

2 ÖNORM M7135, 2000, Compressed wood or compressed bark in natural state – Pellets and briquettes – Requirements and test specifications. Österreichisches Normungsinstitut (ON), Vienna.

3 ASAE 269.4, 1996, Cubes, pellets and crumbles – Definitions and methods for determining density, durability and moisture content – the society for engineering in agriculture, food and biological systems (ASAE), St Joseph, USA.

4 Böhm T., Hartmann H., 2004, Measuring particle density of wood pellets, in Proceedings of the 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14/05/2004, Rome, Italy

5 Lequeux P., Carré J., Hébert J., Lacrosse L., Schenkel Y., 1990, Energie et biomasse, la densification, Les presses agronomiques de Gembloux, Belgium, 188 p

6 Lehtikangas P., 2000, Storage effects on pelletised sawdust, logging residues and bark, Biomass & Bioenergy, Pergamon

7 Carré J., Hebert J., Lacrosse L., 1983, Critical analysis of the dry process improvement of ligneous materials for energy-producing purposes, Final Report (CEC - ADE/958/83 - BDF/1), Centre de Recherches Agronomiques de l'état (CRA) Gembloux, 245 p

8 ISO 5725.2, 1994, Exactitude (justesse et fidélité) des résultats et méthodes de mesure ; Partie 2 : Méthodes de base pour la détermination de la répétabilité et de la reproductibilité d’une méthode de mesure normalisée.

9 Sachs, L, 1997, Angewandte Statistik - Anwendung statistischer Methoden. 8th edition, Springer-Verlag, Berlin, Germany, (Applied statistics - Application of statistical methods, in German Language)

10 ISO 3310-2 Test sieves – Technical requirements and testing – Part 2: Test sieves of perforated metal plate

11 Technical Specification CEN/TS 14961: Solid biofuel – Fuel specification and classes

12 Rabier.F, Temmerman.M Report on equipment for particle density determination (pellets and briquettes)II.4.D1 November 2003

13 Rabier.F, Temmerman.M Report on equipment for durability determination (pellets and briquettes)II.4.D2 August 2003

14 Rabier.F, Temmerman.M Report on interaction and fundamental findings in testing durability and particle density of pellets and briquettes II.4.D3 December 2004



15 Rabier.F, Temmerman.M Report on extra round robin test (ASAE/ÖNORM) October 2003

9 Glossary

St1 Stereometric measurement 1 method (particle density for briquette)

St2 Stereometric measurement 2 method (particle density for briquette)

St Stereometric measurement method (particle density for pellet)

HP Hydrostatic and paraffin coating method (particle density pellet and briquette)

BP Buoyancy and paraffin coating method (particle density pellet and briquette)

BW Buoyancy without paraffin coating method (particle density briquette)

HWA Hydrostatic without paraffin coating and addition of wetting agent method (particle density pellet)

BWA Buoyancy without paraffin coating and addition of wetting agent method (particle density pellet)

r Repeatability

r% Relative repeatability

R Reproducibility

R% Relative reproducibility

MC Moisture content % wet basis

StD Standard deviation

Part 2 Chemical Tests


Part 2.3 Chemical Tests

Part 2.3.1 Task III.1 - Sulphur, chlorine and nitrogen content

Final report prepared by: Martin Englisch1) Contributing co-authors: Susanne Westbourg2), Juan Carrasco3), Frits Bakker4), Edith Thomsen5), Christos Agrafiotis6), Raili Vesterinen7), Conny Haraldsson8), Knut Niebergall9)

1) OFI - Austrian Research Institute for Chemistry and Technology, Austria 2) FORCE Technology (former dk-Teknik Energy & Environment), Denmark 3) Centro de investigaciones energéticas, medioambientales y technologicas, Spain 4) Institute for Energy Process Engineering and Chemical Engineering, Freiberg, Germany 5) Elsam A/S (Techwise), Enstedvaerket, Denmark 6) Centre for Research and Technology Hellas - Process Engineering Research Institute, Greece 7) Technical Research Centre of Finland, Espoo, Finland 8) Swedish National Testing and Research Institute, Sweden 9) IfE Analytik GmbH, Germany

1 Summary

Solid biofuels contain sulphur, chlorine and nitrogen. For energy conversion processes, the knowledge of the concentration of these elements in the fuels is of high importance. The concentrations of sulphur and chlorine are of relevance for deposit formation and corrosion, SOx, HCl and PCDD/F emissions as well as aerosol formation. The fuel nitrogen content is responsible for NOx emissions.

Up to now European standards that define analytical procedures for the determination of these elements in solid biofuels are not available. Thus, different approaches and a wide variety of analytical methods are used. Aim of this task (III.1) of the BioNorm project was the evaluation of the existing methods and the development of best practise guidelines (see deliverable III.1.D3) for the correct determination of sulphur, chlorine and nitrogen in solid biofuels in order to support chemists analysing such materials. The best practise guidelines additionally contain recommendations for appropriate sample preparation, for the digestion (or combustion) of the samples and for the element detection procedures. This report should be understood as a supplement to the best practise guidelines summarizing selected research results that were the basis for the best practise guidelines and the corresponding European draft Technical specifications [1, 2, 3].

The work in task III.1 focussed on the comparison and evaluation of available analytical methods. Eight different biofuels, covering a wide range of concentrations of the investigated elements were used for the experiments performed. Based on the experimental results and the



experience of the participating laboratories, the best and most practical methods were selected and were further investigated for the most critical experimental parameters.

This task report summarizes:

• Selection of samples and sample preparation

• Inter-laboratory comparison of the methods developed and used in the participating laboratories

• Evaluation of the experimental results and discussion based on the experience of the involved experts

• Sensitivity analysis of some critical parameters especially for the bomb combustion method, such as the influences of the receiving solution, the oxygen pressure or the pressure release.

It was concluded, that the recommended method for total sulphur and total chlorine analysis is the combustion in an oxygen bomb and determination by ion chromatography. For sulphur analysis, acid digestion in a closed vessel and determination by ICP systems is also suitable. Since the chlorine content in solid biofuels is mainly present in form of water soluble chlorides, the determination using the method for water soluble chlorine is an alternative and simple method to achieve information about the chlorine content. For nitrogen determination, the use of automatic analysers based on combustion and detection of the gaseous compounds is the recommended method. The Kjeldahl method is especially suitable for samples with low nitrogen contents.

2 Objectives

The objectives of Task III.1 were:

• to investigate the applicability of existing methods for the determination of sulphur, chlorine and nitrogen for a wide range of biofuels. The selection and preparation of the investigated biofuels is summarized in chapter 3.

• to obtain a comprehensive evaluation and identification of the best appropriate methods for the determination of sulphur, chlorine and nitrogen in solid biofuels. After the evaluation of national and international experienc, the most promising methods available in the laboratories of the participating institutes were selected for experimental evaluation (see chapter 4).

• to evaluated the performance of the existing laboratory methods and the respective equipment in comparative trials under technical, economic and work-efficiency aspects.

• to summarize the results of this evaluation and develop best practise guidelines for the determination of sulphur, chlorine, nitrogen, major and minor elements (see deliverable III.1.D3 of the BioNorm project, [13]). The best methods are furthermore currently standardized on European level in TC 335 WG5 "Solid Biofuels, chemical test methods” [1, 2, 3].



• to investigate experimental influences on the performance of the methods (sensitivity analysis). Selected results are compiled in chapter 4.7.

Conditions influencing the measurements, especially the sample preparation and particle size reduction but also parameters like moisture content, handing procedures, machine variables were investigated and discussed during the research work. Notes in the best practise guidelines and the corresponding draft standards are added were appropriate.

3 Selection and preparation of the biofuel samples

At the task kick-off meeting in Vienna (April 2002) 8 biofuel samples were selected with the intention to cover a wide range of biofuels relevant for the market (coniferous wood without bark, woodchips, bark, hardwood with glue, straw, rape straw, orujillo (olive residues) and cynara2, see Table 1). The selected fuels cover the most frequently traded biofuels in Europe and cover also a wide range of concentrations of the investigated elements sulphur, chlorine and nitrogen. The original samples were delivered from partners in Spain, Denmark, the Netherlands, Germany and Austria to either Stuttgart (3 samples) or Vienna (5 samples) for sample preparation.

The sample preparation was more difficult than expected. Especially the amount of the samples (60 kg each) exceeded the laboratory capacity. Furthermore, the milling and homogenisation showed unexpected difficulties. Take the biomass sample orujillo as example, which contained, a high amount of stones and other impurities, due to the natural storage on ground (see Figure 1). Thus, the sample had to be sieved and had to be cleaned manually before milling and homogenisation could proceed. Other samples (e.g. cynara, bark) had to be milled in steps because of their large original particle size. Furthermore, the naturally different parts of some plants caused problems during sample preparation and homogenization. The sample cynara, as example, contained stems, seeds and other plant parts, (see Figure 2) that behave different when fed into mills. The very light seeds tend to be lost by small air movements during the sample handling while the stems are difficult to mill due to their tearproof fibres. Some of the experience from the sample preparation is included in the “best practice guideline for sample preparation” prepared delivered in task III.2. [4].

2 Cynara (Cynara Cardunculusis L.) is a herbaceous perennial plant with high lignocellosic biomass yield that

can be also used as solid biofuel.



Table 1: Samples used in task III.1 for the determination of sulphur, chlorine and nitrogen

The sample preparation in the ofi was done in two steps using a coarse cutting mill equipped with a 10 mm sieve and a laboratory cutting mill (Fritsch Pulverisette 19) equipped with WC cutting tools and a 1 mm sieve (alternatively 0.5 mm), see Figure 3.

After the sample preparation homogeneity tests were carried out on all eight samples using bomb combustion and determination of chloride and sulphate by IC. The results of sulphur and chlorine determinations were evaluated with statistical methods according to ISO 5725. After successful homogeneity testing (no possible or statistical outliers were identified) the samples were distributed to the partners.

Figure 1: Mineral impurities and other screenings of the sample orujillo

3 This sample was prepared in the last project year to verify the draft methods that were developed

Requirement/specification Sample Delivered by Prepared by Also used in

Woodchips (with bark) woodchips Stuttgart USTUTT II.2, III.2

Wood, low N, Cl, S coniferous wood without bark Austria ofi

Wood, residues hardwood with glue Netherlands USTUTT

Bark oak-bark Austria ofi

S-rich fuel rape straw Austria ofi

Cl-rich (and difficult combustion behaviour) wheat straw Denmark USTUTT II.2, III.2

awkward biofuel: (low ash melting point, high halogen and N-content)

cynara cardunculus Spain ofi

agricultural residues olive residues (orujillo) Spain ofi Additional sample for verification of the draft methods3 Hemp Austria ofi



Figure 1: Seeds, stems and woody parts of the original sample cynara

Figure 3: Interior view of the cutting mill used for sample preparation (Fritsch Pulverisette 19; WC tools)

Figure 4: Samples after preparation



4 Description of experiments and results

4.1 Methods for the determination of sulphur, chlorine and nitrogen in solid biofuels

For the determination of total chlorine and total sulphur a wide variety of analytical methods is used for different samples ranging from solid fuels to waste, minerals, soil or material samples. The existing methods include these principles:

• combustion followed by methods for the quantification of the elements or molecules

• digestion followed by methods for the quantification of the elements or molecules

• leaching, applying different solvents

• non destructive methods like XRF.

The following methods were selected and evaluated for their suitability for solid biofuels:

• combustion in closed oxygen bomb

• high temperature combustion in tube furnace

• combustion according to the Wickbold-method

• combustion and detection in automated analyzers (e.g. CHN or CHNS-Analyzers, AOX-Analyzer)

• acid digestion in closed vessel

• acid digestion in open vessel

• digestion according to the Eschka-method

• Kjeldahl-method

• neutron activation analysis (reference method)

• x-ray fluorescence analysis (XRF)

• ICP, IC and other methods were used for determination of sulphate and chloride

Standard procedures exist for most methods on an international or European level, but these do not exist for solid biofuels in general. At the task kick-off meeting the available methods were compiled, as stated in Table 2 and Table 3.

The methods used in the participating laboratories were originally based on standards (especially for coal analysis) but the available equipment and the individual experience with biofuels led to deviations from the original operating procedure since most of the equipment is not optimized for biofuels. In the Annex one can see a summary of the methods tested in this task, the details concerning the laboratory equipment used and the experimental parameters (like sample amounts, reference materials, combustion temperatures) and some statistic parameters (like detection limits).



Table 2: Chlorine determination, tested methods in Task III.1

Bomb Combustion

+ IC detection (or other

det. method)

Pressure Digestion

+ ICP-OES

Eschka Method

Water soluble Cl

+ IC Detection

Tube Combustion

+ Coulometric

/ IC- Detection

Irradiation (INAA)

AOX Analyser

Wickbold + IC

detection XRF

dk Elsam dk Dk ARCS ECN ARCS IFE PAN1

VTT CIEMAT Elsam ECN Elsam ECN Ofi

ofi ofi ECN IFE SP

CPERI 1 Analysis performed by PanAnalytical, not funded within this project

For the determination of nitrogen only two methods were taken into account and were investigated in this project: automated analyzers and the Kjeldahl method. The automated methods replaced formerly developed chemical methods more or less completely (for which International Standards still exist). Automated or instrumental methods, which are in widespread and regular use for the analysis of nitrogen in all organic samples, are standardized for solid mineral fuels in an ISO Technical Specification, which was developed recently4. However, some of the automated analyzers show limitations when low concentrations of nitrogen have to be determined precisely. Therefore, the Kjeldahl method was also investigated. The methods available in this project are summarized in Table 4, details are compiled in the Annex.

Table 3: Sulphur determination, tested methods in Task III.1

Bomb Combustion + IC detection

Microwave / Pressure

Digestion + ICP-OES

AA – Eltra, Leco, Erba,

Fisons

Tube Combustion + IC- Detection

Eschka / Turbidimetric Determination

XRF

Dk CIEMAT VTT ARCS Dk PAN1 ECN Elsam CIEMAT ofi ARCS Elsam ARCS Elsam

ofi ECN SP CPERI ECN

ARCS CPERI

AA...Automatic Analyzer 1 Analysis performed by PanAnalytical, not funded within this project

4 for further details see ISO/TS 12902:2001



Table 4: Nitrogen determination, tested methods in Task III.1

Automated analyzers – Leco, Vario EL, Fisons, Erba Kjeldahl method

Dk ARCS VTT ECN

Elsam CPERI ARCS ECN SP

4.2 Sequence of experiments

After distribution of the samples to the participating laboratories, analysis of chlorine, sulphur and nitrogen was performed using all methods available. The intermediate test results provided by the laboratories were subjected to a statistical evaluation in accordance with the provisions of ISO 5725. The results were compiled in Excel tables and were distributed to the laboratories in December 2002. In January 2003, the results were discussed in a work group meeting in Milano. The discussion of the general suitability of the investigated methods led to the identification of testing methods not suitable for biofuels and so these were not considered for further investigations. This decision based on the experimental results gained, as well as on the experience of the participating laboratories.

These discarded methods are:

• combustion according to Schöninger [e.g. 5]

• combustion according to Wickbold [e.g. 6, 7]

• automated AOX analyzers

• acid digestion in open vessels

• high temperature combustion in tube furnace [e.g. 8, 9, 10, 11]

They were considered to be not suitable for solid biofuels because they were either not sensitive enough or difficult to handle or the reproducibility and repeatability was poor.

The results obtained by the other investigated methods also showed a significant demand for improvement. Deviating results for some samples suggested the replication of experiments. This was in particular necessary for those laboratories being not so experienced with the analysis of solid biofuels, or for laboratories just implementing new methods. As a general conclusion it is highly recommended for laboratories to participate in round robins to evaluate their performance of the analytical procedures, when analyzing biofuels.

Additional results were collected and results, that could be assigned to systematic errors were replaced with replicate results. Deviating results that could not clearly be assigned to systematic errors or to experimental mistakes were not removed from the compiled results in order to present a picture representing the analytical practise. This process continued until December 2004. A summary of the results for the samples investigated including the calculated reproducibility and repeatability and a few examples of results for individual samples are compiled in chapters 4.3, 4.4 and 4.5. All experimental results are compiled in the Annex.



Till the midterm meeting in September 2003 the best methods for the determination of chlorine, sulphur and nitrogen were selected. With exception of the Kjeldahl method these selected methods were also chosen for standardization. The selected methods are described in detail in the “Best practice guidelines for the determination of sulphur, chlorine, nitrogen, major and minor elements” [13] and in the corresponding draft standards:

• Chlorine:

∗ combustion in an oxygen bomb and determination of chloride by IC [2]

∗ water soluble chlorine [2]

• Sulphur:

∗ combustion in an oxygen bomb and determination of sulphate by IC [2]

∗ acid digestion in closed vessels and determination by ICP [2, 14]

• Nitrogen

∗ automated nitrogen analyzers [1]

∗ Kjeldahl method. For the Kjeldahl method, suitable methods are existing and are standardized, see [15, 16, 17, 18]

To identify the most critical steps in the analytical process, a sensitivity analysis was performed for selected parameters. The following aspect were investigated between autumn 2003 and autumn 2004:

• influence of moisture content on the nitrogen determination

• influence of the particle size on the chlorine, sulphur and nitrogen determination (see [13])

• influence of different receiving solutions in the oxygen bomb for the chlorine and sulphur determination

• influence different oxygen pressures in the bomb for the chlorine and sulphur determination.

• influence of the time elapsed for pressure release after digestion in the bomb for the chlorine and sulphur determination.

• sulphur remaining in combustion residues

Selected results are summarized in chapter 4.7.

To separate systematic deviations of the independent analytic steps (digestion and quantification) for the sulphur and chlorine determination, both steps were investigated separately using this approach:

• a digest solution (bomb digestion according to [2]) was prepared by one laboratory and was analyzed for chloride and sulphate concentration by the other laboratories

• all laboratories carried out a bomb combustion according to [2] and the chloride and sulphate concentration was measured by one laboratory



The results of these investigations are summarized in the Annex. Finally, the developed methods were tested using an additional sample (hemp), these results are also summarized in chapter 4.7.

4.3 Results of the method evaluation for chlorine analysis

In Table 5 the results of the method comparison are compiled for all samples. In Figure 5 and Figure 6 results for the analysis of the samples orujillo and cynara are shown. Most results represent mean values for multiple determinations (3-5 experiments each), the error range is indicated by the standard deviation for the replicate measurements.

Table 5: Results for the chlorine analysis using different methods: mean chlorine contents, reproducibility and repeatability

Chlorine content Reproducibility Repeatability

mean [wt %, d.b.]

R abs [%]

R rel [%]

r abs [%]

r rel [%]

wood without bark 0,003 0,006 174 0,003 93 woodchips 0,005 0,010 185 0,006 103 bark 0,010 0,010 99 0,005 53 hardwood with glue 0,015 0,012 83 0,007 51 straw 0,11 0,030 26 0,016 14 orujillo 0,20 0,043 21 0,013 7 rapestraw 0,28 0,065 23 0,027 10 cynara 1,59 0,44 27 0,18 11

chlorine content orujilo

0,00

0,05

0,10

0,15

0,20

0,25

(AOX) A

RCS

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) IFE

(Bom

b + IC

) ofi

bomb dk

- IC of

i

(Bom

b + IC

) SP

(Bom

b + tit

ration)

VTT

(Bom

b + IC

) CPERI

(Esc

hka)

CIEMAT

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(tube

furna

ce) A

RCS

(tube

furna

ce) E

CN

(wate

r soluble

Cl) d

k

(wate

r soluble

Cl) o

fi

(wate

r soluble

Cl) E

CN

(wate

r soluble

Cl) E

lsam

water solu

ble dk/ofi

INAA E

CN

PAN Analyt

. XRF

%

Figure 5: Example for chlorine analysis: content of the sample orujillo, analysed by different methods in [wt%,

d.b.]



chlorine content cynara

0,0

0,5

1,0

1,5

2,0

2,5

(AOX) A

RCS

(Bom

b + Titra

tion)

dk

(Bom

b + Titra

tion)

dk

(Bom

b + Titra

tion)

dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) IFE

(Bom

b + IC

) ofi

bomb dk

- IC of

i

(Bom

b + IC

) SP

(Bom

b + tit

ration)

VTT

(Bom

b + IC

) CPERI

(Esc

hka)

CIEMAT

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(tube

furna

ce) A

RCS

(tube

furna

ce) E

CN

(wate

r soluble

Cl) d

k

(wate

r soluble

Cl) o

fi

(wate

r soluble

Cl) E

CN

(wate

r soluble

Cl) E

lsam

water solu

ble dk/ofi

%

Figure 6: Example for chlorine analysis: content of the sample cynara, analysed by different methods in [wt%,

d.b.]

4.4 Results of the method evaluation for sulphur analysis

In Table 6 the results of the method comparison are compiled for all samples. In Figure 7 and Figure 8 results for the analysis of the samples bark and straw are shown. Most results represent mean values for multiple determinations (3-5 experiments each), the error range is indicated by the standard deviation for the replicate measurements.

Table 6: Results for the sulphur analysis using different methods: mean contents, reproducibility and repeatability

Sulphur content Reproducibility Repeatability

mean [wt %, d.b.]

R abs [%]

R rel [%]

r abs [%]

r rel [%]

wood without bark 0,006 0,008 143 0,004 60 woodchips 0,009 0,004 50 0,005 54 hardwood with glue 0,017 0,007 42 0,004 24 bark 0,080 0,041 51 0,014 17 straw 0,11 0,029 26 0,020 16 orujillo 0,13 0,036 27 0,013 9 cynara 0,20 0,065 33 0,026 13 rapestraw 0,21 0,043 21 0,019 9



sulphur content bark

0,00

0,02

0,04

0,06

0,08

0,10

0,12

0,14

0,16

0,18

0,20

(Bom

b + tu

rbid.)

dk

(Bom

b + tu

rbid.) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) ofi

(Bom

b + IC

P) IFE

(Erba)

ARCS

(Erba)

CIEMAT

(HPA +

ICP) E

lsam

(HPA +I

CP) ARCS

(HPA +I

CP) CIE

MAT

(HPA +I

CP) ECN

(Leco) E

lsam

(Leco) S

P

PAN Analyt

. XRF

[%]

Figure 7: Example for sulphur analysis: content of the sample bark, analysed by different methods in [wt%,

d.b.]

sulphur content straw

0,000,020,040,060,080,100,120,140,160,18

(Bom

b + tu

rbid.)

dk

(Bom

b + tu

rbid.) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) ofi

(Bom

b + IC

P) IFE

(Erba)

ARCS

(Erba)

CIEMAT

(Erba)

ECN

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(HPA +I

CP) ARCS

(HPA +I

CP) CIE

MAT

[%]

Figure 8: Example for sulphur analysis: content of the sample straw, analysed by different methods in [wt%,

d.b.]

4.5 Results of the method evaluation for nitrogen analysis

In Table 7 the results of the method comparison are compiled for all samples. In Figure 9 and Figure 10 results for the analysis of the samples straw and woodchips are shown. Most results represent mean values for multiple determinations (3-5 experiments each), the error range is indicated by the standard deviation for the replicate measurements.



Table 7: Results for the nitrogen analysis using different methods: mean contents, reproducibility and repeatability

Nitrogen content Reproducibility Repeatability

mean [wt %, d.b.]

R abs [%]

R rel [%]

r abs [%]

r rel [%]

wood without bark 0,060 0,060 99,2 0,027 45,1 woodchips 0,106 0,087 82,00 0,079 74,7 hardwood with glue 0,342 0,144 42,2 0,036 10,5 rapestraw 0,406 0,159 39,2 0,090 22,1 bark 0,669 0,221 33,1 0,076 11,4 straw 0,713 0,191 26,7 0,078 10,9 cynara 1,006 0,241 24,0 0,107 10,6 orujillo 1,314 0,333 25,3 0,090 6,8

nitrogen content straw

0,00,10,20,30,40,50,60,70,80,91,0

(Erba

) ARCS

(Erba

) ECN

(Erba

) Elsa

m

(Fisons)

CIE

MAT

(Kjel

dahl)

ARCS

(Kjel

dahl)

ECN

(Leco

) SP

(Leco

) VTT

(Vari

o EL) d

k

(Kjel

dahl

I) CPERI

(Kjel

dahl

II) C

PERI

(Kjel

dahl)

IFE

[%, d

.b.]

Figure 9: Example for nitrogen analysis: content of the sample straw, analysed by different methods in [wt%,

d.b.]



nitrogen content woodchips

0,00

0,05

0,10

0,15

0,20

0,25

0,30

(Erba

) ARCS

(Erba

) ECN

(Erba

) Elsa

m

(Fisons)

CIE

MAT

(Kjel

dahl)

ARCS

(Kjel

dahl)

ECN

(Leco

) SP

(Leco

) VTT

(Vari

o EL) d

k

(Kjel

dahl

I) CPERI

(Kjel

dahl

II) C

PERI

(Kjel

dahl)

IFE

[%, d

.b.]

Figure 10: Example for nitrogen analysis: content of the sample woodchips, analysed by different methods in [wt%, d.b.]

4.6 Performance of the method for total sulphur and total chlorine

When the draft for the standard procedures were completed, a new sample (hemp) was prepared and was analysed by the laboratories strictly applying the developed standard procedure for the (probably) most common analytical combination consisting of a bomb combustion and determination by IC. The results are compiled in Figures 11 and 13.

%w/w Cl, dry basis

0,00

0,02

0,04

0,06

0,08

0,10

dk ECN Elsam IFE ofi SP Ciemat VTT INAAECN

Figure 11: Performance of the standardized bomb method for the determination of total chlorine for a hemp

sample (bomb digestion and determination by IC and titration) and result for the determination using INAA

For the evaluation of the bomb digestion step, the sample was combusted in the participating laboratories applying the procedure described in the draft standard. The digest solutions were sent to one laboratory (ofi) where they were analyzed for the concentration of chloride and sulphate using an IC. In this way, the measurement error of the two steps of the overall



method (digestion and determination) were separated. The results are summarized in Figures 12 and 14. The results clearly show, that the combustion in the oxygen bomb is a highly reproducible method for both, sulphur and chlorine analysis. To improve the results for the complete method, care has to be taken to analyse the concentration of chloride and sulphate in the digest solutions properly.

%w/w Cl, dry basis

0,00

0,02

0,04

0,06

0,08

0,10

ofi Force ECN Elsam Ciemat VTT

Figure 12: Performance of the standardized bomb method for the determination of total chlorine for a hemp sample, bomb digestion by the laboratories listed in the figure, determination by IC in one laboratory (ofi)

%w/w S, dry basis

0,00

0,01

0,02

0,03

0,04

0,05

dk ECN Elsam IFE ofi SP Ciemat VTT SP Leco

Figure 13: Performance of the standardized bomb method for the determination of total chlorine for a hemp sample (bomb digestion and determination by IC and titration) and result for the determination using INAA



%w/w S, dry basis

0,00

0,01

0,02

0,03

0,04

0,05

ofi dk ECN Elsam Ciemat VTT

Figure 14: Performance of the standardized bomb method for the determination of total chlorine for a hemp sample, bomb digestion by the laboratories listed in the figure, determination by IC in one laboratory (ofi)

4.7 Results of the sensitivity analysis for the bomb method

The relatively poor reproducibility and repeatability of the sulphur and chlorine determination suggested a detailed investigation of the parameters influencing the results significantly (sensitivity analysis). It was already decided, that the best available method selected for standardization will consist of a combustion in an oxygen bomb and determination by ion chromatography (IC). Thus, mainly parameters for this method were investigated.

The determination of chloride and sulphate is a fairly well established method [19] and there are in general no problematic matrix effects observed in the receiving solution obtained from the combustion in the bomb that may influence the result of the quantification of these two anions (provided that a proper separation column is used). Thus, parameters influencing the determination by IC were not investigated.

4.7.1 Oxygen pressure

The oxygen pressure in the combustion bomb was varied in a range between 5 and 35 bars. The lower pressure is the minimum pressure to obtain a complete combustion, the higher pressure is the limit for the used equipment (stainless steel bomb). Statistically significant differences between the obtained results for different oxygen pressures in the bomb combustion could not be found. However, the experimental results indicate a tendency that the lowest oxygen pressures result in slightly higher chlorine and sulphur contents (see Figure 15 and 16).

This may be due to the observed connection between the oxygen pressure and the degree of melting of the combustion residues. For the experiments at 10 bar oxygen pressure, the ash residue was only melted a little (sintered). With increasing oxygen pressure, the melting of the combustion residues increased. This indicates higher combustion temperatures for increasing oxygen pressures which matches the expectations. A higher combustion temperature indicates a more complete combustion but on the other hand this may lead to chlorine and sulphur combined in the sintered or even melted inorganic combustion residues. This speculation is supported by the fact, that the combustion residues contain a significant amount of sulphur



(determined by a digestion according to [20]). The combustion residues of rape straw contained 0.33% sulphur, the combustion residues of cynara contained 0.72% sulphur. The combustion of biofuels with a high ash content and a low ash melting temperature may lead to an underestimation of sulphur and possibly also of chlorine. It is interesting to note, that for samples with a high ash content, the water soluble chlorine is higher than the “total” chlorine determined by the bomb combustion. This may be explained by the fact, that chlorine is highly mobile in biofuels and can be quantitatively eluted with water, see also [13] but it cannot be completely liberated in the combustion processes.

% w/w S, dry basis

0,160,170,180,190,200,210,22

35 30 25 20 15 10oxygen pressure [bar]

Laboratory 1 Laboratory 2

Figure 15: Dependence of the sulphur content for the sample cynara on the oxygen pressure in the combustion

bomb in [wt%, d.b.]

% w/w Cl, dry basis

1,00

1,20

1,40

1,60

1,80

2,00

35 30 25 20 15 10oxygen pressure [bar]

Laboratory 1 Laboratory 2

Figure 16: Dependence of the chlorine content for the sample cynara on the oxygen pressure in the combustion

bomb in [wt%, d.b.]

4.7.2 Influence of the pressure release time

During the discussion of the experimental procedure for the bomb method, experts speculated, that the pressure release may be a critical point for the correct determination of the sulphur and chlorine content since chlorine and sulphur may be lost when hydrogenchloride and sulphuroxide are not quantitatively absorbed in the water inside the bomb. However, it was found, that there is no significant influence on the analysis result with respect to the time elapsed until the pressure release is done, see Figure 17.



% w/w S, dry basis

0,10

0,110,12

0,130,14

0,15

A B C D E

Figure 17: Influence of the pressure release on the sulphur content, sample orujilo (Explanation: A: pressure release at once, B: pressure release 1 min, C: pressure release 2 min, D: pressure release 25 min, E: valve not rinsed)

It is important to note, that the pressure release itself is a very critical point since the pressure release may lead to a loss of a fraction of the receiving solution during the venting procedure. These potential losses strongly depend on the specific design of the combustion bomb and especially on the design of the venting valve. An indication for losses is small droplets close to the valve after the pressure release. Depending on the bomb design it may be necessary to release the pressure very carefully and to flush the valve of the bomb with water.

4.7.3 Amount of water in the bomb and alternative receiving solutions

Usually, 1 ml of deionised water is added to the bomb as receiving solution to absorb the acidic gases formed in the combustion. European laboratories used different procedures in the past, ranging from the use of no water to the use of up to 10 ml alkaline receiving solution. The variation of different amounts of water and also the use of alkaline solutions (0.1 n KOH and 0.1 n KOH/H2O2) or the use of the mobile solvent for the IC (NaHCO3) as receiving solution did not show a significant influence on the sulphur or chlorine determination. Therefore, deviations from the standard procedure are possible but should be validated if used, see [2]:

• Water may be omitted (practise showed, that the water formed in the combustion is usually sufficient to solve the acidic gases formed) or more water (up to 5 ml) may be used.

• If the content of chlorine or sulphur exceeds 2 % (m/m), alkaline solutions may be used to neutralize the acidic compounds produced.

• If the content of chlorine or sulphur exceeds 2 % (m/m), a part of the acidic components may not be solved in the receiving solution and may be lost during venting. In this case, the combustion gas should be let through a gas washing bottle with a disk to assure that all acidic gas components are dissolved. The solution from this gas washing bottle may be combined with the bomb washings or may be analysed separately. The fraction recovered via this washing bottle shall in no case exceed 10 % of the total determined chlorine and sulphur.

• When ion chromatography is used for determination, the absorption solution may be the mobile phase e.g. a carbonate/bicarbonate solution.



• If the chlorine content of the sample is very low, the cotton thread usually used for ignition may contribute significantly to the measured chlorine content. This can be avoided by using highly pure combustible crucibles without cotton threats.

• In all cases, the calibration of the method and the blank tests have to be performed with the same amount and kind of receiving solution.

4.7.4 Influence of sulphur and chlorine combined in the combustion residues

The combustion residues of two (combustion) methods were investigated for the content of sulphur: the high temperature combustion in a tube furnace (1250°C) and the combustion in an oxygen bomb. While the combustion residues from the tube furnace can be collected and subsequently be analyzed for their chemical composition fairly easy, the collection of the combustion residues in the bomb is difficult since these residues are washed and a part of the residues is lost (e.g. spilled out of the crucible or lost during a vigorous combustion). Furthermore, the combustion residues from the combustion of 1 gram sample are very small especially for samples with low ash contents. Therefore the results obtained from these analysis are to be seen as a first qualitative indication.

The combustion residues obtained from the bomb combustion of the sample rape straw contained 0.33 % sulphur. For this sample with an ash content of 5.3 % (determined at 550 °C) this indicates, that 0.02 % sulphur of the biofuel is combined in the combustion residues. Added to the sulphur content determined by the bomb combustion method of 0.2 % (average of the analysis results for the bomb method) it results in a total sulphur content of 0.22 % for the rape straw sample which is exactly the same amount as determined with digestion methods (average of the analysis results for the acid digestion methods).

The combustion residues obtained in the bomb combustion of the sample cynara contained 0.72 % sulphur. For this sample with an ash content of 10 % (determined at 550°C) this indicates, that 0.072 % sulphur of the biofuel is combined in the combustion residues. Added to the sulphur content determined by the bomb combustion method of 0.18 % (average of the analysis results for the bomb method) it results in a total sulphur content of 0.25 % for the cynara sample which exceeds the amount determined with digestion methods (average of the analysis results for the acid digestion methods) of 0.22 %.

The residues of the tube combustion contained different amounts of sulphur and chlorine that could not be correlated to the total content of these elements. Since the results of the tube combustion did not match with the results of the other methods regardless of taking the sulphur and chlorine content in the residues into account or not, the analysis values are not discussed. However, it can be concluded, that a fraction of sulphur and chlorine is always combined in combustion residues when a method is used that is based on combustion. It can be speculated, that the amount of combined sulphur and chlorine increases with an increasing ash content.

4.7.5 Influence of analysis sample particle size

The preparation of the analysis sample is very critical with respect to the particle size. The particle size influences are summarized in “Solid Biofuels – Best practice guidelines for the determination of sulphur, chlorine, nitrogen, major and minor elements, chapter 3.3” [13].



4.8 Chlorine determination using the Eschka method

The Eschka-method is a widespread method for the determination of chlorine in coal samples. Two widely recognized standard procedures are available, ISO 587 [21] and ASTM D 2361 [22]. Within the BioNorm project, the methods were evaluated by two laboratories. The results from one laboratory were in line with the results of other methods while one laboratory found systematically too low values. Thus, the details of the methods were discussed and it was found, that the laboratory obtaining correct values used a modified Eschka procedure. The differences are outlined in the Annex. From experiments with modified Eschka methods, following conclusions were drawn:

• An elevated combustion temperature of 775 °C instead of 675 °C as defined in the standards [21, 22] has no significant influence on the result.

• A higher amount of Eschka mixture and especially a higher amount used as cover improves the results (giving higher results of chlorine content).

• The correct treatment of the incinerated mixture, according to ISO 587:1997 Cl. 7.2.3 and ASTM D 2361-95 (adding 20 ml conc. nitric acid) improves the results considerably.

• It should however be mentioned that the treatment of the incinerated mixture causes complications:

∗ after the addition of the 20 ml conc. nitric acid the pH-value of the solutions is in the area of –0.9 to –0.2. This means that the solution has to be neutralized (with 4 M NaOH);

∗ due to the very high salt content of the solutions problems with the chloride determination (potentiometric titration using automatic titration equipment) were experienced. The problem was that the equipment sometimes missed the inflection point for the blanks and for samples with low chlorine content. Therefore, the criterion for the “true” inflection point has to be defined carefully.

Conclusion and recommendation: The Eschka method is a suitable method for the determination of chlorine (with deviations described above) in biofuels although it is not included in the standard procedures in the draft standards for the determination of chlorine and sulphur in solid biofuels [2].

4.9 Other methods for chlorine and sulphur (not investigated in detail)

Two methods were used in this project but were not considered for standardisation. Therefore, they were not investigated in detail but experiences are briefly summarized:

4.9.1 Neutron activation analysis (reference method)

The neutron activation analysis was an appropriate method for the determination of chlorine in all samples. Since it is a method that requires extremely expensive equipment it is therefore not designed for routine analysis for solid biofuels. This method may be of interest for further scientific investigations, as a method to obtain absolute values and to improve and validate analytical methods. For this purpose, the neutron activation analysis showed to be very helpful.



4.9.2 X-ray fluorescence analysis (XRF)

As fast and non-destructive analysis, X-ray fluorescence analysis is used successfully in other fields. Originally, it was not intended to include this method in the BioNorm project. However, a company (PanAnalytical) provided measurements that were included in the analytical results. These measurements showed, that XRF might be an accurate and fast method for the analysis of solid biofuels. XRF may also be used for the chemical characterization of major (ash forming) and some minor elements. However, at the moment this method may only be used when it is validated in a laboratory using certified reference materials.

5 Conclusions and recommendations

5.1 Chlorine and sulphur determination

The method evaluation for total sulphur and total chlorine analysis led to the following conclusions and recommendations:

• The currently best method is the combustion in an oxygen bomb and the determination of sulphate and chloride in the receiving solution applying the procedures currently standardized on European level [1]. However, biofuels with high ash contents tend to combine sulphur and chlorine in the combustion residues resulting in low values.

• For chlorine, the method for the water soluble chlorine leads to similar results as the bomb combustion method, at least for all untreated biofuels investigated in this project (an exception was the sample hardwood with glue were not all chlorine was water soluble but this sample contained non-natural additives). For samples with high ash content higher values were obtained as with the bomb combustion method.

• The required repeatability and reproducibility can only be obtained for sulphur and chlorine determination when strictly kept to the standard procedures.

• Biofuels with low contents of sulphur and chlorine are difficult to analyze, the reproducibility is still not satisfying. More research and method improvement is required.

5.2 Nitrogen determination

The comparison of the results of automated analyzers (different and same brands and types) and of the Kjeldahl method led to following conclusions:

• No systematic deviations were found with respect to a certain type or brand of automated analyzers but differences were found between different laboratories (persons operating the systems), thus the operation and especially the calibration is critical for the results obtained.

• Automated analyzers are suitable for the determination of nitrogen in biofuels and are consequently currently standardized on European level [1]. Since it was recognized, that different designs of automated analyzers led to comparable results, no specific designs of systems are presented in the draft standard because there is a range of



components and configurations available, which can be used to carry out the test method satisfactorily.

• The apparatus shall, however, meet the functional requirements defined in [1].

During the experimental work in this project the following mistakes appeared that should especially be considered in laboratory practise:

• The samples shall always be within the calibration range.

• When low nitrogen concentrations are to be analyzed, an increased sample amount may improve the result.

• No influence of moisture content on the nitrogen concentration could be found (at least for the investigated instruments), thus, it is not necessary to used dried samples which are difficult to handle since they are in general very hygroscopic.



6 References

1 prCEN/TS 15104: Solid biofuels – Determination of total content of carbon, hydrogen and nitrogen – Instrumental methods. European Committee for standardization, Brussels, Belgium; 2004.

2 CEN/TC 335 WG 5 N 124 draft: Solid biofuels — Determination of total content of sulphur and chlorine. European Committee for Standardization, Brussels, Belgium; 2004.

3 prCEN/TS 15105: Solid biofuels – Methods for determination of the water soluble content of chloride, sodium and potassium. European Committee for Standardization, Brussels, Belgium; 2004.

4 Puttkamer T, Baernthaler G: Solid Biofuels - Determination of major and minor elements - Best practice guidelines for the preparation of analysis samples. Report on Deliverable III.2 D3 of the EU funded BIONORM project. http://www.energetik-leipzig.de/BioNorm/Standardisation.htm ; 2004.

5 EN ISO 1158:1998: Plastics - Vinyl chloride homopolymers and copolymers - Determination of chlorine content.

6 ISO 4260:1987: Petroleum products and hydrocarbons - Determination of sulfur content - Wickbold combustion method.

7 DIN 53474:1998: Prüfung von Kunststoffen, Kautschuk und Elastomeren - Bestimmung des Chlorgehaltes (Aufschluß nach Wickbold).

8 ISO 352:1981: Solid mineral fuels – Determination of chlorine – High temperature combustion method.

9 BS 1016:Part 8:1977: Methods for Analysis and testing of coal and coke. Part 8. Chlorine in coal and coke. High temperature method.

10 DIN 51727:2001: Prüfung fester Brennstoffe. Bestimmung des Chlorgehaltes.

11 ISO 351:1996: Solid mineral fuels - Determination of total sulphur - High temperature combustion method.

12 DIN 51724-1: Testing of solid fuels – Determination of sulphur content, Part 1: Total sulphur (1250°C)

13 Baernthaler G, Englisch M, Obernberger I: Solid Biofuels – Best practice guidelines for the determination of sulphur, chlorine, nitrogen, major and minor elements. Report on Deliverable III.1 D3 of the EU funded BIONORM project. http://www.energetik-leipzig.de/BioNorm/Standardisation.htm ; 2005.

14 CEN/TC 335 WG 5 N190 9th draft: Solid biofuels – Determination of major elements. European Committee for Standardization, Brussels, Belgium, 2004.



15 ON-EN 13342: Characterisation of sludges – Determination of Kjeldahl nitrogen. Austrian Standards Institute, Vienna; 2000.

16 ISO 333: Coal – Determination of nitrogen – Semi-micro Kjeldahl method. International Organisation for Standardization, Geneva, Switzerland; 1996.

17 DIN 51722-1: Testing of solid fuels - determination of nitrogen content - semi-micro Kjeldahl method. Deutsches Institut für Normung, Berlin, Germany; 1990.

18 ISO 11261: Soil quality - Determination of total nitrogen - Modified Kjeldahl method. International Organisation for Standardization, Geneva, Switzerland; 1997.

19 EN ISO 10304-1, Water quality – Determination of dissolved fluoride, chloride, nitrite, orthophosphate, bromide, nitrate and sulphate ions, using liquid chromatography of ions – Part 1: Method for water with low contamination.

20 CEN/TC 335 WG 5 N190 9th draft.: Solid biofuels – Determination of major elements. European Committee for Standardization, Brussels, Belgium, 2004.

21 ISO 587:1997: Solid mineral fuels – Determination of chlorine using Eschka mixture.

22 ASTM D 2361 – 95. Eschka method. Standard Test Method for Chlorine in Coal.

7 Annex

The Annex contains additional information about the methods and results used in Task III.1. Some experimental details of used methods are summarized in:

• Table 1: Chlorine determination

• Table 2: Sulphur determination

• Table 3: Nitrogen determination

Results of the determination of chlorine, sulphur and nitrogen in the 9 samples used within this task are compiled in Figures 1-26.

The chlorine determination using an Eschka method was investigated in detail by two laboratories, CIEMAT and FORCE. Since the Eschka method is not included in the standard procedure, it was not in the focus of the current research. Nevertheless it was found in task III.1 that it is a fairly cheap method suitable for solid biofuels. Important for obtaining correct analysis results are some experimental details that are described and discussed in this Annex by Susan Westborg, FORCE.



Table A-1: Chlorine determination, experimental details of used methods (to be continued next page)

C hlo rineac c o rd ing to o r

b as ed o n A p p aratus p rinc ip le s p ec ific d etailsrec eiving s o lu tio n

A p p aratus d e term inatio n

p rinc ip le d ete rm inatio n

gas flo w o r gas p res s u re (if ap p lic ab le)

c o m -b us tio n T em p .

referenc e m ateria l fo r c alib ratio n

refe renc e m ateria l fo r d eterm ina tio n

S am p le am o un t

s am p le p rep aratio n

s am p le (w ater c .)

s am p le p artic le s ize

typ ic al b lank values d etec tio n lim it

m l/m in o r b ar ° C m g p o w d er, p elle t, . . . ar, an , w f m mS P

b o m b + ICIS O 1928 , S S

187177 P arr 1261IS O 1928, w ith

d o d ec ane ad d ed

b o m b is rin s ed 3 tim es w ith 20 m l

H 20 , filled to 100m l 1 m l H 2O IC D io nex Io n

c h ro m ato grap hy 30 b ar ?C l- C ertp ur (M erc )

1g/l 600 p e llet < 1

0,3 g d o d ec ane c o m b us ted and

d ilu ted to 100 m l gives ap p ro x 0 ,2

p p m C l 0 ,001dk -T E K N IK

B o m b + P o tentio m etric T itratio n

D an is h " R ec o m m end ed m etho d no . 9"

P arr 1261 , b o m b : P arr 1108 (s ta in les s

s teel, 350m l), s tain les s s tee l

c ruc ib les

b o m b c o m b us tio n and p o ten tio m etric

titratio nN iC r-ign itio n w ire;

fill to 100m l

no w ater ad d ed to

b o m b

R ad io m eter T IM 900 ; A gA g P 4040

and referenc e elec tro d e

R ad io m eter K 601

T itra tio n w ith A gN O 3, end p o in t is the vo ltage w ith

the g reates t c hange in the p o tential d u ring c o ns tant

ad d itio n o f A gN O 3 ad d itio n 30 -

n-d o d ec ane us ed in every

d eterm inatio n ; p res s ure releas e

> 2m in

d eterm inatio n: V H I Q C D W B - d rink ing

w ater R eferenc e , C h lo rid e s tand ard

s o lu tio ns 1000-1100 p e llet an < 1

0 ,5g B enc o ic ac id and 0 ,12-0 ,18g n-

d o d ec ane : < 0 ,5m g C l/l in

s o lu tio n 0,01%

E s c hka + P o tentio m etric T itratio n IS O 587 :1997

m uffle fu rn ic e, g lazed p o rc elain

25m l, M erc k E s c hka m ix .

c o m b us tio n o f s am p le w ith E s c hka m ix tu re fo r 2 ho urs

o r m o re

M etho d no t us ed fo r b io fue ls , lo w

va lues fo r e .g .s traw -

p o ten tio m etric titra tio n lim it 0 ,5

m gC l/l



ad d itio n o f A gN O 3 ad d itio n - 675



s o lu tio ns 1000 p o w d er an < 1

S am e p ro c ed ure as fo r s am p les -

b u t w itho u t s am p le 0,02%

w ater s o lub le C l + P o tentio m etric T itratio n

D an is h " R ec o m m end ed m etho d no . 10" T eflo n b o ttles

ex trac tio n w ith 50 m l p ure H 2O 120° C fo r 1 ho ur -

p o ten tio m etric titra tio n lim it 0 ,5

m gC l/l



ad d itio n o f A gN O 3 ad d itio n - -



s o lu tio ns 1000 p o w d er an < 1

S am e p ro c ed ure as fo r s am p les -

b u t w itho u t s am p le 0,01%

C IE M A T

E s c hka + V o lhard T itratio n

A S T M D 2361-66 and U N E 32024 furnac e

as hing at 775° C w ith E s c hka m ix ing ,

d is s o lved w ith n itric ac id

heating rate < 15 ºC /m in. 1 ho ur a t

775 ºC . 20 m l c o nc en trated nitric

ac id -invers e V o lhard

titratio n - 775

s p ruc e need les B C R C R M 101 , and

o thers 500-4000 p o w d er w f < 0,5 0 ,0001C P E R I

b o m b + IC c alo rim etric b o m b 5m l w aterIC , M etro hm 761 w ith s up p res s io n 30 ? 100-1000

A R C S

A O X - A nalyzer - L H G -E S C 2000

B urning o f the s am p le and c o u lo m etric

d etec tio n 0 ,0007

T ub e F urnac e (H 2O 2) + IC D IN 51727

S trö h lein tub e furnac e

c o m b us tio n w ith o xygen 1% H 2O 2 D io nex D X -500

Io n c h ro m ato grap hy ~ 2L /m in 1250 200 - 1000

all, p referab ly an < 0,5 0 ,01

E C N

w ater s o lub le C l + IC d an is h gu id eline

B ergho f b o m b 250 m l

ex trac tio n w ith 50 m l p ure H 2O ,d iluted to

100 m l 120 ° C fo r 1 ho ur IC

D io nex D X 600 w ith E G 40, c o lum n A S

17 , 2m mIo n

c h ro m ato grap hy - - 1000 p o w d er < 0,5 c a. 0 .05 m g/l 0 ,05 m g/l

T ub e F urnac e (1200° ) + c o ulo m etric

c o m b us tio n gas es are d ried b y H 2S O 4

c o nc . O 2/A r flo wac etic ac id

s o lutio nE C S -2000-C o u lo m eter

generatio n o f s ilver io ns , d etec tio n b y

C o ulo m eter 1200 1-50 < 0,2 o r < 0,1 0 ,01

B o m b (P A R R ) + IC

c a lo rim eter P arr 1241 , b o m b P arr

1108 -

rins e the b o m b w ith O 2; inc o m p lete

c o m b us tio n: m o is t lo o s e s am p le w ith 200 m l H 20 o r ad d

S iO 2 o r b enzo ic ac id 25 ? 1000-1200 p e llet < 0,2



Chlorineaccording to or

based on Apparatus principle specific detailsreceiving solution

Apparatus determination

principle determination

gas flow or gas pressure (if applicable)

com-bustion Temp.

reference material for calibration

reference material for determination

Sample amount

sample preparation

sample (water c.)

sample particle size

typical blank values

ECN

Irradiation + Gammaspectrometric

Nuclide parameters: Radionuclide Cl-38,

T1/2=38min Ey=1642 keV

neutron activation multi-element analysis

via short lived nuclides

samples are weighted in 0,8 ml polyethylene vessels. Samples, standards and blanks are irridated and

analyzed gammaspectrometrically 10

min on a Ge detector -

samples are irridated in a thermal neutron

flux of 3x10e11 cm-2 s-1, after a decay

time of 30 min samples are analyzed gammaspectrometric

ally - - NIST CRM 3182 - 500 0,0021mgVTT

Bomb + Potentiometric Titration ASTM 2361 Bomb: Parr 1108

bomb combustion and potentiometric

titration

admit oxygen slowly, 15 min in cooling bath after

combustion, pressure release > 2min, rinse with hot water

5ml (10 g (NH4)2CO3 in 100 ml H2O)

automatic titrator or ph-meter and

chlorine electrode

0,025n AgNO3, silver-silver chloride

electrode 30 ? 1000 pellet an <0,5Elsam

Bomb (Parr)+ IC

Parr 1261 Isoperibol bomb with 1108CL Oxygen combustion

chlorine-resistant bomb -

after combustion gas is let through a tube into water (100 or 200 ml volumetric

flask), content of bomb and capsule is transferred to

flask 1 ml MQ H2O

DIONEX DX500, column AS14, eluent 3,5 mM Na2Co3, 1

mM NaHCO3, injection through filtercaps, 20mm IC 30 bar Oxygen ?

Standards: Merck NaCl

Wheat straw and Spruce wood,

German RR 2000 800-1000 pellet ar < 1 < DL

Berhof or Microwave + ICP-OES

Berghof DAB III system with Temp. Regler BTP 843, 250ml vessels or

Microwave system Milestone ETHOS

1600, 100 ml vesselsdifferent temperature

programs

2x2 ml H2O2, the first two to wet the sample, 7,5 ml

HNO3, 0,75ml HF; neutralization with 7,5ml

4%H3BO3; fill to 50ml with H2O -

ICP-OES Spectro Ciros axial

120-800 nmWavelenght Cl:

134,724 nm -

heating rate see

ELSAM sheets

acid matrix matched

standards: Merck HCl

wheat straw and spruce wood,

German RR 2000 500 ar < 1 < DL

water soluble Cl + IC danish guideline blue cap bottles

extraction with 50 ml MQ H2O

120°C for 1 hour, possible evaporation is checked by weighing the bottle after

cooling -As for Parr bomb

method IC - -as for Parr bomb

methodas for Parr bomb

method 1000 ar < 1 < DLIFE

water soluble Cl danish guidelineofi

Bomb (10ml) + IC DIN 51900Calorimeter IKA C

5000 bomb combustion10 ml 0,5n

KOH

Dionex DX 320 with EG 40, column AS

17, 2mm IC 30 ?Benzoic acid

(NIST standard)liquid standard

(MERC 1000 mg/l) 700-1000 pellet 23/50 (an) <0,5 0,1 mg/l

Tube Furnace (1000°) + IC - Carbolite RT 1000

tube furnace, pure oxygen 200 ml H2O


17, 2mm IC 50 l/h 1000liquid standard

(MERC 1000 mg/l) 1000 Powder 23/50 (an) <0,5 0,1 mg/l

water soluble Cl + IC danish guideline blue cap bottles

extraction with 50 ml pure H2O 120°C for 1 hour -


17, 2mm IC - -liquid standard

(MERC 1000 mg/l) 1000-2000 Powder 23/50 (an) <0,5 0,1 mg/l



Table A-2: Sulphur determination, experimental details of used methods (to be continued next pages)

Sulphuraccording to or

based onApparatus digestion principle digestion

receiving solution Apparatus determination principle determination additional information

gas pressure or gas flow

com-bustion T .



Sample amount

sample preparation

SP ml/min or bar °C mg

Leco (Automatic Analyzer) SS 187177 LECO SC432

Combustion in high temperature furnace - LECO SC432

Combustion in high temperature furnace- IR

determination of SO2 4000; 90s 1370Coal, ASCRM-013;

S=0,46± 0,014% - 300

pellet, half of pellet for

analysisdk-TEK NIK

Bomb + Turbidimetric Determination

ASTM D 3177 Method B

Parr 1261, bomb: Parr 1108

(stainless steel, 350ml), stainless

steel cruciblesNiCr-ignition wire;

fill to 100mlno water added

to bomb

nephelometer HACH Ration, filtration folded

filter, turbidimetric method detection limit 1 mgSO4/l

(instead of gravimetric)

Precipitation of BaSO4 by adding BaCl2, measurement of

turbidity by a nephelometer

n-dodecane used in every determination;

pressure release > 2min; interferences for

turbidimetric determination 30 -

determination: VHI QC DWB - drinking

water Reference, Sulfat standard

solutions 1000-1100 pellet

Eschka + Turbidimetric Determination ISO 334:1992

muffle furnice, quartz crucibles

30ml, Merck Eschka mix.

combustion of sample with Eschka mixture for 2 hours

or more -

nephelometer HACH Ration, filtration folded

filter, turbidimetric method detection limit 1 mgSO4/l

(instead of gravimetric)

Precipitation of BaSO4 by adding BaCl2, measurement of

turbidity by a nephelometer

Method not used for biofuels, low values for

e.g.straw - 800

determination: VHI QC DWB - drinking

water Reference, Sulfat standard

solutions 1000 powderCIEMAT

Microwave + ICP-OES internal procedure

microwave (closed vessel)

1st stage: HNO3-H2O2-HF. 2nd stage:

H3BO3 50ICP THERMO JARREL

ASHatomic emission

spectrometry - -

prepared multielemental

standard

bush branches and leaves. GBW-

07602, and others 500 powder

Fisons (Automatic Analyzer) internal procedure - Fisons CHNS-1108

combustion in He/O2, SOx are reduced to

SO2, separation by a column, analyzed by

ECD 120 ml He 1000 Coal GBW-11101

Orchard leaves. LECO B214, and

others 4 powderCPERI

Leco (Automatic Analyzer) LECO CS-200 induction furnace - chromatography, IR

ceramic crucible, accelerator material (?) 850

NIST SRM 334, 0,019%S 100

Leco (Automatic Analyzer)

LECO CHNS 932 induction furnace - chromatography, IR 950

LECO 502-062soil, 160ppmS 2

Gravimetric method

Method 426A, American Public

Health Association 1985

oxidation of S to SO4 by Br2 in CCl4 -

precipitation with BaCl2, gravimetric

bomb +IC calorimetric bomb 5ml waterIC, Metrohm 761 with

suppression conductivity detector 30bar 100-1000







com-bustion T.



Sample amount

sample preparation

ARCS

Open digestion + ICP-AES none

open digestion with HNO3/HClO4 2% HClO4 VARIAN Liberty II ICP/AES

Merck-Standard (H2SO4) none

Tube Furnace (H2O2) + IC

ISO 351, DIN 51724-1

Ströhlein tube furnace

combustion with wet oxygen 1% H2O2 Dionex DX-500 IC ~ 2L/min 1250 200 - 1000

Carlo Erba (Fisons)ÖNORM G 1071-

T1 Erba EA 1108 flash combustion - chromatog., TCD sample in tin capsulecarrier gas: 100ml/min 1000

various, e.g. sulfanilamid 2-15 powder

ECN

Berghof + ICP-OESteflon vessels, 10 ml acid mixture

acid mixture: 950 g HNO3, 50g HCl or 500g HNO3, 50 g

HCl, 450 g HF - ICP-AES - see ECN guide - - 500

Carlo Erba (Automatic Analyzer) Carla Erba 1106

combustion in a reaction column with

He/O2 -Porapak QS 0,1meter,

105°C

separation by a porapak QS column,

detection by katharometer sample in tin capsule 1010

Sulphanilamide, BCR Coke CRM 1

to 4; BCR CRM 065 coal - up to 5 powder

Bomb (PARR) + IC

calorimeter Parr 1241, bomb Parr

1108

rinse the bomb with O2; incomplete

combustion: moist loose sample with 200 ml H20 or add

SiO2 or benzoic acid 25 1000-1200 pelletVTT

Eltra CS 500 (Automatic Analyzer) Eltra CS 500

tube furnace, pure oxygen - - IR

ASTM D 4239, method C 2min 1350 coal - 500

Elsam

bomb (Parr)+ IC

Parr 1261 Isoperibol bomb

with 1108CL Oxygen

combustion chlorine-resistant

bomb

after combustion gas is let through a tube into water (100 ml volumetric flask),

content of bomb and capsule is transferred

to flask1ml MQ

water

DIONEX DX500, column AS14, eluent 3,5 mM

Na2Co3, 1 mM NaHCO3, injection through filtercaps,

20mm IC 30bar oxygen -Standards:

Merck Na2SO4


German RR 2000 800-1000 pellet







com-bustion T.



Sample amount

sample preparation

Elsam

Berhof or Microwave + ICP-OES

Berghof DAB III system with

Temp. Regler BTP 843, 250ml

vessels or Microwave

system Milestone ETHOS 1600, 100 ml vessels

2x2 ml H2O2, the first two to wet the

sample, 7,5 ml HNO3, 0,75ml HF; neutralization with

7,5ml 4%H3BO3; fill to 50ml with H2O,

different temperature programs -

ICP-OES Spectro Ciros axial 180-800 nm -

wavelength S: 182,034nm -

heating rate see ELSAM sheets

acid matrix matched standards:

Merck H2SO4


German RR 2000 500

Leco SC 432 (Automatic Analyzer) LECO SC 432 Resistance furnace - LECO SC 432 IR

the sample is covered by Com-aid for solids

(LECO 501-426) 1350

LECO 501-001 calibration sample

(coal),


German RR 2001, coal BCR 180- 200-400

IFEEschkabomb + ICP

ofi

Bomb (10ml) + IC DIN 51900Calorimeter IKA

C 5000 bomb combustion10 ml 0,5n

KOHDionex DX 320 with EG 40, column AS 17, 2mm IC 30

about 1250

Benzoic acid (NIST standard)

liquid standard (MERC 1000 mg/l) 700-1000 pellet

Tube Furnace (1000°) + IC

Carbolite RT 1000

tube furnace, pure oxygen 200 ml H2O

Dionex DX 320 with EG 40, column AS 17, 2mm IC 50 l/h 1000

liquid standard (MERC 1000 mg/l) 1000 powder



Table A-3: Nitrogen determination, experimental details of used methods (to be continued next pages)

Nitrogen Apparatus principlegas flow (if applicable)

additional information

com-bustion temp.


Sample amount

sample preparation

sample (water

content)sample

particle size


°C mgpowder, pellet,... ar, an, wf mm

SP

Leco (Automatic Analyzer) LECO CHN 2000

CHN-Analyser; Determination by IR and

GC-TCD

4,5 l/min for 45s, 2L/min for

35 s 1000EDTA from LECO: N:

9,57±0,04% 150-200

pressed in a tin foil capsule

using a LECO pellet press

Normally <1 mm

aprox 0,0004

dk-TEKNIK

Vario EL (Automatic Analyzer) Vario-EL

combustion in He/O2, NOx are reduced to N2, purified from other gases (traps), analyzed by TCD

first sample measurement is discarded (for a

tripple determination 4 samples are

prepared) 950

basic calibration: sulfanilic acid,

acetanilide, atropin, daily factor: 6 x

acetanilid, control standards: atropin,

benzoic acid (blank), BCR 180 (coal) 5-25 powder

wf (if H is determined

<1mm (heterogenous

samples smaler

CIEMAT

Fisons (Automatic Analyzer) Fisons CHNS-1108

combustion in He/O2, NOx are reduced to N2, separation by a column,

analyzed by TCD 120 ml He 1000 Coal GBW-11101 4 powder wf <0,5 mmCPERI

tube furnace with gas detectors

quartz tube furnace, O2/He 20:80, NOx-analyser Thermo Environmental

Instruments 42C, NO2 analyzer

Horiba VIA510)

combustion, detection by chemiluminescence

detector (NOx) or NDIR (N2O) does not work! 800-900





com-bustion temp.


Sample amount

sample preparation

sample (water

content)sample

particle size



CPERI

Kjeldahl, method 1

sample, catalyst and 20ml conc. H2SO4are put into Kjeldahl flask, heated to boiling point until a clear solution is obtained and additional 60 min. After cool down, transfer to a

distillation unit, add 60 ml water and 90ml 32% NaOH Up to 200ml

distillate -titration with H2SO4

standard solution - 1000

Kjeldahl, method 2

Photometer DR/2000 Hach

company; wavelength 460nm, range 0-150 mg/l, precision 0,8 mg/l

sample dissolved in 3ml H2SO4, digestion

temperature 440°C for 3-5 min.; add 10 ml H2O2 50%

to give a clear solution. Acess H2O2 is boiled off. Cool down and add water to volume 100ml. 1-2 ml

transferred to 25ml graduated cylinder,

neutralized with KOH, d -

Add 1ml Nessler reagent and analyse

by photometer - 100-200ARCS

Carlo Erba (Automatic Analyzer) Erba EA 1108

flash combustion, chromatography, TCD

carrier gas: 100ml/min 1000

various, e.g. sulfanilamid 2 - 15 powder

all, preferably an <0,2

Kjeldahl

Heating block assembly, distillation apparatus according to DIN 51722-1 200 - 2000 powder

all, preferably an <0,2





com-bustion temp.


Sample amount

sample preparation

sample (water

content)sample

particle size



ECN

Carlo Erba (Automatic Analyzer)

Carla Erba 1106, Porapak QS

0,1meter, 105°C

combustion in a reaction column with He/O2,

separation by a column, detection by katharometer - 1030

Trishydroxymethylaminomethane, BCR Coke CRM 1 to 4;

BCR CRM 065 coal up to 5 powder wf < 0,2

KjeldahlVTT

Leco (Automatic Analyzer) Leco CHN-2000

ASTM D 5273: combustion in O2,

detection by thermal conductivity (N) EDTA 200 an

Elsam

Carlo Erba (Automatic Analyzer) Carlo Erba 1108

Flash dynamic combustion in tin capsules

carrier gas: 125ml/min 1000

BBOT standard 6,51%N for cal.

curve: 0,013-0,13 mg N total Wheat straw

and Spruce wood, German RR 2000,

BCR 180-coal 4-20 wf < 1mmIFE

Kjeldahl

250ml Kjeldahl decomposition

flasks, distillation apparatus 12 (Gerhardt,

Erlenmeyer 250ml

DIN ISO 11261 with deviations, determination

by DIN 38406-E 5-2 (titrimetric) - - 500-1000 powder an < 0,1



chlorine content wood without bark

0,000

0,001

0,002

0,003

0,004

0,005

0,006

0,007

0,008

0,009

0,010

(AOX) A

RCS

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) IFE

(Bom

b + IC

) ofi

bomb dk

- IC of

i

(Bom

b + IC

) SP

(Bom

b + tit

ration)

VTT

(Bom

b + IC

) Cperi

(Esc

hka)

CIEMAT

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(tube

furna

ce) A

RCS

(tube

furna

ce) E

CN

(wate

r soluble

Cl) d

k

(wate

r soluble

Cl) o

fi

(wate

r soluble

Cl) E

CN

(wate

r soluble

Cl) E

lsam

water solu

ble dk/ofi

(Wick

bold) IF

E

INAA E

CN

PAN Analyt

. XRF

%

Figure A-1: chlorine content of the sample wood without bark, different methods[wt%, d.b.]

chlorine content woodchips

0,000

0,002

0,004

0,006

0,008

0,010

0,012

0,014

(AOX) A

RCS

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) IFE

(Bom

b + IC

) ofi

bomb dk

- IC of

i

(Bom

b + IC

) SP

(Bom

b + tit

ration)

VTT

(Bom

b + IC

) Cperi

(Esc

hka)

CIEMAT

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(tube

furna

ce) A

RCS

(tube

furna

ce) E

CN

(wate

r soluble

Cl) d

k

(wate

r soluble

Cl) o

fi

(wate

r soluble

Cl) E

CN

(wate

r soluble

Cl) E

lsam

water solu

ble dk/ofi

(Wick

bold) IF

E

INAA E

CN

PAN Analyt

. XRF

%

Figure A-2: chlorine content of the sample woodchips, different methods [wt%, d.b.]



chlorine content bark

0,000

0,002

0,004

0,006

0,008

0,010

0,012

0,014

0,016

0,018

0,020

(AOX) A

RCS

(Bom

b + Titra

tion)

dk

(Bom

b + Titra

tion)

dk

(Bom

b + Titra

tion)

dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) IFE

(Bom

b + IC

) ofi

bomb dk

- IC of

i

(Bom

b + IC

) SP

(Bom

b + tit

ration)

VTT

(Bom

b + IC

) Cperi

(Esc

hka)

CIEMAT

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(Tube fu

rnace

) ARCS

(Tube fu

rnace

) ECN

(wate

r soluble

Cl) d

k

(wate

r soluble

Cl) o

fi

(wate

r soluble

Cl) E

CN

(wate

r soluble

Cl) E

lsam

water solu

ble dk/ofi

(Wick

bold) IF

E

INAA E

CN

PAN Analyt

. XRF

%

Figure A-3: chlorine content of the sample bark, different methods [wt%, d.b.]

chlorine content hardwood with glue

0,00

0,01

0,01

0,02

0,02

0,03

0,03

(AOX) A

RCS

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) IFE

(Bom

b + IC

) ofi

bomb dk

- IC of

i

(Bom

b + IC

) SP

(Bom

b + tit

ration)

VTT

(Bom

b + IC

) CPERI

(Esc

hka)

CIEMAT

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(tube

furna

ce) A

RCS

(tube

furna

ce) E

CN

(wate

r soluble

Cl) d

k

(wate

r soluble

Cl) o

fi

(wate

r soluble

Cl) E

CN

(wate

r soluble

Cl) E

lsam

water solu

ble dk/ofi

(Wick

bold) IF

E

INAA E

CN

PAN Analyt

. XRF

%

Figure A-4: chlorine content of the sample hardwood with glue, different methods [wt%, d.b.]



chlorine content straw

0,00

0,020,04

0,060,08

0,100,12

0,140,16

0,180,20

(AOX) A

RCS

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) IFE

(Bom

b + IC

) ofi

bomb dk

- IC of

i

(Bom

b + IC

) SP

(Bom

b + tit

ration)

VTT

(Bom

b + IC

) CPERI

(Esc

hka)

CIEMAT

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(tube

furna

ce) A

RCS

(tube

furna

ce) E

CN

(wate

r soluble

Cl) d

k

(wate

r soluble

Cl) o

fi

(wate

r soluble

Cl) E

CN

(wate

r soluble

Cl) E

lsam

water solu

ble dk/ofi

(Wick

bold) IF

E

INAA E

CN

PAN Analyt

. XRF

%

Figure A-5: chlorine content of the sample straw, different methods [wt%, d.b.]

chlorine content orujilo

0,00

0,05

0,10

0,15

0,20

0,25

0,30

0,35

(AOX) A

RCS

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) IFE

(Bom

b + IC

) ofi

bomb dk

- IC of

i

(Bom

b + IC

) SP

(Bom

b + tit

ration)

VTT

(Bom

b + IC

) CPERI

(Esc

hka)

CIEMAT

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(tube

furna

ce) A

RCS

(tube

furna

ce) E

CN

(wate

r soluble

Cl) d

k

(wate

r soluble

Cl) o

fi

(wate

r soluble

Cl) E

CN

(wate

r soluble

Cl) E

lsam

water solu

ble dk/ofi

INAA E

CN

PAN Analyt

. XRF

%

Figure A-6: chlorine content of the sample orujilo, different methods [wt%, d.b.]



chlorine content rapestraw

0,00

0,05

0,10

0,15

0,20

0,25

0,30

0,35

0,40

0,45

(AOX) A

RCS

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) IFE

(Bom

b + IC

) ofi

bomb dk

- IC of

i

(Bom

b + IC

) SP

(Bom

b + tit

ration)

VTT

(bomb+

IC) C

PERI

(Esc

hka)

CIEMAT

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(tube

furna

ce) A

RCS

(tube

furna

ce) E

CN

(wate

r soluble

) dk

(wate

r soluble

) ofi

(wate

r soluble

) ECN

(wate

r soluble

) Elsa

m

water solu

ble dk/ofi

INAA E

CN

PAN Analyt

. XRF

%

Figure A-7: chlorine content of the sample rapestraw, different methods [wt%, d.b.]

chlorine content cynara

0,0

0,5

1,0

1,5

2,0

2,5

(AOX) A

RCS

(Bom

b + Titra

tion)

dk

(Bom

b + Titra

tion)

dk

(Bom

b + Titra

tion)

dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) IFE

(Bom

b + IC

) ofi

bomb dk

- IC of

i

(Bom

b + IC

) SP

(Bom

b + tit

ration)

VTT

(Bom

b + IC

) CPERI

(Esc

hka)

CIEMAT

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(tube

furna

ce) A

RCS

(tube

furna

ce) E

CN

(wate

r soluble

Cl) d

k

(wate

r soluble

Cl) o

fi

(wate

r soluble

Cl) E

CN

(wate

r soluble

Cl) E

lsam

water solu

ble dk/ofi

%

Figure A-8: chlorine content of the sample cynara, different methods [wt%, d.b.]



chlorine content hemp

0,000,020,040,060,080,100,120,14

ofi Force ECN Elsam Ciemat VTT

%

Figure A-9: chlorine content of the sample hemp, standardized bomb method, determination by IC in one

laboratory [wt%, d.b.]

sulphur content wood without bark

0,0000,0020,0040,0060,0080,0100,0120,0140,0160,018

(Bom

b + tu

rbid.)

dk

(Bom

b + tu

rbid.) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) ofi

(Bom

b + IC

P) IFE

(Erba)

ARCS

(Erba)

CIEMAT

(Erba)

ECN

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(HPA +I

CP) ARCS

(HPA +I

CP) CIE

MAT

(HPA +I

CP) ECN

(Leco) E

lsam

(Leco) S

P

PAN Analyt

. XRF

(LecoCS20

0) CPERi

(LecoCS20

0) CPERi

(Grav

i.) CPERI

(Bom

b+IC

) CPERI

(Eltra

) VTT

(tube

+ IC) A

RCS

[%]

Figure A-10: sulphur content of the sample wood without bark, different methods [wt%, d.b.]

sulphur content woodchips

0,0010,003

0,0050,0070,0090,011

0,0130,0150,0170,019

(Bom

b + tu

rbid.)

dk

(Bom

b + tu

rbid.) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) ofi

(Bom

b + IC

P) IFE

(Erba)

ARCS

(Erba)

CIEMAT

(Erba)

ECN

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(HPA +I

CP) ARCS

(HPA +I

CP) CIE

MAT

(HPA +I

CP) ECN

(Leco) E

lsam

(Leco) S

P

PAN Analyt

. XRF

[%]

Figure A-11: sulphur content of the sample woodchips, different methods [wt%, d.b.]



sulphur content hardwood with glue

0,001

0,006

0,011

0,016

0,021

0,026

(Bom

b + tu

rbid.)

dk

(Bom

b + tu

rbid.) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) ofi

(Bom

b + IC

P) IFE

(Erba)

ARCS

(Erba)

CIEMAT

(Erba)

ECN

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(HPA +I

CP) ARCS

(HPA +I

CP) CIE

MAT

(HPA +I

CP) ECN

(Leco) E

lsam

(Leco) S

P

PAN Analyt

. XRF

(LecoCS20

0) CPERi

(LecoCS20

0) CPERi

(Grav

i.) CPERI

(Bom

b+IC

) CPERI

(Eltra

) VTT

(tube

+ IC) A

RCS

[%]

Figure A-12: sulphur content of the sample hardwood with glue, different methods [wt%, d.b.]

sulphur content bark

0,00

0,02

0,04

0,06

0,08

0,10

0,12

0,14

(Bom

b + tu

rbid.)

dk

(Bom

b + tu

rbid.) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) ofi

(Bom

b + IC

P) IFE

(Erba)

ARCS

(Erba)

CIEMAT

(HPA +

ICP) E

lsam

(HPA +I

CP) ARCS

(HPA +I

CP) CIE

MAT

(HPA +I

CP) ECN

(Leco) E

lsam

(Leco) S

P

PAN Analyt

. XRF

[%]

Figure A-13: sulphur content of the sample bark, different methods [wt%, d.b.]



sulphur content straw

0,000,020,040,060,080,100,120,140,160,18

(Bom

b + tu

rbid.)

dk

(Bom

b + tu

rbid.) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) ofi

(Bom

b + IC

P) IFE

(Erba)

ARCS

(Erba)

CIEMAT

(Erba)

ECN

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(HPA +I

CP) ARCS

(HPA +I

CP) CIE

MAT

[%]

Figure A-14: sulphur content of the sample straw, different methods [wt%, d.b.]

sulphur content orujilo

0,001

0,051

0,101

0,151

0,201

(Bom

b + tu

rbid.)

dk

(Bom

b + tu

rbid.) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) ofi

(Bom

b + IC

P) IFE

(Erba)

ARCS

(Erba)

CIEMAT

(Erba)

ECN

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(HPA +I

CP) ARCS

(HPA +I

CP) CIE

MAT

(HPA +I

CP) ECN

(Leco) E

lsam

(Leco) S

P

PAN Analyt

. XRF

[%]

Figure A-15: sulphur content of the sample orujilo, different methods [wt%, d.b.]



sulphur content cynara

0,00

0,05

0,10

0,15

0,20

0,25

0,30

0,35

(Bom

b + tu

rbid.)

dk

(Bom

b + tu

rbid.) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) ofi

(Bom

b + IC

P) IFE

(Erba)

ARCS

(Erba)

CIEMAT

(Erba)

ECN

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(HPA +I

CP) ARCS

(HPA +I

CP) CIE

MAT

(HPA +I

CP) ECN

(Leco) E

lsam

(Leco) S

P

PAN Analyt

. XRF

[%]

Figure A-16: sulphur content of the sample cynara, different methods [wt%, d.b.]

sulphur content rapestraw

0,00

0,05

0,10

0,15

0,20

0,25

0,30

0,35

(Bom

b + tu

rbid.)

dk

(Bom

b + tu

rbid.) dk

(Bom

b + IC

) ECN

(Bom

b + IC

) Elsa

m

(Bom

b + IC

) ofi

(Bom

b + IC

P) IFE

(Erba)

ARCS

(Erba)

CIEMAT

(Erba)

ECN

(Esc

hka)

dk

(HPA +

ICP) E

lsam

(HPA +I

CP) ARCS

(HPA +I

CP) CIE

MAT

(HPA +I

CP) ECN

(Leco) E

lsam

(Leco) S

P

PAN Analyt

. XRF

[%]

Figure A-17: sulphur content of the sample rapestraw, different methods [wt%, d.b.]

sulfur content hemp

0,00

0,01

0,02

0,03

0,04

0,05

ofi dk ECN Elsam Ciemat VTT

%

Figure A-18: sulphur content of the sample hemp, standardized bomb method, determination by IC in one laboratory

[wt%, d.b.]



nitrogen content wood without bark

0,00

0,02

0,04

0,06

0,08

0,10

0,12

(Erba)

ARCS

(Erba)

ECN

(Erba)

Elsam

(Fisons

) CIEMAT

(Kjel

dahl)

ARCS

(Kjel

dahl)

ECN

(Leco) S

P

(Leco) V

TT

(Vari

o EL) dk

(Kjel

dahl

I) CPERI

(Kjel

dahl

II) CPERI

(Kjel

dahl)

IFE

%

Figure A-19: nitrogen content of the sample wood without bark, different methods [wt%, d.b.]

nitrogen content hardwood with glue

0,00

0,10

0,20

0,30

0,40

0,50

(Erba)

ARCS

(Erba)

ECN

(Erba)

Elsam

(Fisons

) CIEMAT

(Kjel

dahl)

ARCS

(Kjel

dahl)

ECN

(Leco) S

P

(Leco) V

TT

(Vari

o EL) dk

(Kjel

dahl

I) CPERI

(Kjel

dahl

II) CPERI

(Kjel

dahl)

IFE

%

Figure A-20: nitrogen content of the sample hardwood with glue, different methods [wt%, d.b.]



nitrogen content woodchips

0,00

0,05

0,10

0,15

0,20

0,25

(Erba)

ARCS

(Erba)

ECN

(Erba)

Elsam

(Fisons

) CIEMAT

(Kjel

dahl)

ARCS

(Kjel

dahl)

ECN

(Leco) S

P

(Leco) V

TT

(Vari

o EL) dk

(Kjel

dahl

I) CPERI

(Kjel

dahl

II) CPERI

(Kjel

dahl)

IFE

%

Figure A-21: nitrogen content of the sample woodchips, different methods [wt%, d.b.]

nitrogen content rapestraw

0,00

0,10

0,20

0,30

0,40

0,50

0,60

0,70

(Erba)

ARCS

(Erba)

ECN

(Erba)

Elsam

(Fisons

) CIEMAT

(Kjel

dahl)

ARCS

(Kjel

dahl)

ECN

(Leco) S

P

(Leco) V

TT

(Vari

o EL) dk

(Kjel

dahl

I) CPERI

(Kjel

dahl

II) CPERI

(Kjel

dahl)

IFE

%

Figure A-22: nitrogen content of the sample rapestraw, different methods [wt%, d.b.]

nitrogen content bark

0,00

0,20

0,40

0,60

0,80

1,00

1,20

(Erba)

ARCS

(Erba)

ECN

(Erba)

Elsam

(Fisons

) CIEMAT

(Kjel

dahl)

ARCS

(Kjel

dahl)

ECN

(Leco) S

P

(Leco) V

TT

(Vari

o EL) dk

(Kjel

dahl

I) CPERI

(Kjel

dahl

II) CPERI

(Kjel

dahl)

IFE

%

FigureA-23: nitrogen content of the sample bark, different methods [wt%, d.b.]



nitrogen content straw

0,00

0,20

0,40

0,60

0,80

1,00

1,20

(Erba)

ARCS

(Erba)

ECN

(Erba)

Elsam

(Fisons

) CIEMAT

(Kjel

dahl)

ARCS

(Kjel

dahl)

ECN

(Leco) S

P

(Leco) V

TT

(Vari

o EL) dk

(Kjel

dahl

I) CPERI

(Kjel

dahl

II) CPERI

(Kjel

dahl)

IFE

%

Figure A-24: nitrogen content of the sample straw, different methods [wt%, d.b.]

nitrogen content cyanara

0,00

0,200,40

0,600,80

1,00

1,201,40

1,601,80

(Erba)

ARCS

(Erba)

ECN

(Erba)

Elsam

(Fisons

) CIEMAT

(Kjel

dahl)

ARCS

(Kjel

dahl)

ECN

(Leco) S

P

(Leco) V

TT

(Vari

o EL) dk

(Kjel

dahl

I) CPERI

(Kjel

dahl

II) CPERI

(Kjel

dahl)

IFE

%

Figure A-25: nitrogen content of the sample cynara, different methods [wt%, d.b.]

nitrogen content orujillo

0,00

0,50

1,00

1,50

2,00

(Erba)

ARCS

(Erba)

ECN

(Erba)

Elsam

(Fisons

) CIEMAT

(Kjel

dahl)

ARCS

(Kjel

dahl)

ECN

(Leco) S

P

(Leco) V

TT

(Vari

o EL) dk

(Kjel

dahl

I) CPERI

(Kjel

dahl

II) CPERI

(Kjel

dahl)

IFE

%

Figure A-26: nitrogen content of the sample orujilo, different methods [wt%, d.b.]



7.1 Chlorine determination using the Eschka method

7.1.1 Background

The results of FORCE (former dk-TEKNIK) in the initial test for the chlorine content in the 8 samples using the ISO 587/Eschka method were systematically too low. Both compared to the results obtained by FORCE using other methods (Bomb method and water soluble method) and to the results obtained by the other laboratories. This was no surprise as FORCE previous had realized that the Eschka method was not a suitable method for the determination of chlorine content in solid biofuels due to too low results.

From the results of the initial test it however appeared, that the results reported by CIEMAT, also using an Eschka method, were satisfactory. By that the conclusion of FORCE that the Eschka method is not suitable for solid biofuels seems to be wrong. In table 4 is reproduced some of the results of the initial test for chlorine, according to the two result sheets distributed at the WPIII.1 meeting, October 2003 in Madrid.

Table 4: Reported values of chlorine content for the initial test [wt %]. Results from FORCE, CIEMAT and mean values.

FORCE (dk-TEKNIK) 1) CIEMAT Sample

Bomb meth. Water-soluble Eschka Eschka Mean,

all results Wood without bark < 0.01 < 0.01 < 0.02 < 0.01 0.0022 Wood + bark < 0.01 < 0.01 < 0.02 0.013 0.0048 Bark < 0.01 < 0.01 < 0.02 0.010 0.0094 Wood with glue 0.014 0.011 < 0.02 < 0.01 0.014 Straw 0.11 0.11 0.085 0.11 0.11 Rape straw 0.27 0.27 0.18 0.25 0.26 Orujillo 0.18 0.19 0.16 0.19 0.18 Cynara 1.3 2) 1.6 1.2 1.6 1.4 1)Results reported on as determined (moist) basis. The basis of the reported results from CIEMAT and the

other laboratories is not known by FORCE. 2)At the subsequent experiments, September 2003, a chlorine content of 1,5 % was found when using a

reduced pressure release time of 60 second.

7.1.2 Comparison of the Eschka methods used by FORCE and CIEMAT

FORCE used the method described in ISO 587:1997. According to ISO 587:1997 however three different methods for the completion of the determination are described in Cl. 7.2; the Volhard (7.2.1), the Mohr (7.2.2) and the ISE (7.2.3) method. FORCE use the ISE (potentiometric titration) method, BUT due to an oversight by the emergence of ISO 587:1997 the incinerated mixture was still prepared as described in 7.2.2 for the Mohr method (in the former edition of ISO 587 the ISE method was not included). For some unknown reason the preparation of the incinerated mixture is different for the Mohr method (7.2.2) than for the other two methods (7.2.1 and 7.2.3):

Cl. 7.2.2: After the transfer of the incinerated mixture to a beaker and the washing of the crucible with hot water, the solution (in the beaker) is heated to boiling



point and filtered. The filtrate is neutralized with conc. nitric acid.

Cl. 7.2.1+ Cl. 7.2.3: After the transfer of the incinerated mixture to a beaker and the washing of the crucible with hot water, 20 ml conc. nitric acid is added to the beaker. The content of the beaker is swirled/stirred and if necessary, filtered.

According to the review of the methods (also distributed at the Madrid meeting), CIEMAT use the method described in ASTM D2361-66 combined with the national UNE32024. After the Madrid meeting Mr. Miguel Fernández provided FORCE with the actual used method at CIEMAT for the preparation of the sample (before analysis of the chloride content in the solution). In table 5 is summarized the principal differences between the procedures used by FORCE and by CIEMAT.

Table 5: Principal differences between FORCE and CIEMAT procedures

Condition FORCE CIEMAT Comment

Size of test portion 2 g - < 0.1 %Cl

1 g - 0.1 – 2 % Cl 0,5 g - > 2 % Cl

5 g - 0.005-0.4 %Cl 2,5 g - 0.01–0.07%Cl 0,5 g - 0.05- 4 % Cl

CIEMAT are using larger testportions at

low Cl contents. ISO1) and ASTM2) prescribe

approx. 1 g coal sample.

Size of crucible 25 ml 50 ml ISO1) prescribe 25 ml, ASTM2) 30 ml

Amount Eschka Mixture, total 4 g 5 g

Distribution of the Eschka mixture

Bottom: 0.5g Mixed with sample:2.5g

Cover: 1.0g

Bottom:- g Mixed with sample: 3.0g

Cover: 2.0g

Amount and distribu-tion of Eschka mix-ture according to ISO1) and ASTM2), repectively

Combustion temperature 675 oC 775 oC Both ISO1) and ASTM2)

prescribe 675 oC

Preparation of the incinerated mixture

After the transfer of the incinerated mixture to a

beaker and the washing of the crucible with hot water, the solution (in the beaker) is heated to boiling point

and filtered. The filtrate is neutrally-zed with conc.

nitric acid.

After the transfer of the incinerated mixture to a

beaker and the washing of the crucible with hot water,

20 ml conc. nitric acid is added to the beaker. The content of the beaker is

agitated and if necessary, filtered.

FORCE procedure according to ISO1)

Cl.7.2.2. CIEMAT procedure

equivalent to ASTM2) (20 ml conc. nitric acid used in stead of 40 ml

(1+1) nitric acid).

Determination of chloride content in the solution

ISE (potentiometric titration) Volhard

1) ISO 587:1997 2) ASTM D 2361-95 (ASTM D 2361-66 not available at FORCE)

7.1.3 Follow-up tests

For the samples “Cynara” and “Rape straw” (the samples with the highest content of chlorine) determination of the chlorine content were carried out using a method modified with the deviant



conditions of CIEMAT according to table 5 – except that crucibles of 25 ml and the ISE method were still used.

As the obtained results for both samples were higher than previous reported by FORCE (dk-TEKNIK) for the Eschka method, two more modifications were investigated in order to identify the influential conditions, see Table 6.

Table 6: Content of chlorine, [wt %] dry basis, by different Eschka procedures, FORCE lab

Sample Reported

2002 (dk-TEKNIK)

Modification I Mean

(double determination)

Modification II Mean


Modification III Mean


Rape straw 0.19 0.23 (0.227; 0.228)

0.23 (0.234;0.229)

0.21 (0.216;0.209)

Cynara 1.3 1.6 (1.56;1.57)

1.6 (1.59;1.59)

1.5 (1.45;1.49)

Modification I : CIEMAT procedure, except for crucibles of 25 ml and ISE method Modification II : As modification I, but a combustion temperature of 675 oC Modification III : FORCE procedure (as used in the initial test) but modified with treatment of

the incinerated mixture according to CIEMAT procedure (addition of 20 ml conc. nitric acid).

It was hereby concluded that:

• The difference in combustion temperature (675 oC >< 775 oC) used had no significant influence on the result;

• A higher amount of Eschka’s mixture and especially a higher amount used as cover improve the results (giving higher results of chlorine content);

• The correct treatment of the incinerated mixture, according to ISO 587:1997 Cl. 7.2.3 and ASTM D 2361-95 (adding the 20 ml conc. nitric acid) improve the results considerably.

By this, the conclusion of the experiments is, that the procedure used by FORCE was not optimal regarding samples of solid biofuels and that a procedure as described in ASTM D2361 (or by CIEMAT) should be used if content of chlorine in a solid biofuels is to be determined by the Eschka method.

The need of the higher amount of Eschka mixture for the cover may be due to the low density of solid biofuels (compared to coal).

The deviant preparation in ISO 587:1995 Cl. 7.2.2 for the completion by the Mohr method, compared to the preparation described in Cl. 7.2.1 and 7.2.3, is unaccountable. It should be investigated whether this deviant procedure also is influential for coal.

Based on the results the procedure of FORCE has been modified with the preparation of the incinerated mixture as described in ISO 587:1995 Cl. 7.2.3 (and ASTM D2361-95) and for solid biofuels with the amount and distribution of Eschka mixture as described in ASTM D 2361-95.

Using the modified procedure, triple determinations of all the 8 samples were carried out.



7.1.4 Results

The obtained results using the modified Eschka procedure appear from the excel document “Bionorm WPIII.1 – follow-up on initial test results of chlorine Eschka method”.

In Table 7 is reproduced the obtained mean results for the 8 samples, together with the previous reported results.

Table 7: Chlorine content for the 8 test samples

FORCE (dk-TEKNIK) [wt %] Cl, dry sample

Eschka method

CIEMAT [wt %] Cl

Eschka method

All results [wt %] Cl

Diff. Methods Sample

Reported 2002 Mod. Method 2004 According to data sheets distributed at the WG 5 meeting in Madrid 6.10.2003

Wood without bark < 0.022 < 0.011 < 0.01 0.0022 Wood + bark < 0.022 < 0.011 0.013 0.0048 Bark < 0.022 < 0.011 0.010 0.0094 Wood with glue < 0.022 < 0.011 < 0.01 0.014 Straw 0.092 0.10 0.11 0.11 Rape straw 0.19 0.23 0.25 0.26 Orujillo 0.17 0.19 0.19 0.18 Cynara 1.3 1.6 1.6 1.4

From Table 7 can be seen, that the modification of the FORCE procedure has improved the method as the results for the 2 straw samples and the Orujillo and Cynara samples has got closer to the results of CIEMAT and all laboratories. The results of the two straw samples are however still a bit on the low side.

Concerning the 4 “wooden” samples, still no chlorine could be detected. It should however be mentioned that the correct treatment of the incinerated mixture complicated the procedure:

• after the addition of the 20 ml conc. nitric acid the pH-value of the solutions were in the area of –0.9 to –0.2. This meant that we had to neutralize the solution (with 4 M NaOH);

• due to the very high salt content of the solutions we experienced problems with the chloride determination (potentiometric titration using automatic titration equipment). The problem was that the equipment from time to time for the blanks and for samples with low chlorine content missed the inflection point – and by that the results. The criterion for the “true” inflection point by that shall be very carefully defined.



Part 2.3.2 Task III.2 – Determination of major and minor elements

Final report prepared by: Georg Bärnthaler1), Ingwald Obernberger1) Contributing co-authors: Conny Haraldsson2), Edith Thomsen3), Tatjana Arnold4), Thore von Puttkamer5), Andrea Vityi6), Michael Zischka7)

1) Institute for Resource Efficient and Sustainable Systems, Graz University of Technology, Austria 2) Swedish National Testing and Research Institute, Sweden 3) Elsam A/S (Techwise), Denmark 4) Institute for Energy Process Engineering and Chemical Engineering, Technical University

Bergakademie Freiberg, Germany 5) Institute of Process Engineering and Power Plant Technology, University of Stuttgart, Germany 6) University of West-Hungary, Technical Institute of Forestry and Environmental Sciences,

Department of Energetics, Hungary 7) Institute for Analytical Chemistry, Graz University of Technology, Austria

1 Summary

Task III.2 of WP III was concerned with the development of reliable and appropriate methods for the accurate determination of major (Al, Ca, Fe, K, Mg, Na, P, Si, Ti) and minor (As, Ba, Cd, Co, Cr, Cu, Hg, Mn, Mo, Ni, Pb, Sb, Tl, V, Zn) ash forming elements in solid biofuels in order to support the compilation of respective CEN standards. As first step of the project work a study was carried out to give a survey of the state-of-the-art of existing determination methods, which can be recommended for analyses of the elements in question. The next step of the project work was concerned with the preparation of sample materials. A sufficient amount of homogeneous material was needed for the tests planned within Task III.2. Pre-tests revealed that particle sizes <0.25 mm provided satisfactory homogeneity ranges. 10 kg wood+bark and straw materials were milled to this particle size and filled in bottles of 20 g each. Subsequently, F-tests were carried out to check the within and between bottle homogeneity. Summarising the outcome of these tests, both materials proved to be suitable for the further investigations planned within Task III.2. Additional investigations revealed that materials with particle sizes <1 mm also provided satisfactory homogeneities. The experiences gained from the sample preparation are summarised in best practice guidelines for the preparation of analysis samples. The next step of the project work was concerned with the optimisation of analytical techniques. For this purpose, wood+bark, straw, olive residues and a reference material were analysed using several digestion and determination methods. The digestion methods investigated included wet digestion in closed vessels with different acid mixtures as well as dry-ashing techniques. The determination systems examined included FAAS, GFAAS, CVAAS, ICP-OES, ICP-MS, XRF as well as direct Hg determination. To summarise these examinations, wet digestion with H2O2 / HNO3 / HF / H3BO3 for major elements and wet digestions with H2O2 / HNO3 / HF for minor elements proved to be the most suitable methods for decomposition. The most suitable determination methods included FAAS, GFAAS, ICP-OES, ICP-MS, CVAAS and direct Hg determination. In order to evaluate the performance of these methods validations were carried out. For this purpose wood+bark, straw and reference materials were analysed and the following statistical and



technical parameters were calculated: overall mean, repeatability limit r, reproducibility limit R, critical difference to certified reference values, detection limit as well as application ranges of methods for the different elements. Summarising the validation results, the applied digestion methods may be recommended for solid biofuel decomposition. For determination of the different elements in solid biofuels, the following detection systems may be recommended: FAAS for Ca, Fe, K, Mg, Na, Si, Mn, Zn; ICP-OES for Al, Ca, Fe, K, Mg, Na, P, Si, Ti, Mn, Zn, Ba, Cr, Cu, Mn, Ni, V, Zn; GFAAS for Cd, Cr, Cu, Ni, Pb; ICP-MS for P, Ti, As, Ba, Cd, Co, Cr, Cu, Mn, Mo, Ni, Pb, Sb, V, Zn and direct Hg determination as well as CVAAS for Hg.

The experiences gained from the methodological investigations as well the resulting recommendations from validation are summarised in the “Best practice guidelines for the determination of sulphur, chlorine and nitrogen as well as major and minor elements in solid biofuels”.

2 Objectives

The application of fuels of known quality is an essential prerequisite for secure and efficient biomass combustion. Therefore, producers, traders and users of solid biofuels relay on accurate standardised methods for fuel quality determinations. Important parameters for assessing the quality of solid biofuels are the contents of major (Al, Ca, Fe, K, Mg, Na, P, Si, Ti) and minor (As, Ba, Cd, Co, Cr, Cu, Hg, Mn, Mo, Ni, Pb, Sb, Tl, V, Zn) ash forming elements. While major elements are of key relevance regarding ash melting, deposit and slag formation as well as corrosion, minor elements are of special importance for particulate emissions as well as the environmental assessment of the ashes produced and their subsequent utilisation. Specific standard operating procedures for the determination of these elements in biofuels exist (or are under development) only in a few cases [1]. Most laboratories thus determine major and minor elements in biofuels using “in house” methods or standards originally developed for solid mineral fuels. This often leads to considerable deviations between the results of different laboratories as shown by several round robins on various solid biofuels [2, 3, 4].

Therefore, Task III.2 was concerned with the development of standardised procedures for the determination of major and minor elements in solid biofuels. The objectives were to test, select, evaluate and validate analytical methods for the correct measurement of these elements. The work foreseen to reach this goal included the performance of a state-of-the-art study concerning appropriate analytical methods, the preparation of homogeneous materials for the investigations planned and the testing of the homogeneity of these materials, examinations to compare suitable analytical methods, the improvement of the most suitable ones and the validation of the most appropriate methods. By this way statistically proven methods with documented uncertainty should be worked out and best practice guidelines for the correct determination of these elements in solid biofuels should be prepared. The findings of these investigations were foreseen to be fed into the development of respective CEN standards.




Two for European conditions representative types of biomass,

• a coniferous wood+bark mixture (particle size <0.25 mm) and

• wheat straw (particle size <0.25 mm),

were used for the majority of the investigations. For preliminary examinations, different particle sizes of these materials (<1 mm, <0.25 mm, <0.1 mm) were compared. Olive residues (particle size <0.25 mm) as well as certified biomass reference materials (GBW 07602 (NCS DC 73348) “Bush Branches and Leaves”, issued 1997 by the China National Analysis Centre for Iron and Steel, Beijing, or NIST SRM 1575a “Trace elements in Pine Needles”) were also included for some investigations.


The work within Task III.2 was divided into the following 5 Subtasks:

• Subtask III.2.1 “Definition of the state-of-the-art”

• Subtask III.2.2 “Sample homogenisation and homogeneity testing”

• Subtask III.2.3 “Method testing and improvement – part 1 and 2”

• Subtask III.2.4 “Method comparison and validation”

• Subtask III.2.5 “Method evaluation and compilation of best practice guidelines”

4.1 Sub-task III.2.1 “Definition of the state-of-the-art”

As first step of the project work, a study was carried out to give a survey of the state-of-the-art of those existing determination methods which are commonly applied for analyses of the elements in question and to work out their advantages and disadvantages [5]. For this purpose relevant literature was reviewed and the knowledge of the partners was summarised. Based on the conclusions and recommendations of this study, promising analytical methods were chosen for the practical tests.

4.2 Subtask III.2.2 “Sample homogenisation and homogeneity tests”

An adequate supply of homogeneous material was needed for the tests planned within Subtask III.2.3 and III.2.4. Pre-tests were performed in order to determine the particle size to which the final materials should be reduced in order to obtain satisfactory homogeneity. XRF and ICP-MS determinations of major and minor elements, respectively, were performed on three different particle sizes (<1 mm, <0.25 mm, <0.1 mm) and the relative standard deviations (RSD) calculated. The results showed that particle sizes <0.25 mm delivered satisfactory homogeneity ranges with RSD-values below 10% for most of the elements tested. The results furthermore revealed that special care has to be taken to avoid contamination during size reduction. Intensive milling led to introduction of impurities, probably caused by abrasion from the inner parts of the mills used. A detailed description of these investigations



is given in [6]. The next step of the practical work dealt with the preparation of sufficient wood+bark and straw materials for the further investigations planned in Task III.2. 10 kg wood+bark and straw material were milled to the agreed particle size (<0.25 mm) and filled in bottles of 20 g each. In addition, for Task II.2 and Task III.1 of the BioNorm project, 150 kg and 100 kg of each material (grinded to a particle size <1 mm), respectively, were also prepared.

In order to investigate the homogeneity of the prepared materials, XRF and ICP-MS determinations were performed once again. Subsequently, F-tests were carried out to check the within and between bottle homogeneity (see Table 1). Satisfactory RSD-values below 10% and 30% were obtained for most of the major and minor elements determined. Furthermore, the vast majority of the elements did not show significant F-values, indicating homogeneous distributions of the materials throughout the different bottles, with the exception of Zn in wood+bark as well as Fe and Cd in straw. To summarise the outcome of the homogeneity tests, both materials proved to be suitable for the further investigations planned within Task III.2. In order to verify whether a particle size <1 mm would also yield satisfactory homogeneities additional wood and straw materials with particle sizes <1 mm were prepared and investigated. The respective investigations revealed that particle sizes <1 mm also yielded satisfactory results [6].

The experiences gained from sample preparation together with the results of the homogeneity studies and relevant literature sources were summarised in the report “Best practise guidelines for the preparation of analysis samples” [7]. These guidelines are based on the Technical specification “Solid Biofuels – Methods for sample preparation” developed in CEN/ TC 335.



Table 1: Homogeneity tests for major and minor elements in wood+bark and straw

Explanations: major and minor elements were measured by XRF and ICP-MS, respectively; method precision was verified on the basis of 9 to 12 repeated measurements of 1 measurement portion; within homogeneity was verified by 9 to 12 independent determinations of the content of one bottle; between homogeneity was verified on the basis of 9 to 12 independent determinations from 9 to 12 bottles; RSDM, RSDW, RSDB…relative standard deviations (in %) of method precision, within and between bottles; URSD… respective uncertainties in % (defined as U ≈ RSD/√2n, where n is the number of replicates; calculation according to [8]); F-values…test statistics of the F-tests performed to check the within- and between-bottle homogeneity; *)…significant F-value (p ≤ 0.01)

Wood+barkRSDM URSDM RSDW URSDW RSDB URSDB

Major elementsAl 2.4 0.5 9.0 2.1 15.9 3.6 3.18Ca 2.9 0.6 3.2 0.7 1.8 0.4 3.42Fe 2.0 0.4 5.7 1.3 4.0 0.9 2.28K 2.3 0.5 1.0 0.2 0.9 0.2 1.40Mg 2.6 0.5 2.5 0.5 3.2 0.7 1.67Mn 1.9 0.4 1.7 0.4 1.0 0.2 3.34Na 4.2 0.9 4.5 1.0 2.4 0.5 3.43P 2.9 0.6 4.3 1.0 2.5 0.6 2.68Si 1.5 0.3 6.9 1.5 13.7 3.1 3.61Ti 1.4 0.3 1.5 0.3 1.1 0.3 1.67

Minor elementsBa 0.4 0.1 1.6 0.3 1.6 0.3 1.05 Cd 2.1 0.5 3.3 0.7 3.7 0.8 1.13 Co 1.8 0.4 3.6 0.7 4.8 1.0 1.79 Cr 6.8 1.5 16.1 3.4 14.1 3.3 1.65 Cu 1.8 0.4 9.3 2.0 14.7 3.3 3.48 Mo 12.4 2.8 38.9 8.0 44.1 9.0 1.00 Ni 1.6 0.3 19.8 4.0 29.0 5.9 2.15 Pb 1.9 0.4 26.9 5.7 50.6 10.3 4.48 Sb 1.6 0.4 18.6 4.0 27.1 5.8 2.53 Tl 2.3 0.5 2.3 0.5 3.7 0.8 2.64V 2.5 0.6 23.3 4.8 21.3 4.3 1.12 Zn 0.7 0.2 3.1 0.7 8.9 1.9 7.77 *)

StrawRSDM URSDM RSDW URSDW RSDB URSDB

Major ElementsAl 1.4 0.3 5.0 1.1 5.7 1.3 1.09Ca 1.0 0.2 2.4 0.5 2.0 0.5 1.54Fe 1.1 0.2 7.0 1.6 2.1 0.5 13.32 *)

K 1.5 0.3 2.7 0.6 2.3 0.5 1.55Mg 1.6 0.3 3.5 0.8 2.6 0.6 2.00Mn 0.9 0.2 1.1 0.2 1.7 0.4 2.37Na 3.9 0.8 6.5 1.5 3.3 0.7 3.94P 1.8 0.4 3.2 0.7 2.2 0.5 1.86Si 1.1 0.2 3.6 0.8 2.6 0.6 1.84Ti 1.3 0.3 2.0 0.4 1.7 0.4 1.37

Minor ElementsBa 0.5 0.1 1.7 0.4 1.0 0.2 3.26 Cd 1.7 0.4 4.0 0.8 1.8 0.4 4.96 *)

Co 1.5 0.3 6.4 1.3 4.5 0.9 1.83 Cr 11.1 2.5 15.8 3.2 13.3 2.8 1.88 Cu 2.1 0.5 3.8 0.8 6.6 1.3 2.29 Mo 5.4 1.2 6.7 1.4 7.5 1.5 1.26 Ni 1.9 0.4 45.4 9.3 27.8 5.7 1.54 Pb 1.1 0.3 10.8 2.4 12.1 2.5 1.11 Sb 3.9 0.9 17.8 3.6 8.1 1.6 4.37 Tl 1.5 0.3 2.4 0.5 3.6 0.7 1.48V 1.8 0.4 5.0 1.0 7.2 1.5 2.10Zn 0.8 0.2 7.1 1.5 7.0 1.4 1.09

Between Homogeneity F-Value

F-Value Between HomogeneityWithin HomogeneityMethod Precision

Within HomogeneityMethod Precision



Working group 3 [9] and discuss assets and drawbacks of different sample preparation methods in terms of technical issues and feasibility. Emphasis is laid on the discussion of suitable equipment for size and mass reduction. In special notes practical examples for the application of the guidelines are described.

4.3 Subtask III.2.3 “Method testing and improvement – part 1”

Subtask III.2.3 - part 1 was concerned with the testing of several suitable analytical methods and the choice of the most advisable methods for further investigations within Subtask III.2.3 - part 2. For this purpose 16 digestion and 8 determination methods were investigated (see Table 2). The main results can be summarised as follows.

• Wet digestion in closed vessels with a mixture of HNO3, (H2O2) and HF, HCl, H2O or HClO4, heated conventionally or by microwaves, with or without neutralisation (complexation) by H3BO3 (digests No. 1-9, see Table 2), followed by determination with FAAS, GFAAS, CVAAS, ICP-OES or ICP-MS, showed, for most of the elements, a good conformity between the results gained. However, for Si and Ti, large fluctuations between the measured concentrations were observed. The results of the Na, P, Si, Cr and Zn determinations are shown in Figure 1 as examples (a compendium of all results is given in the Annex.) Lower concentrations of Si and Ti were especially stated when the digestions have been performed without addition of HF. This suggests the requirement of this acid for digestion in case Si and Ti are foreseen for determination.

• Pre-ashing at 550°C, followed by Li-metborate fusion, dissolving in H2O/HNO3 (digest No. 10, see Table 2; applied for major element determination according to ASTM D3682) or pre-ashing (550°C) and subsequent ash-digestion with HNO3/HF/H3BO3 (digest No.11, see Table 2; applied for major and (some) minor element determinations according to ASTM D3683) resulted in good conformity with the above mentioned wet digestion procedures (see Figure 1). Because of restricted time and budget within the project working group, however, dry ashing was not included in the following investigations within Task III.2.

• No difference with respect to the applied heating device (conventional (resistance) or microwave heating) was observed.

• Digest No. 3 (see Table 2) which, among others, prescribes filtration through filter paper after decomposition, showed, for several elements, strong deviations from other results (see as examples the results for P and Zn in Figure 1).

• Results obtained for K, Mg, Na and P further revealed that direct XRF-measurements (performed on wood and straw pellets; digest No. 16, see Table 2) deviated from results gained with the other measurement systems. Similar observations were made for Al, Fe, Na, P, Cd, Cr, Ni, Co, Mn, Sb when XRF-measurements were applied after pre-ashing and fusion in Li-tetraborate (digests No. 13-15, see Table 2). As examples of these observations, see the results for Na, P and Cr in Figure 1 These deviations suggest calibration problems for the applied XRF-systems. Therefore, this detection system was not included in the following investigations within Task III.2.



Table 4-1 Participating laboratories and applied detection and digestion methods for the investigations within Subtask III.2.3 “Method testing and improvement – part 1”

Code Laboratory A AC-TUG D IEC-DPI E ELSAM I IEC N NYME S SP U USTUTT V VT-TUG Dection systems applied

C CVAAS D Direct Hg determination E FAES F FAAS G GFAAS M ICP-MS O ICP-OES X XRF Digestion methods applied

1 H2O2/HNO3 (microwave heating) 2 H2O2/HNO3/HClO4 (microwave heating) /HF 3 H2O2/HNO3 (conventional heating, ½ hour cooking, filtration through filter paper) 4 H2O2/HNO3/HCl (microwave heating) 5 H2O2/HNO3/HClO4 (conventional heating) /HF 6 HNO3/HF (microwave heating) 7 H2O2/HNO3/HF (conventional heating) /H3BO3 8 H2O2/HNO3/HF (microwave heating) /H3BO3 9 HNO3/HF (microwave heating) /H3BO3

10 Preashing at 550 °C, Li2BO2 (1050 °C), HNO3/H2O (ASTM D 3682) 11 Preashing at 550 °C, HNO3/HF/H3BO3 (ASTM D 3682) 12 Direct Hg determination 13 Preashing at 450 °C/Li2B4O7/NH4NO3 (only straw was measured) 14 Preashing at 525 °C/Li2B4O7/NH4NO3 (only straw was measured) 15 Preashing at 815 °C/Li2B4O7/NH4NO3 (only straw was measured) 16 Pellets (direct measurement of palletised material)

4.4 Subtask III.2.3 “Method testing and improvement – part 2”

Based on the outcomes of Subtask III.2.3. - part 1, it was decided to focus further work on the optimisation of a wet chemical digestion procedure. The amount of HF required as well as different digestion temperatures (105°C, 190°C and 220°C) were investigated. In order to assure a broader applicability of the analytical methods under investigation, olive residues as well as a certified biomass reference material (GBW 07602 (NCS DC 73348) “Busch Branches and Leaves”) were included in the examinations in addition to straw. Because of restricted time and budget within the project working group, wood+bark was excluded from these examinations.



Wood+bark Straw

Nam

g/kg

d.b

.Grand Mean 57.2

AO2

AO5

DO

8

EO7

EO8

IF4

IF8

SO

10

SO

11

UX1

6

VM6

VM9

0

200

400

600

800

1000

1200

P

mg/

kg d

.b.

Grand Mean 86.6

AO

2

AO

5

DO

4

DO

8

EO

7

EO

8

SO

10

SO

11

UX

16

VM

6

VM

9

59

69

79

89

99

109

Na

mg/

kg d

.b.

Grand Mean 186

AO

2A

O5

EO

7E

O8

IF4

IF8

IX13

IX14

IX15

NO

3SO

10SO

11U

X16

VF6

VF9

VM6

VM9

0

300

600

900

1200

1500

1800

P

mg/

kg d

.b.

Grand Mean 1310

AO2

AO5

DO

4D

O8

EO7

EO8

IX13

IX14

IX15

NO

3SO

10SO

11U

X16

VM6

VM9

840

1040

1240

1440

1640

Si

mg/

kg d

.b.

Grand Mean 491

AO

2

DG

4

EO

7

EO

8

IF4

IF8

NO

3

SO

10

SO

11

UX1

6

0

200

400

600

800

1000

1200

Si

mg/

kg d

.b.

Grand Mean 14500

AO

2D

G4

DG

8D

O4

DO

8E

O7

EO

8IF

4IF

8IX

13IX

14IX

15N

O3

SO

10S

O11

UX1

6V

F6V

F90

4

8

12

16

20

24(X 1000)

Cr

mg/

kg d

.b.

Grand Mean 1.83

AO

2A

O5

DG

4D

G8

DM

4D

M8

DO

4D

O8

EO

1E

O7

EO

8IF

4IF

8S

M1

SO11

VM

6V

M9

0

1

2

3

4

Cr

mg/

kg d

.b.

DG

4D

G8

DM

4D

M8

DO

4D

O8

EO

1E

O7

EO

8IF

4IF

8IX

13IX

14IX

15S

M1

SO

11V

M6

VM

9

0

3

6

9

12

15

18Grand Mean 1.77

Zn

mg/

kg d

.b.

Grand Mean 22.3

AO2

AO5

DM

4D

M8

DO

4D

O8

EO1

EO7

EO8

IF4

IF8

NO

3SM

1S

O11

VM6

VM9

15

25

35

45

55

Zn

mg/

kg d

.b.

Grand Mean 22.4

AO2

AO5

DM

4D

M8

DO

4D

O8

EO1

EO7

EO8

IF4

IF8

IX13

IX14

IX15

NO

3SM

1SO

11VM

6VM

9

0

20

40

60

80

Figure 1: Results of Na, P, Si, Cr and Zn determinations within Subtask III.2.3 – part 1

Explanations: the boxplots have been calculated from n=3-6 independent determinations; the codes on the x axes refer to the participating laboratories (1st character) as well as the detection systems (2nd character) and digestion methods (digit) applied (see Table 2)



The following investigations were performed within Subtask III.2.3 – part 2:

• Digestion tests on straw and olive residues with H2O2, HNO3 and varying amounts of HF at 190°C and 220°C (105°C by partner NYME).

• Analyses of a certified reference material (GBW 07602 (NCS DC 73348) „Bush Branches and Leaves“) digested with varying amounts of HF as well as standard digestion procedures of the participating laboratories.

• Special digestion tests on olive residues by partner ELSAM (performed with varying amounts of H2O2, HNO3 and HF at 190°C and 220°C, and with or without subsequent heating of the digest solution with H3BO3).

The results obtained can be summarised as follows:

• For most of the elements investigated a good agreement between the results of most laboratories can be stated. Furthermore, most of the elements (Fe, P, As, Cd, Cu, Mo, (Ni), Pb, Sb, Tl, Zn) showed no systematic trends with respect to the applied HF quantity at digestion temperatures of 190°C and 220°C. As examples for these observations the results for Fe and Pb are shown in Figure 2 (a compendium of all results is given in the Annex).

• Systematic trends could be observed for Al, Na, Si, Ti, Ba and Cr (as examples for these observations the results for Al, Si and Cr are shown in Figure 3). Digestions without HF showed lower concentrations than digestions with HF for

∗ Al in straw and the reference material,

∗ Na in straw and olive residues,

∗ Si and Ti in straw, olive residues and the reference material,

∗ Ba in olive residues and the reference material and

∗ Cr in olive residues.

In addition, the results of some laboratories revealed a tendency towards higher concentration with rising amounts of HF for

∗ Si in olive residues and the reference material,

∗ Ti in straw and

∗ Cr in olive residues,

with the highest mean values usually shown when 80 µl HF per 100 mg sample had been used for digestion. For Si and Ti the same conclusions can be drawn from the findings of partner ELSAM who performed special investigations on olive residues (Figure 4). These results suggest the necessity of HF for digestion.

• Concerning olive residues in particular, some laboratories found decreasing concentrations with rising amounts of HF for some major elements like

∗ Ca (partner DPI and SP),

∗ K (partner DPI, IEC, SP and LH) and



∗ Al (partner AC-TUG, LH and partner SP; the latter found decreasing concentrations after an initial concentration-increase from the digestions without HF).

As examples for these observations the results for K are shown in Figure 2 and for Al in Figure 3. These results indicate the formation of insoluble fluorides (e.g. CaF2) caused by an excess of HF. Such precipitations can be brought into solution by heating with H3BO3 (formation of stable, soluble complexes). The assumption of fluoride formation is supported by the findings of laboratory E, which found in olive residues decreasing values of Al and Mg with rising amounts of HF in case the HF excess has not been neutralised with H3BO3 (see Figure 4). Important for neutralisation is heating of the solution. Addition of H3BO3 without heating was not effective. These results suggest the necessity of the usage of H3BO3 for the determination of major elements.

• In addition to the already mentioned results the special digestion tests on olive residues performed by partner ELSAM (see Figure 4) revealed no difference with respect to the applied temperatures (190°C, 220°C). No clear difference could also be stated with respect to the different H2O2 and HNO3 quantities applied.

• The investigations also revealed for many elements that the results of the laboratory, which performed the digestions at 105°C, deviated from results of the other laboratories, where digestions with the same amount of HF, but at higher temperatures, have been performed. As example of this trend see the results for Al in Figure 4.

4.5 Subtask III.2.4 “Method comparison and validation”

Based on the results of the method development investigations (see sections 4.3 and 4.4), it was decided to validate different digestion methods for major and minor elements (see Table and Table ). The detection systems applied for validation included FAAS, GFAAS, ICP-OES, ICP-MS, CVAAS and direct Hg determination. The validation parameters examined included accuracy (precision and trueness) and application ranges as well as estimations of detection limits. A detailed description of these investigations together with the statistics applied is given in [10].

4.5.1 Definitions, purpose and calculation of validation parameters

Accuracy (precision and trueness) Accuracy expresses the closeness of a result to a true value. It is normally studied as two components: precision and trueness [11, 12]. Precision is a measure of how close results are to one another. Important precision parameters are: overall mean Xrb, repeatability limit r and the reproducibility limit R (see Table).

The repeatability limit r enables the analyst to decide, whether the difference between duplicate analyses of a sample, determined under repeatability conditions (applying the same method on identical test material in the same laboratory by the same operator using the same equipment within short intervals of time), is significant. A further application of r includes the calculation of a critical difference between mean values of two series of investigations performed under repeatability conditions [11, 12, 13, 14].



The reproducibility limit R enables the analyst to decide whether the difference between duplicate analyses of a sample, determined under reproducibility conditions (applying the same method on identical test material in different laboratories with different operators using different equipment), is significant. Further applications of R and r include the calculation of critical differences for the following cases: comparison of means under reproducibility conditions, comparison of a laboratory mean with a reference value as well as comparison of several laboratory means with a reference value [11, 12, 13, 14].

The following investigations were performed in order to determine the accuracy measures mentioned above. Three independent digestions of reference materials (GBW 07602 (NCS DC 73348) “Bush Branches and Leaves” or NIST SRM 1575a “Trace elements in Pine Needles”) and 6 independent digestions of wood+bark and straw (no digestions were carried out for direct Hg determination) were measured by each laboratory (interlaboratory study). Subsequently, Xrb (overall robust mean), r and R were calculated using the robust method given in [14]. This method does not require exclusion of outliers and is gaining increasing acceptance [15].

The trueness (of a method) is defined as ‘the closeness of agreement between the average value obtained from a large set of test results and an accepted reference value’ [12]. For this purpose critical differences between the found (overall robust means) and certified element concentrations of the investigated reference materials (GBW 07602 (NCS DC 73348) “Bush Branches and Leaves”, or NIST SRM 1575a “Trace elements in Pine Needles”) were calculated according to [12, 14].

The calculated precision and trueness parameters are shown in Table 4.

Detection limits and application ranges of methods tested When measurements are made at low analyte levels it is important to know the lowest concentration of the analyte that can be confidently detected by the method (detection limit). According to [11], it is usually sufficient for validation purposes to provide an indication of the level at which detection becomes problematic. Detection limits vary between different brands and instruments. Furthermore, it has to be considered that the detection limits do not represent levels at which quantitation is not possible. ‘It is simply that the size of the associated uncertainties approach comparability with the actual result in the region of the detection limit’ [11]. For critical decisions, the relevant values need to be re-determined in line with actual operating performance [11]. In the present study, the detection limit is defined as three times the standard deviation of the blank (when the analytical work is in support of regulatory or specification compliance, a more exact approach such as that described by [13] is more appropriate). For this purpose, the entire analysis consisting of acid digestion, dilution and instrument run was done in exactly the same way as for the accuracy measurements, with the only exception that no sample material was added to the digestion mixture. 20 blank determinations were performed using this procedure. The average value of two measurement results was used to state one determination value, which corresponds with 10 double determinations. These 10 values were applied for the calculation of the detection limit.

The purpose of the application range investigations was to clarify the range of the sample element content up to which the applied digestion and determination methods can be used. This range should cover element concentrations to be expected for different types of solid biofuels as well as limit values specified in standards relevant for biofuels. Typical



concentration ranges for biofuels are given in [16], limit values for some minor elements in wood pellets are given in [17]. As materials other than wood+bark and straw were not available, different sample concentrations were simulated by the addition of defined amounts of the elements of interest (spiking). With the exception of Hg, the elements were added to the final solutions of the wood+bark and straw digests. In the case of Hg determinations, the spikes were added to the wood+bark and straw samples, either prior to measurements (in case of direct Hg determination) or prior to digestion (in case CVAAS was used for Hg determination).

Depending on the determination system applied, 3 (GFAAS), 5 (FAAS) or 6 (ICP-OES, ICP-MS, CVAAS, direct Hg determination) element additions were performed and the concentrations of the original and the spiked solutions measured in duplicate (for direct Hg determinations, the original and spiked wood+bark and straw samples were measured). Subsequently, the linearity between measured and added concentrations was checked by the Mandel test as described in [18]. For this purpose, linear and quadratic regression lines were calculated. In case of significant deviations from linearity (significantly better fitting of quadratic regression), the concentration range was constricted and the evaluation repeated until linearity was confirmed by a non-significant test value.

Table 6 and Table 7 summarise the results of the detection limit and application range investigations. They provide an overview of the detection limits to be expected and the concentration ranges to be covered with the digestion methods and detection systems given.

Table 3: Validated digestion method for major elements

Explanations: 1)…temperature referred to digestion solution; 2)…temperature referred to heating device (e.g. oven); sample masses applied: 250 to 500 mg (d.b.)

Acid quantities (per 100 mg sample (d.b.)): 0.6 ml H2O2 (30%), 1.6 ml HNO3 (65%), 0.2 ml HF (40%)

Temperature programme: Microwave heating1)

Step 1: Step 2:

in 15 min heat to 190°C, rate 11.3°C/min

hold for 20 min at 190°C

Resistance heating2) Step 1: Step 2:

in 1 h heat to 220°C, rate 3.33°C/min

After cooling to room temperature, addition of 2 ml H3BO3 (4%) (= 10 ml H3BO3 / ml HF)

per 100 mg sample (d.b.): Temperature programme:

Microwave heating1) Step 1: Step 2:

heat as fast as possible up to 150°C

hold for 15 min at 150°C

Resistance heating2) Step 1:

heat as fast as possible up to 180°C



Table 4: Validated digestion method for minor elements

Explanations: 1)…temperature referred to digestion solution; 2)…temperature referred to heating device (e.g. oven); sample masses applied: 250 to 500 mg (d.b.)

Acid quantities (per 100 mg sample (d.b.)): 0.5 ml H2O2 (30%), 1 ml HNO3 (65%), 0.08 ml HF (40%)

Temperature programme: Microwave heating1)

Step 1: Step 2:

in 15 min heat to 190 °C, rate 11.3°C /min hold for 20 min at 190°C

Resistance heating2) Step 1: Step 2:

in 1 h heat to 220°C, rate 3.33°C/min

hold for 1 h at 220°C

Part 2 Part 2.3 Chemical Tests


Fe-Straw

200

250

300

350

400

450

HF 20 40 80 80 80-

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 220 220 105 220

Det M O F O O O O O

Lab V D I E S A N L

mg/

kg d

.b.

Fe-Olive Residues

1300

1400

1500

1600

1700

1800

1900

2000

HF 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 220 220 105 220

Det M O F O O O O O

Lab V D I E S A N L

mg/

kg d

.b.

Fe-Reference Material

600

700

800

900

1000

1100

1200

1300

HF 80 80 64 400 80 200 80 100 0 20 40 80 20 80 80

- # * - # # # * -

°C 220 190 190 210 220 190 220 220 105

Det M O F O O O O

Lab V D I E S A N

mg/

kg d

.b.

CertifiedValue

Pb-Reference Material

4

5

6

7

8

9

10

11

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M G G M M O

Lab V D E S A N

mg/

kg d

.b.

CertifiedValue

Pb-Straw

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

HF 20 40 80 80 80-

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 220 105 220

Det M G G M M O G

Lab V D E S A N L

mg/

kg d

.b.

Pb-Olive Residues

0

2

4

6

8

10

12

14

HF 20 40 80 80 80 -

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M G G M M O G

Lab V D E S A N L

mg/

kg d

.b.

K-Straw

9000

10000

11000

12000

13000

14000

15000

16000

17000

18000

HF 20 40 80 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 190 220 190 190 220 190 220 220 105 220

Det M F O F O O O O O

Lab V V D I E S A N L

mg/

kg d

.b.

K-Olive Residues

21000

22000

23000

24000

25000

26000

27000

28000

29000

30000

31000

HF 20 40 80 80 80 -

20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 80 20 80

°C 190 220 190 220 190 190 220 190 220 105 220

Det M F O F O O O O

Lab V V D I E S N L

mg/

kg d

.b.

K-Reference Material

6000

6500

7000

7500

8000

8500

9000

9500

10000

HF 80 80 80 80 64 400 80 200 80 100 0 20 40 80 20 80 80

- - # * - # # # * -

°C 220 220 190 190 210 220 190 220 220 105

Det M F O F O O O O

Lab V V D I E S A N

mg/

kg d

.b.

CertifiedValue

Figure 2: Results of Fe, Pb and K determinations within Subtask III.2.3 – part 2

Explanations: digestions were performed with 0.5 ml H2O2 / 1 ml HNO3 (acid quantities are referred to 100 mg sample (d.b.)) and varying amounts of HF (µl) for straw and olive residues and with varying amounts of H2O2 / HNO3 and HF (the HF quantities (µl) stated are referred to 100 mg sample (d.b.)) for reference material GBW 07602 (NCS DC 73348) at different temperatures; the solid line refers to the mean value, the dashed line defines the 95% confidence interval; HF...hydrofluoric acid; °C...temperature in degree Celsius; Det...detection system (F...FAAS or FAES, G...GFAAS, M...ICP-MS, O...ICP-OES); Lab... laboratory (V...VT-TUG, D...DPI, I...IEC, E...ELSAM, S...SP, A...AC-TUG, N...NYME, L...LH, U...USTUTT); #...neutralisation (complexation) with H3BO3; -...digestion without H2O2; *...digestion with addition of HCl; the depicted means plots have been calculated from n=2-6 independent determinations

Part 2 Part 2.3 Chemical Tests


Al-Straw

0

50

100

150

200

250

300

350

400

450

500

HF 20 40 80 80 80-

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 220 220 105 220

Det M O F O O O O O

Lab V D I E S A N L

mg/

kg d

.b.

Al-Olive Residues

4821500

1700

1900

2100

2300

2500

2700

2900

3100

HF 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 220 220 105 220

Det M O F O O O O O

Lab V D I E S A N L

mg/

kg d

.b.

Al-Reference Material

4301000

1200

1400

1600

1800

2000

2200

2400

2600

2800

3000

HF 80 80 64 400 80 200 80 100 0 20 40 80 20 80 80

- # * - # # # * -

°C 220 190 190 210 220 190 220 220 105

Det M O F O O O O

Lab V D I E S A N

mg/

kg d

.b.

CertifiedValue

Si-Reference Material

1482000

2500

3000

3500

4000

4500

5000

5500

6000

6500

7000

HF 80 80 64 400 80 200 80 100 0 20 40 80 20 80

- # * - # # # * -

°C 220 190 190 210 220 190 220 220

Det F O F O O O

Lab V D I E S A

mg/

kg d

.b. Certified

Value

Si-Straw

0

5000

10000

15000

20000

25000

30000

HF 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 20 80

°C 190 220 190 190 220 190 220 220 220

Det F O F O O O O

Lab V D I E S A L

mg/

kg d

.b.

Si-Olive Residues

0

2000

4000

6000

8000

10000

12000

14000

HF 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 20 80

°C 190 220 190 190 220 190 220 220 220

Det F O F O O O O

Lab V D I E S A L

mg/

kg d

.b.

Cr-Reference Material

0

1

2

3

4

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M O O M O O

Lab V D E S A N

mg/

kg d

.b.

CertifiedValue

Cr-Straw

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

HF 20 40 80 80 80-

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 220 105 220

Det M O O M O O G

Lab V D E S A N L

mg/

kg d

.b.

Cr-Olive Residues

10

12

14

16

18

20

22

24

HF 20 40 80 80 80 -

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M O O M O O G

Lab V D E S A N L

mg/

kg d

.b.

Figure 3: Results of Al, Si and Cr determinations within Subtask III.2.3 – part 2

Explanations: digestions were performed with 0.5 ml H2O2 / 1 ml HNO3 (acid quantities are referred to 100 mg sample (d.b.)) and varying amounts of HF (µl) for straw and olive residues and with varying amounts of H2O2 / HNO3 and HF (the HF quantities (µl) stated are referred to 100 mg sample (d.b.)) for reference material GBW 07602 (NCS DC 73348) at different temperatures; the solid line refers to the mean value, the dashed line defines the 95% confidence interval; HF...hydrofluoric acid; °C...temperature in degree Celsius; Det...detection system (F...FAAS or FAES, G...GFAAS, M...ICP-MS, O...ICP-OES); Lab... laboratory (V...VT-TUG, D...DPI, I...IEC, E...ELSAM, S...SP, A...AC-TUG, N...NYME, L...LH, U...USTUTT); #...neutralisation (complexation) with H3BO3; -...digestion without H2O2; *...digestion with addition of HCl; the depicted means plots have been calculated from n=2-6 independent determinations



190°C, 0.5 ml H2O2, 1 ml HNO3, 1 ml H3BO3 without heating

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6

Al Ca Fe K Mg Na P Si Ti Ba Mn Ni

rela

tive

conc

entr

atio

n

20 µl HF

40 µl HF

80 µl HF

100 µl HF

220°C, 0.5 ml H2O2, 1 ml HNO3, 1 ml H3BO3 without heating

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6


rela

tive

conc

entr

atio

n

20 µl HF

40 µl HF

80 µl HF

100 µl HF

190°C, 0.8 ml H2O2, 1 .5 ml HNO3, 1 ml H3BO3 without heating

0.60.70.80.91.01.11.21.31.41.51.6


rela

tive

conc

entra

tion

20 µl HF

40 µl HF

80 µl HF

100 µl HF

220°C, 0.8 ml H2O2, 1.5 ml HNO3, 1 ml H3BO3 without heating

0.60.70.80.91.01.11.21.31.41.51.6


rela

tive

conc

entr

atio

n

20 µl HF

40 µl HF

80 µl HF

100 µl HF

220°C, 0.8 ml H2O2, 1.5 ml HNO3, 1 ml H3BO3 with heating

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6


rela

tive

conc

entra

tion

20 µl HF

40 µl HF

80 µl HF

100 µl HF

Figure 4: Digestion tests on olive residues by partner ELSAM

Explanations: digestions have been performed in a microwave system with varying amounts of H2O2, HNO3 and HF (acid quantities referred to 100 mg sample (d.b.)) at 190°C and 220°C and with or without heating of digest to 150°C with H3BO3; relative concentrations…concentrations normalised to 1 for digestion with 0.8 ml H2O2, 1.5 ml HNO3 at 220°C and subsequent heating with 1 ml H3BO3; the values have been calculated from n=2 independent determinations



Table 5: Precision data for major and minor elements in wood+bark, straw and reference materials

Explanations: Det. System…determination system; F...FAAS, O...ICP-OES, G...GFAAS, M...ICP-MS, C...CVAAS, dHg...direct Hg determination; 1)...except for Hg where NIST SRM 1575a was analysed; 2)...except for wood+bark where FAAS was not applied; Xrb…overall robust mean; r…repeatability limit; R…reproducibility limit; N…number of participating laboratories; Cert. conc….certified concentration; Crit. diff…critical difference between certified concentration and Xrb; Meas. diff…measured difference between certified concentration and Xrb; measurement values from n=6 independent digestions of wood+bark and straw or n=3 independent digestion of reference materials per laboratory (no digestions were carried out for direct Hg determination) were used to calculate r and R according to the robust method given in [14]

Wood+bark Straw GBW 07602 (NCS DC 73348)1) Ele-ment

N Det. System

Xrb [mg/kg d.b.]

r mg/kg d.b.]

R [mg/kg d.b.]

Xrb [mg/kg d.b.]

r [mg/kg d.b.]

R [mg/kg d.b.]

Xrb [mg/kg d.b.]

r [mg/kg d.b.]

R [mg/kg d.b.]

Cert. conc.[mg/kg d.b.]

Crit. diff.[mg/kg d.b.]

Meas. diff. [mg/kg d.b.]

Major elements Al 6 F2), O, M 83 6 28 323 18 48 2,220 80 380 2,140 99 80 Ca 7 F, O 1,780 90 190 5,650 270 580 22,400 800 2,000 22,200 494 200 Fe 7 F, O 85 14 14 314 15 32 1,020 30 80 1,020 21 0 K 7 F, O 710 27 178 12,700 500 2300 8,500 180 610 8,500 158 0

Mg 7 F, O 212 13 22 1,370 50 140 2,930 100 250 2,870 62 60 Na 7 F, O 17 7 11 109 12 23 11,000 - 1,200 11,000 315 0 P 6 O, M 87 9 14 1,330 60 170 838 35 58 830 15 8 Si 6 F, O 541 101 192 19,400 1,100 1,500 5,910 230 790 5,800 221 110 Ti 6 O, M 11 1 3 134 9 40 104 4 12 95 3 9 Minor elements

As 4 G, M 0.07 0.08 0.13 0.22 0.07 0.15 1.02 0.07 0.37 0.95 0.13 0.07 Ba 6 O, M 25 2 3 78 3 13 19.6 0.6 2.6 19.0 0.7 0.6 Cd 5 G, M 0.35 0.03 0.14 0.15 0.02 0.10 0.20 0.04 0.12 0.14 0.04 0.06 Co 4 G, M 0.42 0.16 0.16 0.22 0.05 0.53 0.34 0.05 0.27 0.39 0.10 0.05 Cr 9 O, G, M 2.0 0.8 1.0 1.2 0.3 0.8 2.24 0.42 0.61 2.30 0.12 0.06 Cu 9 O, G, M 2.1 0.9 1.6 4.1 0.8 1.3 5.1 0.6 1.2 5.2 0.3 0.1 Hg 3 C, dHg 0.007 0.003 0.005 0.027 0.006 0.013 0.042 0.005 0.010 0.040 0.004 0.002 Mn 9 F, O, M 298 13 75 43 2 6 58 2 6 58 1 0 Mo 4 G, M 0.09 0.14 0.14 0.37 0.06 0.08 0.304 0.031 0.102 0.260 0.035 0.044 Ni 9 O, G, M 1.7 0.6 0.7 0.7 0.2 0.4 1.7 0.2 0.9 1.7 0.2 0.0 Pb 5 G, M 1.0 0.4 0.6 1.3 0.3 0.5 7.1 0.8 1.5 7.1 0.4 0.0 Sb 4 G, M 0.03 0.12 0.13 0.07 0.03 0.08 0.17 0.02 0.45 0.08 0.16 0.09 Tl 3 G, M 0.01 - - 0.02 - - 0.02 - - - - - V 7 O, G, M 0.18 0.10 0.32 0.60 0.13 0.38 2.4 0.2 0.5 2.4 0.1 0.0 Zn 9 F, O, M 21 3 6 20 3 3 21.1 0.9 2.9 20.6 0.6 0.5



Table 6: Detection limits and application ranges for major elements

Explanations: the figures given were compiled as follows: 2 to 4 labs performed element determinations with one detection system; the resulting n=10 measurement values were used to calculate detection limits for each laboratory; these figures were compared; the highest of these figures (rounded up to the next highest significant digit) is stated in the table as “<value”; for example: four laboratories performed ICP-OES measurements for Al; the calculated detection limits are 4.9, 1.7, 26 and 3.1 mg/kg (d.b.); the highest figure is 26; therefore, the detection limit given for Al is <26; a similar approach was used in stating the application ranges; the results of the different labs were compared and the figures rounded to the next reasonable digit

Major elements

Determination systems

Detection limits [mg/kg]

Application ranges studied [mg/kg]

Al FAAS <23 300 - 2,000

ICP-OES <26 100 - 3,000

Ca FAAS <20 2,000 - 40,000

ICP-OES <4 2,000 - 20,000

Fe FAAS <3 100 - 4,000

ICP-OES <16 100 - 2,000

K FAAS <10 1,000 - 50,000

ICP-OES <75 1,000 - 30,000

Mg FAAS <6 200 - 10,000

ICP-OES <20 200 - 4,000

Na FAAS <1 20 - 1,000

ICP-OES <30 20 - 3,000

P ICP-OES <10 100 - 7,000

ICP-MS <1 100 - 4,000

Si FAAS <60 500 - 150,000

ICP-OES <240 500 - 40,000

Ti ICP-OES <1 10 - 500



Table 7: Detection limits and application ranges for minor elements

Explanations: see Table 6

Minor elements Determination systems Detection limits

[mg/kg] Application ranges studied

[mg/kg] As GFAAS <0.2 0.2 - 5

ICP-MS <0.02 0.1 - 5 Ba ICP-OES <0.3 25 - 100

ICP-MS <0.2 25 - 100 Cd GFAAS <0.08 0.1 - 2

ICP-MS <0.02 0.1 - 5 Co GFAAS <0.4 0.5 20

ICP-MS <0.02 0.2 - 20 Cr ICP-OES <1 2 - 50 GFAAS <0.4 2 - 50 ICP-MS <0.03 2 - 50

Cu ICP-OES <0.8 2 - 40 GFAAS <0.3 2 - 40 ICP-MS <0.1 2 - 40

Hg Direct Hg determin. <0.0006 0.01 - 0.3 CVAAS <0.002 0.03 - 0.5

Mn FAAS <3 50 - 2,500 ICP-OES <0.2 50 - 800 ICP-MS <0.02 50 - 500

Mo GFAAS <0.4 0.1 - 20 ICP-MS <0.015 0.1 - 20

Ni ICP-OES <0.8 1 - 20 GFAAS <0.65 1 - 10 ICP-MS <0.4 1 - 20

Pb GFAAS <0.2 1 - 25 ICP-MS <0.04 1 - 25

Sb GFAAS <0.35 0.5 - 20 ICP-MS <0.02 0.1 - 20

Tl GFAAS <0.08 n.d. n.d. ICP-MS <0.003 0.01 - 20

V ICP-OES <0.35 0.2 - 20 GFAAS <0.40 1 - 10 ICP-MS <0.02 0.1 - 20

Zn FAAS <2 25 - 500 ICP-OES <0.4 20 - 500 ICP-MS <0.09 20 - 200



4.5.2 Results and conclusions from method validation

The results of the trueness examinations proved to be satisfactory (Table 5). With the exception of Ti, Cd and Mo, the concentrations (overall robust means) found in the reference materials (GBW 07602 (NCS DC 73348) “Busch Branches and Leaves” and NIST SRM 157a “Trace elements in Pine Needles”) correspond with the certified values (critical differences were not exceeded).

For the majority of the elements investigated (Al, Ca, Fe, K, Mg, Na, P, Si; Ti, Ba, Cr, Cu, Mn, Ni, Zn), the precision data obtained represent a realistic picture of measurement results which may be expected when different laboratories apply the tested methods for analyses of the respective elements in solid biofuels such as wood+bark, straw, bush branches and leaves (see Table 5). The defined application ranges indicate the suitability of the tested methods for a wide range of concentrations. The detection limits given represent estimations to be expected when the biofuels under consideration are analysed by means of the methods tested (see Table 6 and Table 7). Summarising these findings, the applied digestion methods (H2O2/HNO3/HF/ neutralisation with H3BO3 for major elements (Table 3); H2O2/HNO3/HF for minor elements (Table 4) as well as the applied determination systems (see Table 5) proved to be suitable for analyses of the above mentioned elements.

The precision data obtained for Cd, Pb and Hg must be considered in the light of fact that only 5 (Cd, Pb) or 3 (Hg) laboratories participated in these investigations (see Table 5). This is below the minimal number of 7 that is required for statistical evaluation [14]. Therefore, the data may not be used to calculate critical differences in order to compare analytical measurements. Nevertheless, the data well indicate the performance of the tested methods. The applied determination systems (GFAAS, ICP-MS for Cd and Pb determination; CVAAS and direct measurements for Hg determination) provided comparable results for the materials investigated. The defined application ranges give indications of the concentration ranges, which may be covered by the analytical methods applied. The detection limits given are typical values to be expected when the biofuels under consideration are analysed with the methods tested (Table 7). Summarising these results, digestion by H2O2/HNO3/HF (see Table 4) as well as determination by GFAAS and ICP-MS may be recommended for Cd and Pb analyses, while CVAAS and direct determination may be recommended for Hg analyses.

The precision data for As, Co, Mo and Sb revealed that the R (sometimes also the r) values of As (in wood+bark), Co (in straw), Mo (in wood+bark) and Sb (in wood+bark, straw and GBW 07602 (NCS DC 73348)) exceeded the calculated mean concentrations of these elements in the respective materials (see Table 5). This and the fact that only 4 laboratories participated in the investigations make the data unsuitable for calculating critical differences in order to compare analytical measurements. One reason for the high repeatability (r) and reproducibility limits (R) is the low As, Co, Sb and Mo concentration in wood+bark and straw which makes determinations by GFAAS difficult. This is indicated by high dispersions of the GFAAS measurement results, sometimes non-linear application ranges or relatively high recovery rates as well as relatively poor GFAAS detection limits (Table 7). GFAAS determination is therefore not well suitable for materials containing low concentrations of the above mentioned elements. ICP-MS showed better performance providing smaller spreads of measurement results and lower detection limits. Summarising these results, the applied digestion method (H2O2/HNO3/HF; Table 4) and determination by ICP-MS may be



recommended for As, Co, Mo and Sb analyses. GFAAS determination is only suitable for materials with high concentrations of these elements. For As determinations, hydride generation AAS, offering better performance with respect to dispersion of measurement results and detection limits could be used alternatively.

The precision data for V show that the R value of wood+bark exceeded the calculated mean concentration (Table 5). V was analysed by ICP-OES, ICP-MS and GFAAS. In comparison to ICP-OES and ICP-MS, GFAAS analysis resulted in higher V concentrations, higher dispersion of the measurement results, smaller linear ranges, and, especially in comparison to ICP-MS a poor detection limit (Table 7). This might have caused the high R value found for wood+bark. GFAAS is therefore not well suitable for V determinations in low concentration ranges. Summarising the results, the applied digestion method (H2O2/HNO3/HF, see Table 4) as well as determination by ICP-OES and ICP-MS can be recommended for V analyses. Due to the high dispersion of measurement results and poor detection limits, GFAAS determination is not well suitable for V determinations in lower concentration ranges.

No precision data are given for Tl. This is due to the fact that only 3 laboratories provided measurement results for the wood+bark and straw materials investigated. This low number combined with strong differences between the results of the applied measurement systems (ICP-MS and GFAAS), produced unrealistic calculation data. The linear ranges and detection limits found may be used as an indicator for the applicability of the methods tested. With respect to ICP-MS determination the linear ranges assessed and the low detection limits obtained indicate the potential suitability of this detection system for Tl analyses in solid biofuels (Table 7). GFAAS determination led to non-linear application ranges and poor detection limits. This indicates that GFAAS is not suitable to determine Tl in solid biofuels with low concentrations of this element.

4.6 Subtask III.2.5 “Method evaluation and compilation of best practice guidelines”

This work was performed in close cooperation with Task III.1. and resulted in the report “Solid Biofuels - Best practice guidelines for the determination of S, Cl, N as well as major and minor elements” [6]. The guidelines involve recommendations for the preparation of analysis samples, and, in this context, especially the influence of the particle size on the homogeneity of the prepared samples. Furthermore, based on the results from Subtasks III.2.3 and III.2.4, suitable digestion and determination methods are presented and the parameters and factors, which may influence the analytical results, are discussed. In addition, in order to define the performance of the methods recommended, the validation parameters covering precision, application ranges and detection limits for the different elements are given.


Regarding the results of the work performed within Task III.2 of the BioNorm project, the following overall conclusions can be drawn:

Particle sizes <1 mm or <0.25 mm should provide satisfactory homogeneity for sample materials to be used for major and minor element analyses in wood or straw. However, during size reduction special attention must be taken to avoid contamination from the inner materials



of the used mills. These materials should be chosen depending on the elements to be determined. If, for instance, Cr and Ni have to be determined with a high accuracy at low levels, stainless steel materials should be avoided for the parts of the mill having contact with the sample, using for example tungsten carbide or titanium instead. Due to the higher abrasion rate the use of high-speed mills should generally be avoided.

In conclusion of the validation results, the applied digestion methods (H2O2/HNO3/HF/ neutralisation with H3BO3 for major elements, see Table 3 and H2O2/HNO3/HF for minor elements, see Table 4) may be recommended for solid biofuel analyses.

For the determination of the different elements in solid biofuels, the following detection systems may be recommended:

FAAS: Ca, Fe, K, Mg, Na, Si, Mn, Zn; because of relatively poor detection limits, Al determination by FAAS is not recommended for solid biofuels containing low concentrations of this elements.

ICP-OES: Al, Ca, Fe, K, Mg, Na, P, Si, Ti, Mn, Zn, Ba, Cr, Cu, Mn, Ni, V, Zn. The ICP-OES detection limit given for Al in Table 6 is in the same magnitude as the FAAS detection limit for Al. This figure resulted from one laboratory which found a relatively high detection limit (26 mg/kg (d.b.)), compared to the values found by the other laboratories (4.9, 1.7, 3.1 mg/kg (d.b.)).

GFAAS: Cd, Cr, Cu, Ni, Pb; because of relatively poor detection limits, As, Co, Mo, Sb and V determination by GFAAS is not recommended for solid biofuels containing low concentrations of these elements.

ICP-MS: P, Ti, As, Ba, Cd, Co, Cr, Cu, Mn, Mo, Ni, Pb, Sb, V, Zn

Direct Hg determination, CVAAS:

Hg

6 Recommendations

Size reduction to <1 mm or <0.25 mm can be recommended for major and minor element analyses of solid biofuels.

The digestion methods developed within Task III.2 may be recommended for major and minor element determinations in solid biofuels. The determination systems tested may be recommended provided the detection limits are suitable for the elements to be determined.

XRF detection would be a suitable and fast alternative method for the determination of several major and minor elements. The application of XRF systems for element determination



strongly depends on reliable calibration standards, unless highly sophisticated systems are used. Since such standards are not available for solid biofuels, the application of XRF systems can presently not be recommended for solid biofuel analyses. Therefore, the development of such standards for solid biofuels would be desirable.



7 References

1 Curvers A, Gigler JK. Characterization of biomass fuels: an inventory of standard procedures for the determination of biomass properties. ECN-C--96-032. Netherlands Energy Research Foundation ECN, editor. Petten, the Netherlands; 1996.

2 van der Drift A, Schakellaar D. National round robin of biomass fuels and ashes. Netherlands Energy Research Foundation ECN, editor. Petten, the Netherlands; 1997.

3 Obernberger I, Dahl J, Arich A. Biomass fuel and ash analysis. Report of the European Commission. European Commission DG XII, editor. Brussels, Belgium; 1998. ISBN 92-828-3257-0.

4 Kauter D, Lewandowski I, Puttkamer T. Determination of the chemical composition of solid biofuels - Need of standardised methods and results from a round robin test. In: Proc. 1st World Conf. “Biomass for Energy and Industry”, June 2000, Sevilla, Spain, James & James Ltd., editor. London, UK; 2001, p. 335-8. ISBN 1-902916-15-8.

5 Baernthaler G, Zischka M, Arich A, Obernberger I. Definition of the state-of-the-art for the determination of major and minor ash forming elements in solid biofuels. Report on Subtask III.2.1 of the EU funded BioNorm project (ENK6-CT-2001-00556). http://www.ie-leipzig.de/ BioNorm/Standardisation.htm; 2003.

6 Baernthaler G, Englisch M, Zischka M, Haraldsson C, Obernberger I. Solid Biofuels – Best practice guidelines for the determination of sulphur, chlorine and nitrogen as well as major and minor elements. Report on deliverables III.1.D3 and III.2.D5 of the EU funded BIONORM project (ENK6-CT-2001-00556). http://www.ie-leipzig.de/BioNorm/ Standardisation.htm; 2004.

7 Puttkamer T, Baernthaler G. Solid Biofuels - Determination of major and minor elements - Best practice guidelines for the preparation of analysis samples. Report on Deliverable III.2 D3 of the EU funded BIONORM project (ENK6-CT-2001-00556). http://www.energetik-leipzig.de/BioNorm/Standardisation.htm; 2004.

8 Kramer KJM, Dorten WS, Groenewoud H, Haan E, Kramer GN, Monteiro L, Muntau H, Quevauviller P. Collaborative study to improve the quality of rare earth element determinations in environmental matrices. Journal of Environmental Monitoring 1999; 1:83-9.

9 PrCEN/TS 14790. Solid Biofuels – Methods for sample preparation. European Committee for Standardization, editor. Brussels; 2004.

10 Baernthaler G, Zischka M, Haraldsson C, Obernberger I. Solid biofuels – Determination of major and minor elements – Results of sub-task III.2.5 “Method validation”. Report on deliverable III.2 D4 of the EU funded BioNorm project (ENK6-CT-2001-00556). http://www.energetik-leipzig.de/BioNorm/Standardisation.htm; 2004.

11 Eurachem. The fitness for purpose of analytical methods - A laboratory guide to method validation and related topics. LGC (Teddington) Ltd., editor. Middlesex, UK; 1998. ISBN 0-948926-12-0.



12 ISO 5725. Accuracy (trueness and precision) of measurement results - parts 1 – 6. International Organization for Standardization, editor. Geneva, Switzerland; 1994.

13 International Union of Pure and Applied Chemistry (IUPAC). Compendium on Analytical Nomenclature (The Orange Book), Web edition. http://www.iupac.org/publications/analytical_compendium/; 2002.

14 Walter E, Lischer P et al. Swiss Food Manual (Schweizerisches Lebensmittelbuch) - Chapter 60 - Statistik und Ringversuche. Eidg. Drucksachen und Materialzentrale, editor. Bern, Switzerland; 1989.

15 Bogdanow S. Harmonised methods of the international honey commission. Swiss Bee Research Centre, FAM, editor. Liebefeld, Bern, Switzerland; 2002.

16 CEN/TC 335 - WG2 N94, Final draft. Solid biofuels - Fuel specifications and classes. European Committee for standardization, editor. Brussels, Belgium; 2003.

17 DIN 51731. Testing of solid fuels - compressed untreated wood - requirements and testing. Deutsches Institut für Normung, editor. Berlin, Germany; 1996.

18 Gottwald W. Statistik für Anwender. Gruber U, Klein W, editors. Weinheim, Germany: Willey-VCH; 2000. ISBN 3-527-29780-4

8 Glossary

Chemical symbols and formulae

Al aluminium LiBO2 lithium metaborate As arsenic Li2B4O7 lithium tetraborate Ba barium Mg magnesium Ca calcium Mn manganese Cd cadmium Mo molybdenum Cl chlorine N nitrogen Co cobalt Na sodium Cr chromium NH4Br ammonium bromide Cu copper NH4NO3 ammonium nitrate Fe iron Ni nickel Hg mercury P phosphorus H3BO3 boric acid Pb lead HCl hydrochloric acid S sulphur HClO4 perchloric acid Sb antimony HF hydrofluoric acid Si silicon HNO3 nitric acid Ti titanium H2O2 hydrogen peroxide Tl thallium K potassium V vanadium Li lithium Zn zinc



Abbreviations and definitions CVAAS cold vapour atomic absorption spectrometry: system for

determination of Hg d.b dry basis: condition in which the solid biofuel is free from

moisture direct Hg determinations: system for direct determination (without prior digestion) of Hg (general) analysis sample sub-sample of a laboratory sample having a nominal top size of

1 mm or less and used for a number of chemical and physical analyses

FAAS flame atomic absorption spectrometry: system for element determinations

FAES flame atomic emission spectrometry: system for element determinations

GFAAS graphite furnace atomic absorption spectrometry (also referred to as ETAAS – electrothermal AAS): system for trace element determinations

Hydride-generation AAS hydride generation atomic absorption spectrometry: system for determination of As, Se, (Bi, Sb, Pb)

ICP- MS inductively coupled plasma mass spectrometry: system for multi-element determinations of trace elements

ICP-OES inductively coupled plasma optical emission spectrometry: system for multi-element determinations

laboratory sample combined sample or a sub-sample of a combined sample or an increment or a sub-sample of an increment sent to a laboratory

nominal top size aperture size of the sieve used in the CEN method for determining the particle size distribution of solid biofuels through which at least 95% by mass of the material passes

RSD relative standard deviation sample quantity of fuel, representative of a larger mass for which the

quality is to be determined size reduction reduction of the nominal top size of a sample or sub-sample sub-sample portion of a sample test portion sub-sample of a laboratory sample consisting of the quantity of

material required for a single execution of a test method XRF X-ray fluorescence spectrometry: system for multi-element

determinations



9 Annex

This Annex presents the results of the investigations within WP III/TaskIII.2 - Subtask III.2.3 – part 1 and part 2. For explanations please refer to chapter 4.3 and 4.4 of the final report of WP III/Task III.2.

Table 1: Participating laboratories and applied detection and digestion methods for the investigations within Subtask III.2.3 “Method testing and improvement – part 1”

Code Laboratory A AC-TUG D IEC-DPI E ELSAM I IEC N NYME S SP U USTUTT V VT-TUG Dection systems applied

C CVAAS D Direct Hg determination E FAES F FAAS G GFAAS M ICP-MS O ICP-OES X XRF Digestion methods applied

1 H2O2/HNO3 (microwave heating) 2 H2O2/HNO3/HClO4 (microwave heating) /HF 3 H2O2/HNO3 (conventional heating, ½ hour cooking, filtration through filter paper) 4 H2O2/HNO3/HCl (microwave heating) 5 H2O2/HNO3/HClO4 (conventional heating) /HF 6 HNO3/HF (microwave heating) 7 H2O2/HNO3/HF (conventional heating) /H3BO3 8 H2O2/HNO3/HF (microwave heating) /H3BO3 9 HNO3/HF (microwave heating) /H3BO3

10 Preashing at 550 °C, Li2BO2 (1050 °C), HNO3/H2O (ASTM D 3682) 11 Preashing at 550 °C, HNO3/HF/H3BO3 (ASTM D 3682) 12 Direct Hg determination 13 Preashing at 450 °C/Li2B4O7/NH4NO3 (only straw was measured) 14 Preashing at 525 °C/Li2B4O7/NH4NO3 (only straw was measured) 15 Preashing at 815 °C/Li2B4O7/NH4NO3 (only straw was measured) 16 Pellets (direct measurement of palletised material)



A lm

g/kg

d.b

.

G rand Mean 81.4

AO

2

AO

5

DO

4

DO

8

EO

7

EO

8

IF4

IF8

NO

3

SO

10

SO

11

VM

6

VM

9

0

20

40

60

80

100

120

Ca

mg/

kg d

.b.

G rand Mean 1790

AO

2A

O5

DO

4D

O8

EO

7E

O8

IF4

IF8

NO

3S

O10

SO

11U

X16

VF6

VF9

VM

6V

M9

900

1200

1500

1800

2100

2400

Fe

mg/

kg d

.b.

G rand M ean 85.4

AO

2A

O5

DO

4D

O8

EO

7E

O8

IF4

IF8

NO

3S

O10

SO

11U

X16

VM

6V

M9

58

78

98

118

138

K

mg/

kg d

.b.

G rand Mean 695

AO

2A

O5

DO

4D

O8

EO

7E

O8

IF4

IF8

NE

3S

O10

SO

11U

X16

VF6

VF9

VM

6V

M9

420

520

620

720

820

920

Mg

mg/

kg d

.b.

G rand Mean 212

AO

2A

O5

DO

4D

O8

EO

7E

O8

IF4

IF8

NO

3S

O10

SO

11U

X16

VF6

VF9

VM

6V

M9

160

180

200

220

240

260

A l

mg/

kg d

.b.

G ra n d M e a n 3 6 0

AO

2A

O5

DO

4D

O8

EO

7E

O8

IF4

IF8

IX13

IX14

IX15

NO

3S

O10

SO

11V

M6

VM

9

0

2 0 0

4 0 0

6 0 0

8 0 0

1 0 0 0

C a

mg/

kg d

.b.

G ran d M ean 5 8 5 0

AO

2A

O5

DO

4D

O8

EO

7E

O8

IF4

IF8

IX13

IX14

IX15

NO

3S

O10

SO

11U

X16

VF6

VF9

VM

6V

M94 3 0 0

5 3 0 0

6 3 0 0

7 3 0 0

8 3 0 0

Fem

g/kg

d.b

.

G ran d M ean 3 3 2

AO

2A

O5

DO

4D

O8

EO

7E

O8

IF4

IF8

IX13

IX14

IX15

NO

3S

O10

SO

11U

X16

VM

6V

M92 0 0

3 0 0

4 0 0

5 0 0

6 0 0

K

mg/

kg d

.b.

G ra n d M e a n 1 2 3 0 0

DO

4D

O8

EO

7E

O8

IF4

IF8

IX13

IX14

IX15

NE

3S

O10

SO

11U

X16

VF6

VF9

VM

6V

M9

6 5 0 0

8 5 0 0

1 0 5 0 0

1 2 5 0 0

1 4 5 0 0

1 6 5 0 0

M g

mg/

kg d

.b.

G ran d M ean 1 3 9 0

AO

2AO

5D

O4

DO

8E

O7

EO8

IF4

IF8

IX13

IX14

IX15

NO

3SO

10S

O11

UX

16V

F6V

F9V

M6

VM

9

1 1 0 0

1 2 0 0

1 3 0 0

1 4 0 0

1 5 0 0

1 6 0 0

1 7 0 0

WOOD STRAW

Figure 1: Results of Al, Ca, Fe, K, and Mg determinations within Subtask III.2.3 – part 1




Nam

g/kg

d.b

.Grand Mean 57.2

AO

2

AO

5

DO

8

EO

7

EO

8

IF4

IF8

SO

10

SO

11

UX

16

VM

6

VM

90

200

400

600

800

1000

1200

P

mg/

kg d

.b.

G rand M ean 86.6

AO

2

AO

5

DO

4

DO

8

EO

7

EO

8

SO

10

SO

11

UX

16

VM

6

VM

959

69

79

89

99

109

Ti

mg/

kg d

.b.

G rand Mean 7.3

AO

2A

O5

DG

4D

G8

DM

4D

M8

DO

4D

O8

EO

7E

O8

SO

10S

O11

VM

6V

M9

0

3

6

9

12

15

N a

mg/

kg d

.b.

G ran d M ean 1 8 6

AO

2A

O5

EO

7E

O8

IF4

IF8

IX13

IX14

IX15

NO

3S

O10

SO

11U

X16

VF6

VF9

VM

6V

M90

3 0 0

6 0 0

9 0 0

1 2 0 0

1 5 0 0

1 8 0 0

P

mg/

kg d

.b.

G r and M ean 1 310

AO

2A

O5

DO

4D

O8

EO

7E

O8

IX13

IX14

IX15

NO

3S

O10

SO

11U

X16

VM

6V

M98 4 0

10 40

12 40

14 40

16 40

T i

mg/

kg d

.b.

G rand M ean 88

AO

2A

O5

DG

4D

G8

DM

4D

M8

DO

4D

O8

EO

7E

O8

IX13

IX14

IX15

SO

10S

O11

UX

16V

M6

VM

9

0

30

60

90

120

150

180

Sim

g/kg

d.b

.Grand Mean 14500

AO

2D

G4

DG

8D

O4

DO

8E

O7

EO

8IF

4IF

8IX

13IX

14IX

15N

O3

SO

10S

O11

UX1

6V

F6V

F9

0

4

8

12

16

20

24(X 1000)

Si

mg/

kg d

.b.

Grand Mean 491

AO

2

DG

4

EO

7

EO

8

IF4

IF8

NO

3

SO

10

SO

11

UX

16

0

200

400

600

800

1000

1200

WOOD STRAW

Figure 2: Results of Na, P, Si and Ti determinations within Subtask III.2.3 – part 1




A s

mg/

kg d

.b.

EG

7

EG

8

SM

1

VM

6

VM

90

0 .0 2

0 .0 4

0 .0 6

0 .0 8

0 .1

0 .1 2

Cd

mg/

kg d

.b.

Grand Mean 0.50

AO

2A

O5

DG

4D

G8

DM

4D

M8

DO

4D

O8

EG

7E

G8

IF8

SM

1S

O11

VM

6V

M9

0

0,4

0,8

1,2

1,6

2

2,4

Cr

mg/

kg d

.b.

G rand Mean 1.83

AO

2A

O5

DG

4D

G8

DM

4D

M8

DO

4D

O8

EO

1E

O7

EO

8IF

4IF

8S

M1

SO

11V

M6

VM

9

0

1

2

3

4

A s

mg/

kg d

.b.

EG

7

EG8

SM

1

VM

6

VM

90 .1 3

0 .1 8

0 .2 3

0 .2 8

0 .3 3

0 .3 8

0 .4 3G ran d M ean 0 .2 3

C d

mg/

kg d

.b.

G rand M ean 0.31

DG

4D

G8

DM

4D

M8

DO

4D

O8

EG

7E

G8

IF8

IX13

IX14

SM

1S

O11

VM

6V

M9

0

0 .3

0 .6

0 .9

1 .2

1 .5

1 .8

C r

mg/

kg d

.b.

DG

4D

G8

DM

4D

M8

DO

4D

O8

EO

1E

O7

EO

8IF

4IF

8IX

13IX

14IX

15S

M1

SO

11V

M6

VM

9

0

3

6

9

1 2

1 5

1 8G rand M ean 1.7 7

B a

mg/

kg d

.b.

G ran d M ean 26 .8

AO

2A

O5

DG

4D

M4

DO

4E

O1

EO

7E

O8

IF4

IF8

SO

10S

O11

VM

6V

M9

21

26

31

36

41

46

51

B a

mg/

kg d

.b.

G ran d M ean 89 .7

AO

2A

O5

DG

4D

G8

DM

4D

M8

DO

4D

O8

EO

1E

O7

EO

8IF

4IF

8IX

13IX

14IX

15S

O10

SO

11V

M6

VM

90

40

80

12 0

16 0

20 0

24 0

C o

mg/

kg d

.b.

G rand M ean 0 .27

DG

4

DO

8

EG

7

EG

8

IX15

SM

1

VM

6

VM

9

0

0 .2

0 .4

0 .6

0 .8

1

Co

mg/

kg d

.b.

Grand Mean 0.51

DG

8

DM

8

DO

4

DO

8

EG7

EG8

SM1

SO11

VM6

VM9

0 .29

0.49

0.69

0.89

1.09

WOOD STRAW

Figure 3: Results of As, Ba, Cd, Co and Cr determinations within Subtask III.2.3 – part 1




Cum

g/kg

d.b

.

Grand Mean 3.04

AO

2A

O5

DG

4D

G8

DM

4D

M8

DO

4D

O8

EO

1E

O7

EO

8IF

4IF

8N

O3

SM

1S

O11

VM

6V

M9

0

5

10

15

20

25

30

N i

mg/

kg d

.b.

DG

4

DG

8

DM

4

DM

8

DO

4

DO

8

EO

1

EO

7

EO

8

SM

1

SO

11

VM

6

VM

90.8

1.2

1.6

2

2.4G rand M ean 1 .58

N i

mg/

kg d

.b.

G rand M ean 1 .38

DG

4D

G8

DM

8D

O4

DO

8E

O1

EO

7E

O8

IX13

IX14

IX15

SM

1S

O11

VM

6V

M9

0

2

4

6

8

10

Cu

mg/

kg d

.b.

Grand Mean 5.16

AO2

AO5

DG

4D

G8

DM

4D

M8

DO

4D

O8

EO1

EO7

EO8

IF4

IF8

IX13

IX14

IX15

SM1

SO11

VM6

VM9

0

2

4

6

8

10

12

Hg-1

mg/

kg d

.b.

Grand Mean 0.03

EC

7

EC

8

ED

12

SD

12

SO10

SO11

VM6

VM9

0

0.02

0.04

0.06

0.08

0.1

0.12

Hg-1

mg/

kg d

.b.

Grand Mean 0.15

EC

7

EC

8

ED

12

SD

12

SO

10

SO

11

VM

6

VM

9

0

0.2

0.4

0.6

0.8

1

Mn

mg/

kg d

.b.

G rand Mean 293

AO

2A

O5

DO

4D

O8

EO

1E

O7

EO

8IF

4IF

8N

O3

SM

1S

O10

SO

11V

F6V

F9V

M6

VM

9

90

140

190

240

290

340

390

M nm

g/kg

d.b

.

G ran d M ean 5 3 .8

AO

2A

O5

DO

4D

O8

EO

1E

O7

EO

8IF

4IF

8IX

13IX

14IX

15N

O3

SM

1S

O10

SO

11V

M6

VM

90

5 0

1 0 0

1 5 0

2 0 0

2 5 0

3 0 0

Mo

mg/

kg d

.b.

Grand Mean 0.09

EO

1

EO

7

EO

8

VM

6

VM

90

0.03

0.06

0.09

0.12

0.15

M o

mg/

kg d

.b.

DM

4

EO

1

EO

7

EO

8

VM

6

VM

9

0 .32

0 .37

0 .42

0 .47

0 .52

0 .57

0 .62G rand M ean 0.43

WOOD STRAW

Figure 4: Results of Cu, Hg, Mn, Mo and Ni determinations within Subtask III.2.3 – part 1




P bm

g/kg

d.b

.

DG

4

DG

8

DM

4

DM

8

DO

4

DO

8

EG

7

EG

8

SM

1

SO

11

VM

6

VM

9

0 .5

0 .8

1.1

1.4

1.7

2G rand M ean 1.06

Zn

mg/

kg d

.b.

Grand Mean 22 .3

AO

2A

O5

DM

4D

M8

DO

4D

O8

EO

1E

O7

EO

8IF

4IF

8N

O3

SM

1S

O11

VM

6V

M9

15

25

35

45

55

P b

mg/

kg d

.b.

G rand M ean 1.28

DG

4

DG

8

DM

4

DM

8

DO

4

DO

8

EG

7

EG

8

SM

1

SO

11

VM

6

VM

90

0 .4

0 .8

1 .2

1 .6

2

2 .4

Z n

mg/

kg d

.b.

G rand M ean 22 .4

AO

2A

O5

DM

4D

M8

DO

4D

O8

EO

1E

O7

EO

8IF

4IF

8IX

13IX

14IX

15N

O3

SM

1S

O11

VM

6V

M9

0

20

40

60

80

S b

mg/

kg d

.b.

G rand M ean 0.04

EG

7

EG

8

VM

6

VM

9

0

0 .03

0.06

0.09

0.12

0.15

S b

mg/

kg d

.b.

DM

4

EG

7

EG

8

IX13

IX14

IX15

VM

6

VM

9

0

0 .5

1

1 .5

2

2 .5

3G rand M ean 1.05

Tl

mg/

kg d

.b.

DO

4

DO

8

EG

7

EG

8

VM

6

VM

9-0.2

1.8

3.8

5.8

7.8Grand Mean 1.64 T l

mg/

kg d

.b.

G rand M ean 0 .008

EG

7

EG

8

VM

6

VM

9

-0 .11

-0 .08

-0 .05

-0 .02

0 .01

0 .04

0 .07

V

mg/

kg d

.b.

G ran d M e an 0 .1 1

EO

1

EO

7

EO

8

SM

1

SO

11

VM

6

VM

97 1

9 1

1 1 1

1 3 1

1 5 1(X 0 .0 0 1 )

V

mg/

kg d

.b.

G rand M ean 0 .71

AO

2A

O5

DG

4D

G8

DM

4D

M8

DO

4D

O8

EO

1E

O7

EO

8IX

13IX

14IX

15S

M1

SO

11V

M6

VM

9

0

0 ,3

0 ,6

0 ,9

1 ,2

1 ,5

1 ,8

WOOD STRAW

Figure 5: Results of Pb, Sb, Ts, V and Zn determinations within Subtask III.2.3 – part 1




Al-Straw

0

50

100

150

200

250

300

350

400

450

500

HF 20 40 80 80 80-

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 220 220 105 220

Det M O F O O O O O

Lab V D I E S A N L

mg

/kg

d.b

.

Ca-Straw

4000

4500

5000

5500

6000

6500

7000

7500

HF 20 40 80 80 80-

20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 190 220 220 105 220



mg

/kg

d.b

.

Fe-Straw

200

250

300

350

400

450

HF 20 40 80 80 80-

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 220 220 105 220

Det M O F O O O O O

Lab V D I E S A N L

mg/

kg d

.b.

Al-Olive Residues

4821500

1700

1900

2100

2300

2500

2700

2900

3100

HF 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 220 220 105 220

Det M O F O O O O O

Lab V D I E S A N L

mg/

kg d

.b.

Ca-Olive Residues

11000

12000

13000

14000

15000

16000

17000

HF 20 40 80 80 80 -

20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 190 220 220 105 220



mg/

kg d

.b.

Fe-Olive Residues

1300

1400

1500

1600

1700

1800

1900

2000

HF 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 220 220 105 220

Det M O F O O O O O

Lab V D I E S A N L

mg/

kg d

.b.

Al-Reference Material

4301000

1200

1400

1600

1800

2000

2200

2400

2600

2800

3000

HF 80 80 64 400 80 200 80 100 0 20 40 80 20 80 80

- # * - # # # * -

°C 220 190 190 210 220 190 220 220 105

Det M O F O O O O

Lab V D I E S A N

mg/

kg d

.b.

CertifiedValue

Ca-Reference Material

17000

18000

19000

20000

21000

22000

23000

24000

25000

26000

27000

HF 80 80 80 80 64 400 80 200 80 100 0 20 40 80 20 80 80

- - # * - # # # * -

°C 220 220 190 190 210 220 190 220 220 105

Det M F O F O O O O

Lab V V D I E S A N

mg/

kg d

.b.

CertifiedValue

Fe-Reference Material

600

700

800

900

1000

1100

1200

1300

HF 80 80 64 400 80 200 80 100 0 20 40 80 20 80 80

- # * - # # # * -

°C 220 190 190 210 220 190 220 220 105

Det M O F O O O O

Lab V D I E S A N

mg/

kg d

.b.

CertifiedValue

Figure 6: Results of Al, Ca and Fe determinations within Subtask III.2.3 – part 2

Explanations: digestions were performed with 0.5 ml H2O2 / 1 ml HNO3 (acid quantities are referred to 100 mg sample (d.b.)) and varying amounts of HF (µl) for straw and olive residues and with varying amounts of H2O2 / HNO3 and HF (the HF quantities (µl) stated are referred to 100 mg sample (d.b.)) for reference material GBW 07602 (NCS DC 73348) at different temperatures; the solid line refers to the mean value, the dashed line defines the 95% confidence interval; HF...hydrofluoric acid; °C...temperature in degree Celsius; Det...detection system (F...FAAS or FAES, G...GFAAS, M...ICP-MS, O...ICP-OES); Lab... laboratory (V...VT-TUG, D...DPI, I...IEC, E...ELSAM, S...SP, A...AC-TUG, N...NYME, L...LH, U...USTUTT); #...neutralisation (complexation) with H3BO3; -...digestion without H2O2; *...digestion with addition of HCl; the depictured means plots have been calculated from n=2-6 independent determinations



K-Straw

9000

10000

11000

12000

13000

14000

15000

16000

17000

18000

HF 20 40 80 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 190 220 190 190 220 190 220 220 105 220



mg/

kg d

.b.

K-Olive Residues

21000

22000

23000

24000

25000

26000

27000

28000

29000

30000

31000

HF 20 40 80 80 80 -

20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 80 20 80

°C 190 220 190 220 190 190 220 190 220 105 220

Det M F O F O O O O

Lab V V D I E S N L

mg/

kg d

.b.

Mg-Straw

1000

1100

1200

1300

1400

1500

1600

1700

1800

1900

2000

HF 20 40 80 80 80-

20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 190 220 220 105 220



mg/

kg d

.b.

Mg-Olive Residues

1,5162500

2700

2900

3100

3300

3500

3700

HF 20 40 80 80 80 -

20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 190 220 220 105 220



mg/

kg d

.b.

K-Reference Material

6000

6500

7000

7500

8000

8500

9000

9500

10000

HF 80 80 80 80 64 400 80 200 80 100 0 20 40 80 20 80 80

- - # * - # # # * -

°C 220 220 190 190 210 220 190 220 220 105

Det M F O F O O O O

Lab V V D I E S A N

mg/

kg d

.b.

CertifiedValue

Mg-Reference Material

1400

1600

1800

2000

2200

2400

2600

2800

3000

3200

3400

HF 80 80 80 80 64 400 80 200 80 100 0 20 40 80 20 80 80

- - # * - # # # * -

°C 220 220 190 190 210 220 190 220 220 105

Det M F O F O O O O

Lab V V D I E S A N

mg/

kg d

.b.

CertifiedValue

Na-Reference Material

8000

8500

9000

9500

10000

10500

11000

11500

12000

12500

13000

HF 80 80 80 80 64 400 80 200 80 100 0 20 40 80 20 80 80

- - # * - # # # * -

°C 220 220 190 190 210 220 190 220 220 105

Det M F O F O O O O

Lab V V D I E S A N

mg

/kg

d.b

.

CertifiedValue

Na-Straw284

0

20

40

60

80

100

120

140

160

180

200

HF 20 40 80 80 80-

20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 190 220 220 105 220



mg/

kg d

.b.

Na-Olive Residues 332

50

70

90

110

130

150

170

190

210

230

250

HF 20 40 80 80 80 -

20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 190 220 220 105 220



mg/

kg d

.b.

Figure 7: Results of K, Mg and Na determinations within Subtask III.2.3 – part 2




P-Reference Material 1282

650

700

750

800

850

900

950

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M O O O O O

Lab V D E S A N

mg/

kg d

.b.

CertifiedValue

Si-Reference Material

1482000

2500

3000

3500

4000

4500

5000

5500

6000

6500

7000

HF 80 80 64 400 80 200 80 100 0 20 40 80 20 80

- # * - # # # * -

°C 220 190 190 210 220 190 220 220

Det F O F O O O

Lab V D I E S A

mg/

kg d

.b.

CertifiedValue

Ti-Reference Material

30

40

50

60

70

80

90

100

110

120

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M O O O O O

Lab V D E S A N

mg

/kg

d.b

.

CertifiedValue

P-Straw

1000

1100

1200

1300

1400

1500

1600

1700

1800

1900

HF 20 40 80 80 80-

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M O O O O O O

Lab V D E S A N L

mg/

kg d

.b.

Si-Straw

0

5000

10000

15000

20000

25000

30000

HF 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 20 80

°C 190 220 190 190 220 190 220 220 220

Det F O F O O O O

Lab V D I E S A L

mg

/kg

d.b.

T i-Straw

0

50

100

150

200

250

HF 20 40 80 80 80-

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M O O O O O O

Lab V D E S A N L

mg

/kg

d.b

.

P-Olive Residues

700

900

1100

1300

1500

1700

1900

2100

2300

HF 20 40 80 80 80 -

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M O O O O O O

Lab V D E S A N L

mg/

kg d

.b.

Si-Olive Residues

0

2000

4000

6000

8000

10000

12000

14000

HF 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 20 80

°C 190 220 190 190 220 190 220 220 220

Det F O F O O O O

Lab V D I E S A L

mg/

kg d

.b.

Ti-Olive Residues

50

100

150

200

250

300

HF 20 40 80 80 80 -

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M O O O O O O

Lab V D E S A N L

mg/

kg d

.b.

Figure 8: Results of P, Si and Ti determinations within Subtask III.2.3 – part 2




As-Reference Material 177.3

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6

1.7

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M G G M M O

Lab V D E S A N

mg

/kg

d.b

.

CertifiedValue

Ba-Reference Material

14

16

18

20

22

24

26

28

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M O O M O O

Lab V D E S A N

mg

/kg

d.b

.

CertifiedValue

Cd-Reference Material

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M G G M M O

Lab V D E S A N

mg

/kg

d.b

.

CertifiedValue

As-Straw

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

HF 20 40 80 80 80-

80 #

0 20 40 80 20 20 40 80 80 80

°C 190 220 220 190 190 220 220 105 220

Det M G M M O C

Lab V E S A N L

mg

/kg

d.b

.

Ba-Straw

50

55

60

65

70

75

80

85

90

95

100

HF 20 40 80 80 80-

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 190 220 220 105 220

Det M O F O M O O O

Lab V D I E S A N L

mg

/kg

d.b.

Cd-Straw1.9 1.6 2.0

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

HF 20 40 80 80 80-

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 220 105 220

Det M G G M M O G

Lab V D E S A N L

mg/

kg d

.b.

As-Olive Residues 2.72.52.6

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

HF 20 40 80 80 80 -

80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 220 190 220 220 105 220

Det M G M M O C

Lab V E S A N L

mg/

kg d

.b.

Ba-Olive Residues

10

12

14

16

18

20

22

24

26

HF 20 40 80 80 80 -

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M O O M O O O

Lab V D E S A N L

mg/

kg d

.b.

Cd-Olive Residues0.741.171.44

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

HF 20 40 80 80 80 -

80 #

20 40 80 80 20 80

°C 190 220 220 220 105 220

Det M G M O G

Lab V E A N L

mg/

kg d

.b.

Figure 9: Results of As, Ba and Cd determinations within Subtask III.2.3 – part 2




Co-Reference Material

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

HF 80 80 80 100 0 20 40 80 20 80 80

- # # * -

°C 220 220 190 220 220 105

Det M G M M O

Lab V E S A N

mg

/kg

d.b

.

CertifiedValue

Cr-Reference Material

0

1

2

3

4

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M O O M O O

Lab V D E S A N

mg/

kg d

.b.

CertifiedValue

Cu-Reference Material

3

4

5

6

7

8

9

10

11

12

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M O O M O O

Lab V D E S A N

mg/

kg d

.b.

CertifiedValue

Co-Straw

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

HF 20 40 80 80 80-

80 #

0 20 40 80 20 20 40 80 80

°C 190 220 220 190 190 220 220 105

Det M G M M O

Lab V E S A N

mg

/kg

d.b

.

Cr-Straw

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

HF 20 40 80 80 80-

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 220 105 220

Det M O O M O O G

Lab V D E S A N L

mg/

kg d

.b.

Cu-Straw

0

1

2

3

4

5

6

7

8

HF 20 40 80 80 80-

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 220 105 220

Det M O O M O O G

Lab V D E S A N L

mg

/kg

d.b

.

Co-Olive Residues

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

HF 20 40 80 80 80 -

20 40 80 80 #

0 20 40 80 20 40 80 20 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M G G M M O G

Lab V D E S A N L

mg/

kg d

.b.

Cr-Olive Residues

10

12

14

16

18

20

22

24

HF 20 40 80 80 80 -

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M O O M O O G

Lab V D E S A N L

mg/

kg d

.b.

Cu-Olive Residues

20

22

24

26

28

30

32

34

HF 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 220 220 105 220

Det M O F O M O O G

Lab V D I E S A N L

mg/

kg d

.b.

Figure 10: Results of Co, Cr and Cu determinations within Subtask III.2.3 – part 2




Hg-Reference Material

0

10

20

30

40

50

60

HF 64 400 80 100 80

# * - # #

°C 190 220 105

Det C C O

Lab D E N

mg/

kg d

.b.

Mn-Reference Material

45

50

55

60

65

70

HF 80 80 64 400 80 200 80 100 0 20 40 80 20 80 80

- # * - # # # * -

°C 220 190 190 210 220 190 220 220 105

Det M O F O M O O

Lab V D I E S A N

mg/

kg d

.b.

CertifiedValue

Mo-Reference Material

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

HF 80 80 80 100 0 20 40 80 20 80 80

- # # * -

°C 220 220 190 220 220 105

Det M O M M O

Lab V E S A N

mg/

kg d

.b.

CertifiedValue

Hg-Straw

0.2

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

HF 20 40 80 80 80-

80 #

80 40 40

°C 190 220 220 D D 105 190 220

Det M C D D O C

Lab V E E S N U

mg/

kg d

.b.

Mn-Straw

20

25

30

35

40

45

50

55

HF 20 40 80 80 80-

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 190 220 220 105 220

Det M O F O M O O O

Lab V D I E S A N L

mg/

kg d

.b.

Mo-Straw

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

HF 20 40 80 80 80-

80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 220 190 190 220 220 105 220

Det M O M M O G

Lab V E S A N L

mg/

kg d

.b.

Hg-Olive Residues

0.0730.196

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

HF 20 40 80 80 80 -

80 #

80 40 40

°C 190 220 220 D D 105 190 220

Det M C D D O C

Lab V E E S N U

mg/

kg d

.b.

Mn-Olive Residues

35

36

37

38

39

40

41

42

43

44

45

HF 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 220 220 105 220

Det M O F O M O O O

Lab V D I E S A N L

mg/

kg d

.b.

Mo-Olive Residues

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

HF 20 40 80 80 80 -

80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 220 190 220 220 105 220

Det M O M M O G

Lab V E S A N L

mg/

kg d

.b.

Figure 11: Results of Hg, Mn and Mo determinations within Subtask III.2.3 – part 2

Explanations: digestions were performed with 0.5 ml H2O2 / 1 ml HNO3 (acid quantities are referred to 100 mg sample (d.b.)) and varying amounts of HF (µl) for straw and olive residues and with varying amounts of H2O2 / HNO3 and HF (the HF quantities (µl) stated are referred to 100 mg sample (d.b.)) for reference material GBW 07602 (NCS DC 73348) at different temperatures; the solid line refers to the mean value, the dashed line defines the 95% confidence interval; Hg is not certified within GBW 07602; HF...hydrofluoric acid; °C...temperature in degree Celsius; Det...detection system (F...FAAS or FAES, G...GFAAS, M...ICP-MS, O...ICP-OES); Lab... laboratory (V...VT-TUG, D...DPI, I...IEC, E...ELSAM, S...SP, A...AC-TUG, N...NYME, L...LH, U...USTUTT); #...neutralisation (complexation) with H3BO3; -...digestion without H2O2; *...digestion with addition of HCl; the depictured means plots have been calculated from n=2-6 independent determinations



Ni-Reference Material 4.5

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M O O M O O

Lab V D E S A N

mg/

kg d

.b.

CertifiedValue

Pb-Reference Material

4

5

6

7

8

9

10

11

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M G G M M O

Lab V D E S A N

mg

/kg

d.b

.

CertifiedValue

Sb-Reference Material

0.00

0.04

0.08

0.12

0.16

0.20

HF 80 80 80

- * -

°C 220 220

Det M M

Lab V A

mg/

kg d

.b.

CertifiedValue

Ni-Straw5.2

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

HF 20 40 80 80 80-

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 220 105 220

Det M O O M O O G

Lab V D E S A N L

mg

/kg

d.b

.

Pb-Straw

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

HF 20 40 80 80 80-

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 190 220 220 105 220

Det M G G M M O G

Lab V D E S A N L

mg/

kg d

.b.

Sb-Straw

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

HF 20 40 80 80 80-

20 40 80

°C 190 220 220

Det M M

Lab V A

mg/

kg d

.b.

Ni-Olive Residues

0

2

4

6

8

10

12

14

16

18

20

HF 20 40 80 80 80 -

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M O O M O O G

Lab V D E S A N L

mg/

kg d

.b.

Pb-Olive Residues

0

2

4

6

8

10

12

14

HF 20 40 80 80 80 -

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M G G M M O G

Lab V D E S A N L

mg/

kg d

.b.

Sb-Olive Residues

0.00

0.02

0.04

0.06

0.08

0.10

0.12

HF 20 40 80 80 80 -

20 40 80

°C 190 220 220

Det M M

Lab V A

mg/

kg d

.b.

Figure 12: Results of Ni, Pb and Sb determinations within Subtask III.2.3 – part 2




V-Reference Material

1.5

2.0

2.5

3.0

3.5

HF 80 80 64 400 80 100 0 20 40 80 20 80 80

- # * - # # * -

°C 220 190 220 190 220 220 105

Det M O O M O O

Lab V D E S A N

mg

/kg

d.b

.

CertifiedValue

Zn-Reference Material

12

17

22

27

32

37

HF 80 80 64 400 80 200 80 100 0 20 40 80 20 80 80

- # * - # # # * -

°C 220 190 190 210 220 190 220 220 105

Det M O F O M O O

Lab V D I E S A N

mg/

kg d

.b.

CertifiedValue

Tl-Straw

0.000

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

0.010

HF 80 80-

20 40 80

°C 220 220

Det M M

Lab V A

mg/

kg d

.b.

V-Straw

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

HF 20 40 80 80 80-

80 #

0 20 40 80 20 20 40 80 80

°C 190 220 220 190 190 220 220 105

Det M O M O O

Lab V E S A N

mg/

kg d

.b.

Zn-Straw

0

5

10

15

20

25

30

35

40

HF 20 40 80 80 80-

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 190 220 220 105 220

Det M O F O M O O O

Lab V D I E S A N L

mg

/kg

d.b.

Tl-Olive Residues

0.00

0.02

0.04

0.06

0.08

0.10

0.12

HF 20 40 80 80 80 -

20 40 80

°C 190 220 220

Det M M

Lab V A

mg/

kg d

.b.

V-Olive Residues

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

HF 20 40 80 80 80 -

20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 220 190 220 220 105 220

Det M O O M O O O

Lab V D E S A N L

mg/

kg d

.b.

Zn-Olive Residues

0

5

10

15

20

25

30

35

40

HF 20 40 80 80 80 -

20 40 80 20 40 80 80 #

0 20 40 80 20 20 40 80 80 20 80

°C 190 220 190 190 220 190 220 220 105 220

Det M O F O M O O O

Lab V D I E S A N L

mg/

kg d

.b.

Figure 13: Results of Tl, V and Zn determinations within Subtask III.2.3 – part 2

Explanations: digestions were performed with 0.5 ml H2O2 / 1 ml HNO3 (acid quantities are referred to 100 mg sample (d.b.)) and varying amounts of HF (µl) for straw and olive residues and with varying amounts of H2O2 / HNO3 and HF (the HF quantities (µl) stated are referred to 100 mg sample (d.b.)) for reference material GBW 07602 (NCS DC 73348) at different temperatures; the solid line refers to the mean value, the dashed line defines the 95% confidence interval; Tl is not certified within GBW 07602; HF...hydrofluoric acid; °C...temperature in degree Celsius; Det...detection system (F...FAAS or FAES, G...GFAAS, M...ICP-MS, O...ICP-OES); Lab... laboratory (V...VT-TUG, D...DPI, I...IEC, E...ELSAM, S...SP, A...AC-TUG, N...NYME, L...LH, U...USTUTT); #...neutralisation (complexation) with H3BO3; -...digestion without H2O2; *...digestion with addition of HCl; the depictured means plots have been calculated from n=2-6 independent determinations



Part 2.4 Fuel Quality Assurance

Final report prepared by: Langheinrich, Christian1) Contributing co-authors: Eija Alakangas2), Margaret Biddlecombe3), Peter Jansen4), Martin Kaltschmitt1), Jaap Koppejan4), Jan-Erik Levlin5), Max Nitschke6), Gitte Nyhus6), Dr. D.C. Pike3), Jouni Valtanen5) 1) Institute for Energy and Environment gGmbH, Germany 2) Technical Research Centre of Finland - VTT Processes, Finland 3) Green Land Reclamation Ltd, United Kingdom 4) TNO Environment, Energy & Process Innovation, The Netherlands 5) Finnish Pulp & Paper Research Institute, Finland 6) Elsam Engineering A/S, Denmark

1 Summary

Work-package IV (WPIV) of the BIONORM project provided information to support the Technical Specification on Quality Assurance [4] for solid biofuels that is being drafted by Working Group 2 of CEN/TC335 (WG2). The aim of WPIV was to fill gaps in knowledge about Quality Assurance in the field of solid biofuels, and this has been achieved in three tasks.

In the first task in WPIV (Task IV.1), a review of existing, relevant Quality Management systems [2] as already maintained by different biofuel producers, was elaborated. This review covered solid biofuels and excluded recovered fuels. On the basis of a list of 11 questions, the review yields an indication how Quality Assurance and Quality Control – as parts of Quality Management – are currently performed in ten different cases, representing six different countries (Denmark, Finland, Germany, Netherlands, UK, Sweden) and six different product categories (agricultural products, pelletised animal feed, used wood, straw bales, fresh wood chips, wood pellets). As a result of the review performed in Task IV.1 and the analysis of ISO 9001:2000 as an international Quality Management standard, a report has been produced with the pros and cons of Quality Assurance and Quality Control systems as well as the basic ideas of a Quality Management system especially adapted to solid biofuels. After a description of the results, important conclusions are drawn for the design of a guideline for Quality Assurance (including Quality Control) for solid biofuels, emphasis of the work in IV.2.

In the second task (Task IV.2) a first draft of a guideline for Quality Assurance [3] was elaborated. The guideline sets out a step-by-step methodology to help each operator within a supply chain of solid biofuels to design a manual for Quality Assurance. Based on the ideas of this first draft guideline and conclusions of the review of Task IV.1 draft manuals how to deal with Quality Assurance and Quality Control were tried out in practical situations. This was accomplished at the industrial premises of a range of producers, traders and users of solid biofuels, referred to as “hosts”. These activities are known as the



“field-trials”. The selection of the host companies ensured, that the overall supply chain of solid biofeuls as producing, preparing, trading, handling and/or using of solid biofuels was covered. Furthermore the geographical distribution within Europe was considered. The hosts fell into two broad classes: Class-A companies that buy raw biomass, such as residues from agriculture and/or forestry and convert them into higher-grade biofuels for onward sale to third parties; and Class-B companies that buy such raw biomass and use it in processes to produce electricity and (sometimes) heat for sale. Both classes play key roles in the expanding market for biofuels. Beyond it, there was a wide range of possible circumstances to be covered. Most of those were included in a spectrum lying between two extreme cases: (a) small-scale (especially domestic) users who require high-grade fuels, and (b) large-scale users who can take advantage of lower-cost raw materials by the use of appropriately designed combustion plant. The field trials have underlined the need of a guideline with a general methodology applicable by all operators throughout the overall supply chain. The first draft of the guideline and its successor documents were optimised and improved on the basis of the findings and information gathered during the field trials. The final version of the guideline is the Deliverable IV.2.D4.

Besides the work in the field trials a proposal for a standard for Quality Assurance was elaborated in Task IV.3. This was done in close cooperation with WG2 from CEN/TC 335. WG2 expressed the need of a guideline as supporting document for the Technical Specification (TS) “Fuel QA for solid biofuels” [4] and supporting information to improve this TS. Thereupon WPIV commented the work of WG2 from their scientific point of view and participated in the further elaboration of the TS. This channel of communication and some common membership ensured a good linkage between the work of WPIV and WG2. Due to this close cooperation the proposal for a standard to be developed in Task IV.3 based upon the Technical Specification under development in TC 335/WG 2. It could found an agreement of a common document for the proposal of a standard and the TS from WG2. The outcome of Task IV.2 (the guideline as Deliverable IV.2.D4 [3]) will be further adapted to the Technical Specification from WG2 and published as CEN technical report.

2 Objectives

Solid biofuels can contribute significantly to reach the political goals of the European Union to increase the share of renewable energy within the energy system. For various reasons, however, this is not easily happening. In many cases, the costs of energy provision are higher for biofuels than fossil fuels. In order to develop a more widespread use and to compete with fossil fuels economically the costs of the production, provision and use of solid biofuels have to be reduced by optimising the interaction between fuel-producer, supplier and plant operator/end user.

Standardisation of biofuel properties as well as their sampling and test methods provides information and tools to facilitate the businesses and actions of operators within the market. Therefore standardisation also promotes a more widespread use of biofuels.

But standardisation alone cannot, however, ensure an increase of the market for solid biofuels. To achieve confidence of customers and regulators within this market, it is also essential that it can be demonstrated that the demanded level of quality is reached, and that adequate controls are ensured throughout the overall process-chain. Activities on



standardization therefore have to be accompanied by the development of a Quality Management (QM) system with main focus on Quality Assurance (QA) and Quality Control (QC).

Currently no Quality Management system exists in practice in Europe, which fulfils those demands. CEN/TC 335 has received a mandate from the European Commission (EC) to develop Standards for solid biofuels. The first step in the development of those Standards is the production of Technical Specifications (TS), which, after validation, will be upgraded to full European Standards.

BIONORM was designed to provide supporting information to CEN/TC 335 on solid biofuels. Specifically, Work-package 4 (WPIV) of the BIONORM Project was designed to fill gaps in knowledge about Quality Assurance in this field by a review, guideline and a proposal for a standard. This has been achieved in three tasks.

As one of the results a guideline [3] was elaborated which provides a methodology on how to develop and implement a Quality Assurance system (including Quality Control) within a company dealing with solid biofuels. Another outcome is the corporately with WG2 designed proposal of a standard [4] for Quality Assurance for solid biofuels.


The documents elaborated in WPIV are applicable for operators dealing with solid biofuels originating from the following sources:

• Products from agriculture and forestry

• Vegetable waste from agriculture and forestry

• Vegetable waste from food processing industry

• Wood waste, with the exception of wood waste which may contain halogenated organic compounds or heavy metals as a result of treatment with wood preservatives or coating, and which includes in particular such wood waste originated from construction and demolition waste

• Fibrous vegetable waste from virgin pulp production and from production of paper from pulp, if it is co-incinerated at the place of production and heat generated is recovered

• Cork waste

Solid biofuels from these sources are conform with the scope of the Technical Specification “Fuel specifications and classes”, elaborated by WG2, CEN/TC335 [1] in accordance with the mandate given for the standardisation work.




The results of WPIV of the BIONORM project were achieved in three tasks.

• Task IV.1: Review of Quality Management systems [2]

• Task IV.2: Implementation of Quality Assurance – field trials

• Task IV.3: Development of a proposal for a standard for Quality Assurance [4]

In the following the main results of the Task IV.1 – Task IV.3 will be described.

4.1 Review of quality systems

4.1.1 Approach

The aim of Task IV.1 was to write an elaborated review report of Quality Management systems that are already used by producers of solid biofuels as well as other sectors, operating under similar circumstances as the production, provision and trade of solid biofuels that apply Quality Management systems in accordance with e.g. ISO 9001. The review discussed the pros and cons of existing Quality Management systems as well as the basic ideas of Quality Assurance adapted to solid biofuels.

The extensive review was usefuel for extracting main components and aspects that are relevant for solid biofuel Quality Management, taking into account the whole supply chain. This review covers solid biofuels and excludes recovered fuels. This implies that some materials are to be excluded, such as waste wood that has been glued, painted or otherwise chemically treated, domestic waste, paper, cardboard and peat, all of which are excluded from the definition of “solid biofuels” that has been adopted by CEN/TC335.

It was decided to analyse existing Quality Management systems in industry and to identify the needs and the demands as well as the main components of these existing systems. On the basis of a list of 11 questions, the review yields an indication how Quality Assurance and Quality Control – as parts of Quality Management – are currently performed in ten different cases, representing six different countries and six different product categories. The following table illustrates the geographical coverage and the different products types of the cases covered in the review.

Table 1: Case studies in Task IV.1

Country of case study Type of product

Denmark Finland Germany Netherlands UK SwedenAgricultural products 1 8 Pelletised animal feed 9 Used wood 3 5 6, 7 Straw bales 2 Fresh wood chips 4 6 Wood pellets 10

The following cases were studied (numbers according to above table):



1. Risinge Gods, Denmark 2. Halm80 Aps, Denmark 3. Quality Assurance Manual for Recovered Fuels, Finland 4. Vapo Oy and Biowatti Oy, Finland 5. Bavaria Transport GmbH Lobenstein, Germany 6. Bruins en Kwast, Goor, Netherlands 7. Afvalzorg, Haarlem, Netherlands 8. ACCS quality-management system for food-grains, UK 9. Welsh Feed Producers Ltd, UK 10. Hedensbyn, Skellefteå, Sweden

4.1.2 Results

This review has yielded a number of interesting observations regarding the role Quality Management systems for solid biofuels can play under different conditions. As a result of the review performed in Task IV.1, a report has been produced with the pros and cons of Quality Management systems as well as the basic ideas of a Quality Management system especially adapted to solid biofuels. (Deliverable IV.1.D1). After a description of the results, important conclusions are drawn for the design of a guideline for Quality Management with emphasis on Quality Assurance and Quality Control for solid biofuels, to be done in IV.2.

The observations of the review are concluded following in aspects particular relevant for solid biofuels as the need for adoption of a Quality Management systems for solid biofuels, defining the product quality, traceability, documentation and statistical control, test and sampling.

Need for adoption of a Quality Management system Varying quality of solid biofuel is something that consumers are aware of and consequently, large consumers very often test for parameters (moisture, ash), which are important for the value of the solid biofuel they receive. Usually, the consequence of the test result is some level of reduction in price for the reduced quality and in the extreme, rejection of the lot or batch when the quality is below a certain minimum. If a more constant product quality can be achieved, this may result in a higher product price.

Yet, it takes a strategic decision at a company management level before a Quality Management system can be introduced in any company. This decision needs to be based on the felt needs of the company’s clients, where the costs associated with the introduction of the quality system can ideally be earned back through a higher product price. The examples from Denmark have shown that the benefits do not always outweigh the costs. But Quality Management systems can also have other benefits to the management of the business, such as improved planning, avoidance of failures, and enhanced welfare of employees.

The above interviews show that ISO 9001 and ISO 14000 are the most commonly applied Quality Management standards. For this reason it is proposed that a guideline for Quality Assurance and Quality Control should be prepared using terminology as used in ISO 9000.



Defining product quality Product quality is defined in different manners, depending on the type of product, the size and type of end-user devices where it is used, natural variations that occur in the raw materials used and variations that incidentally occur in the production processes.

It is therefore concluded that product specifications should be defined on a case-by-case basis by the market itself, not by an external organisation. It is therefore suggested not to prescribe normative fuel parameters (these should always be defined) and informative fuel parameters (these should not necessarily be described but may be of interest to some customers) in this BIONORM work package.

Traceability Any solid biofuel can in theory be associated to a growing location or location where the raw material was obtained from. Securing ability to trace a unit of solid biofuel backwards to the specific location of origin will secure ability to trace a lot of other properties and characteristics about that unit of solid biofuel. Therefore this information may be of value to end-users.

In cases where virgin biomass materials are used for the production of solid biofuels (e.g. straw bales, wood chips), documenting growing location may be relatively easy to implement, e.g. by documenting this information in delivery notes. How-ever, one still needs to seriously consider how the origin should be defined to en-able the customer to draw a relation to product quality. For example, the distance of trees near roadsides or the type of soil the tree was grown on, may be more important for end-users to know than the geographical location in a country.

Being usually transported as bulk, solid biofuel will normally be documented by a delivery note per lorry load. In order to enable tracing back biofuel products throughout the production chain, it is recommended that individual batches are tagged with a unique batch identification number. If each company in the chain properly documents the origin(s) of the raw materials used for each batch of product, this system enables tracing back a product throughout the production chain back to the origin.

In cases where different types of inputs are blended or the production takes place batchwise, it may not be feasible to introduce a system of tracing back the origin of the raw materials used.

Documentation The system for solid biofuels should demand that the different owners through the provision chain are responsible for documenting and filing all relevant information on treatment, storing and transport of the biofuel under their specific ownership. It should then additionally demand that the current owner is responsible for making all information back through the provision chain accessible at all times via a batch number.

The batch size can be defined by the standard or the consumer, and in accordance with the specific type of solid biofuel and the characteristics of its production. It is important that the batch identification refers to an amount, which - from the very beginning of the provision chain - is naturally coherent. As mentioned above, it may not be possible in all cases to link product batches to the source of raw materials.



Statistic control, tests and sampling One important part of a Quality Mangement system is the design of an appropriate sampling and testing procedure that provides confidence to customers that promised fuel specifications are always met. This varies greatly for both the amount of analytic work and the amount of documentation, depending on the quality requirements and process means to influence product quality. For straw bales for example, it is accepted by the purchasing end-user that fuel parameters may vary as a result of weather conditions etc. Therefore no analysis or documentation is done in this case. For pelletised animal feed however, all production process parameters are continuously logged automatically and documented.

The costs of implementing a sampling and testing procedure within a biofuel production process will heavily depend on the selection of process parameters to be tested and the frequency of testing. In return, the need for sampling certain process parameters will heavily depend on the type of production process, variability in the raw materials used and the allowable product variations. The sampling and testing procedure will therefore need to be custom made.

Individual biofuel producers need to be aware of variations occurring in the process to design an appropriate sampling and testing strategy. For this purpose, one often needs to perform proficiency testing. In addition to insight in the real fluctuations occurring in the process, the company needs to have an indication of the individual errors occurring in sampling, sample reduction, result reading and repeatability between labs testing.

4.2 The implementation of Quality Assurance in companies – field trials

4.2.1 Approach

The field trials were divided in 3 main steps:

(a) to develop a methodology how to implement a Quality Management system with emphasis on Quality Assurance and Quality Control in companies, and to write a first draft of a guideline [3] for the development and implementation of such a system in companies

(b) to investigate the implementation in different companies throughout Europe

(c) to develop, evaluate and update a guideline for the development and implementation of Quality Assurance and Quality Control in companies

Within the scope of the field trials and evaluation it has been decided to implement the first draft guideline within different practical situations at the industrial premises of a range of producers, traders and users of solid biofuels by the common elaboration of company specific draft manuals. The manuals followed the instructions given in the guideline and reflect specific arrangements and activities of the respective company related to Quality Assurance and Quality Control.

It was ensured that the choice of the companies is well balanced i.e. the main types of European biofuels and conversion processes were covered, and attention was given to geographical distribution (southern European countries, as well as northern ones). Following companies have been selected for the field trials:



• Biowatti Oy, Finland

• Bruins&Kwast, The Netherlands

• C&T, Italy

• The National Forest and Nature Agency, Denmark

• EEE, Austria

• MANN Naturenergie, Germany

• NRGi Handel A/S, Denmark

• Skelleftea Kraft, Sweden

• SODEAN, Spain

• Vapo Oy, Finland

• Wood Flame, The Netherlands

4.2.2 Results

Guideline The field trials have shown the need for a guideline [3] with a methodology applicable by each operator. The guideline as one of the main outcome of WPIV has been therefore developed to provide information about Quality Assurance and Quality Control for solid biofuels, and to present a methodology, which helps operators in the solid biofuels industry to design an appropriate Quality Assurance system (including Quality Control) according to their respective demands. The final version of the guideline is the Deliverable IV.2.D4 [3].

Chapter 1 of the guideline sets out the reasoning for using a Quality Assurance system for solid biofuels, and Chapter 2 defines the intentions of this guideline. The terms used in this guideline are set out in Chapter 3, along with a description of the four pillars of Quality Management used within ISO 9000:2000. Chapter 4 forms the major part of this guideline and sets out a step-by-step methodology to help each operator within a supply chain of solid biofuels to design a manual for Quality Assurance. The methodology used in this guideline is compliant with ISO 9001:2000, which would become important if the company already applies an ISO 9001:2000 Quality Management system. Appendix 1 provides some guidance on the relevant parts of ISO 9001:2000, and further definitions used in this document are provided in Appendix 2. Additionally a list of relevant CEN Technical Specifications can be found in Appendix 3.

The document aims to assist all operators within the supply chains for solid biofuels to write an appropriate manual for Quality Assurance, according to specific needs. It provides a better understanding of the issues, thereby enabling the design of appropriate measures to control and assure quality. The language and methods used in this guideline are compatible with, but not limited to ISO 9000:2000 and ISO 9001:2000, the most commonly used Quality Management standard-family.

The system for Quality Assurance described does not require a system to be already in force in accordance with ISO 9001:2000, but it may be integrated into an existing one. If



the company does not yet have any form of Quality Management, this guideline can help to create adequate confidence between all of the operators active in a supply chain, without creating undue bureaucratic complexity. The document can also be seen as a means of bridging the gap between the generalised text of ISO 9001:2000 (which is applicable to almost all industries), and the specific needs of operators in the solid biofuel market. This guideline does not discuss adaptations to production processes, nor does it set any pre-conditions in respect of specific technologies or technological processes. Rather, it focuses on providing confidence to customers that quality-requirements are complied with.

Important terms The language and methods used in WPIV was compatible with, but not limited to ISO 9000:2000 and ISO 9001:2000. ISO 9001:2000 already provides an overview of the system requirements that should be included when designing Quality Management systems. Quality Management is based on four pillars, as shown in Figure 1. The application of these pillars and their different measures depends on the problem and question regarded.

Quality management (QM)

Quality control (QC)

Quality assurance

(QA)

Quality improvement

(QI)

Quality planning

(QP)

Figure 1: Main pillars of Quality Management according to ISO 9000:2000

Each of these quality tools has its own measures and approaches. The characteristic of the supply chain for solid biofuels places emphasis on Quality Assurance and Quality Control. Quality Assurance measures should (a) be simple to operate, (b) not cause undue bureaucracy, and (c) offer savings in costs to both producers and users.

Quality Control is important to assess the properties of the fuel achieved, but it does not directly affect the quality of a product. In the context of solid biofuels, Quality Control includes the selection and application of appropriate sampling and sample-reduction techniques, as well as test methods. However, the application of sample- and test-methods is expensive and should be applied carefully and not as a matter of routine. An appropriate Quality Assurance system can reduce the frequency of testing and costs accordingly. Wherever possible, means should be sought to exempt parties from unnecessary procedures. Nevertheless, procedures should be drafted so that all steps in the supply chain are fully covered.

Customer requirements Quality Assurance aims to provide confidence that the quality required by the customer is continuously fulfilled. The fulfilling of these requirements leads to customer satisfaction.



The supply chains of solid biofuels can consist of different process chains and/or process steps, which can be distributed among different companies or organisational units. In this definition, the customer is not always the end-user of a solid biofuel but the next operator (company or organisational unit) within the supply chain. Each subsequent process chain or process step within the supply chain can be involved in defining the demanded quality. This is illustrated in Figure 2 taking a common pellet supply chain as example. The considerations are exemplarily shown from the producers point of view.

Trade and delivery of solid biofuels

Reception of solid biofuel by the end user

End user Conversion unit

Packaging, labelling, storage

Processing/ of raw material at the plant, e.g.:

Preliminary storage

Removal of contaminants

Drying, chipping

Mixing/ blending Screening,

cooling

Raw material

Identification and collection of raw material

Preparation/transportation of raw material

Process chain considered

Subsequent process(es)

Previous process(es)

Figure 2: The coherences between previous and subsequent processes

The identification and collection as well as the preparation and transportation of the raw material are to be considered as previous processes in this example. The processing of the raw material to a solid biofuel, packaging, labelling and storage of the solid biofuel exemplify the processes chain considered, i.e. all activities at the company. Trade and delivery as well as the reception of the solid biofuel and the conversion processes at end users premises are demonstrating the subsequent processes in this example.

The customer requirements, i.e. demanded quality, include not just the quality of product but also the quality of the company’s performance, as shown in Figure 3.



Customer Requirements

Product quality Quality of performance

Figure 3: Distinction of customer requirements

Product quality. For solid biofuels to be accepted at the marketplace, it is important that the requirements of the customers, in terms of fuel properties, are fulfilled whether or not those requirements follow a fuel-Standard.

The quality of solid biofuels can be defined in terms of a number of key properties that relate to the suitability of the fuel for a specific use. The selection of these indicators can differ from case to case, depending on the foreseen application and the occurrence of natural variations in fuel characteristics under current production processes.

Quality of performance. This point can usefully be considered by the following questions:

(a) How does the company operate in terms of specific costs per unit of product?

(b) How does the company recognise and fulfil the customer needs?

(c) Is the work carried out both effectively and correctly?

Quality of performance therefore refers to documentation, timing and logistical issues. The documentation of information on raw materials, intermediate- and final products as well as production process parameters may be instrumental in safeguarding proper treatment of the material during the production process and later provide insight into weaknesses of the production process so as to constantly improve product quality.

The methodology to develop and implement Quality Assurance The methodology illustrated in Figure 4 can operators use to design an appropriate Quality Assurance system (including Quality Control) for a specific chain. The methodology ensures – besides an efficient control of the processes considered – also the control over the overall provision chain by an integration of previous and subsequent process steps of other organisational units. Each implementation to these different steps of the methodology within a specific company should be documented in a (site-)specific manual. This manual can serve as an appropriate tool to illustrate to different parties that all the processes and their interaction are fully under control.



1. Description of process chain

2. Determination of customer requirements

3. Analysis of quality influencing factors

4. Selection of appropriate QA measures

Elaboration of a process (site)-specific manual

Figure 4: Methodology to apply and implement Quality Assurance

Step 1: Description of process-chain. The first step in the production of a site-specific Quality Assurance manual is to describe the production process in the company. As this forms the basis for a more detailed assessment on where quality is/can be influenced, a sufficient level of detail of this process description is desired. Hence the description of the production process including the different sub-processes is of vital importance for the following drafting of the Quality Assurance manual. Ultimately the production process could be described at a very detailed level where everything thinkable is taken into consideration. That would not serve any purpose except for a detailed description. It should therefore carefully be assessed how the purpose of an appropriate Quality Assurance with the description in question can be realised.

Step 2: Determination of customer requirements. The position of the process units(s) considered within the overall provision or process-chain is essential for Quality Assurance because the customer requirements depends on previous and subsequent process steps, as shown in Figure 2. In this regard the customer is not always the end-user of the final product. Consequently not in each case the product requirements according a CEN standard or another Technical Specification have to be fulfilled. However, in principle the requirements of the next operator of the overall provision chain have to be fulfilled. In case the end user of the fuel through a lack of knowledge is unable to provide adequate Technical Specifications of the fuel, key properties of the final product should be determined either by a standard that specifies which fuels can be used, or by the fuel supplier that through knowledge of the user needs is able to define such specifications. Such an indirect demand is to be understood as requirement from a subsequent step.

Step 3: Analysis of quality influencing factors. After the description of the process unit(s) and the analysis of customers requirements the process should now be examined concerning factors considered to be most influential in terms of fuel quality. The following factors are involved in general in determining fuel quality and refer to the management of the company:

• Effectiveness of preliminary inspections of fuel sources. This aspect is of importance to establish the suitability of the sources in general.

• Effectiveness of checking of incoming loads. This aspect ensures that the loads appear to be of one of the types already identified as suitable, and that the delivery notes are in harmony with the defined preconditions.



• Appropriateness of applied methods to handle, store and process materials. The necessary buildings and equipments have to be determined and composed to produce a solid biofuel with properties according to the needs of the customers.

• Quality-control measures adopted. The frequency of testing should be managed so as to accomplish the appropriate level of control at the lowest level of cost.

• Company management and responsibility. The responsibility for the quality of the biofuel within the overall provision chain is transferred to the next operator according the agreement of the different parties involved into this chain.

• Qualification and knowledge of staff. The staff needs to know about possible interactions between process-steps and the quality of the fuel, environmental regulations, and the relevant regulations on occupational safety and health.

Step 4: Selection of appropriate Quality Assurance measures. In accordance with the results of step 1 to step 3 appropriate Quality Assurance measures – i.e. measures giving confidence – has to be identified and applied. Following aspects are to be taken into account:

• Allocation of responsibilities

• Identification and documentation of Critical Control Points (CCP`s) and application of Quality Control (QC) measures

• Elaboration of work instructions

• Proper documentation of processes and test results

• System for dealing with nonconforming materials and products

• Training of staff

• System for complaint procedures

• Customer satisfaction and maintenance of the Quality Assurance system

• Preliminary inspection of raw material suppliers and formulating of acceptance criteria

• Enforcement of Quality Assurance meetings

• FME-analysis

Most relevant parts of ISO 9001:2000 for solid biofuels As already mentioned ISO 9001:2000 was analysed in terms of the applicability for solid biofuels. The review of existing quality systems and the field trials have shown the relevance of some aspects of ISO 9001:2000. Therefore the requirements were interpreted for operators dealing with solid biofuels. It ensures that the method described in the guideline can be integrated in an existing Quality Management system according to ISO 9001:2000. For some of the issues listed in Table 2 a general approach is sufficient, while the approach for others is inherently associated with the specific characteristics of solid biofuels.



Table 2: Quality Management system requirements in ISO 9001:2000, with general or specific approach

Quality system requirements in ISO 9000-2000: General or specific approach for biofuel production chains?

4. Quality Management system: 4.1. General requirements General 4.2. Documentation requirements Specific for biofuels 5. Management responsibilities 5.1. Management commitment General 5.2. Customer focus Specific for biofuels 5.3. Quality policy General 5.4. Planning General 5.5. Responsibility, authority and communication Specific for biofuels 5.6. Management review General 6. Resource management: 6.1 Provision of resources General 6.2 Human resources Specific for biofuels 6.3 Infrastructure Specific for biofuels 6.4 Work environment Specific for biofuels 7. Product realization: 7.1 Planning of product realisation Specific for biofuels 7.2 Customer related processes Specific for biofuels 7.3 Design and development Specific for biofuels 7.4 Purchasing Specific for biofuels 7.5 Production and service provision General 7.6 Control of monitoring and measuring devices Specific for biofuels 8. Measurement, analysis and improvement: 8.1 General General 8.2 Monitoring and measurement Specific for biofuels 8.3 Control of nonconforming product Specific for biofuels 8.4 Analysis of data Specific for biofuels 8.5 Improvement General

4.2.3 Conclusions about the field trials

The field trials have clearly shown, that all parties that/who are active in the overall provision chain for solid biofuels (including the suppliers of the raw materials, the producer of the biofuels, the transporter(s) and perhaps one or more retailers) should be integrated within an overall Quality Assurance system. Such a Quality Assurance system should links the various parts, i.e. previous an subsequent process-units, so that there is clarity about the allocation of responsibilities, documentation and Quality Control, etc. When that has been accomplished, it should be possible to reduce costs by taking advantage of pooling information, such as test-results. The application of Quality Assurance measures enable therefore the reduction of costly measures of Quality Control and contribute to a more transparent and efficient biofuel market. Based upon the customer requirements and known strengths and weaknesses of a production process, a producer of biofuels can take appropriate measures to provide confidence that the desired quality is always met.



Based on the methodology described above an appropriate Quality Assurance system can be implemented easily in a company dealing with solid biofuels, which fulfils those demands. The methodology ensures thereby a system that is simple to operate without undue bureaucracy, and which offers savings in costs.

4.3 Proposal for a draft standard for Quality Assurance

4.3.1 Cooperation with WG2

The activities in Task IV.3 describe mainly the cooperation of WPIV and WG2, CEN/TC335.

Besides the work in the field trials there was a close cooperation with WG2 from CEN/TC 335. WG2 has seen the need of a guideline as supporting document for the Technical Specification (TS) “Fuel QA for solid biofuels” [4] and expressed their wish of supporting information to improve this TS. Thereupon WPIV commented the work of WG2 from their scientific point of view. Due to this channel of communication and some common membership a good linkage between the work of WPIV and WG2 could be achieved.

In WPIV it was agreed, that the proposal for a standard to be developed in Task IV.3 must list all the requirements necessary to reach the aim of the standard. It may also contain informative Annexes of value to the user of the standard. A standard may also well be accompanied by separate documents describing how to use and apply it in practise. Therefore, the proposal to be developed in Task IV.3 consists of the standard + Guideline (outcome of Task IV.2 [3]). By compiling input from WPIV members on WG2`s Technical Specification “Fuel-Quality Assurance” [4], WPIV could provide a real input to the development of this TS. Due to this close cooperation the proposal for a standard to be developed in Task IV.3 based upon the TS under development in TC 335/WG 2. It could found an agreement of a common document of the proposal of a standard and the TS from WG2.WG2 decided to publish the draft standard as an official CEN Technical Specification [4] whereas the Guideline will be published as e.g. CEN technical report. The outcome of Task IV.2 (the guideline as Deliverable IV.2.D4 [3]) will be further adapted to the TS from WG2 before it is published as a CEN technical report.

In the following, a description of the ideas and content of the proposal of a standard will be given.

4.3.2 Elements of the proposal of a standard

General information The overall aim of the proposal of a standard/CEN TS "Solid Biofuels - Fuel quality assurance" [4] is to guarantee solid biofuel quality through the whole supply chain from the origin to the delivery of the solid biofuel and to provide adequate confidence that specified quality requirements are fulfilled.

The solid biofuel supply chain usually consists of the following main stages, see Figure 5.



Raw material

Indentification and collection of raw material

Production/preparation of solid biofuels

Trade anddelivery of solidbiofuels

Reception of solid biofuel by end-user

End - user

Combustionunit or other conversionunit

Supply chain activities covered by this Technical Specification

Figure 5 Solid biofuel supply chain

The objective of this document is to serve as a tool to enable efficient trading of biofuels. Thereby the end-user can find a fuel that corresponds to his needs and the producer/supplier can produce a fuel with defined and consistent properties. The system for Quality Assurance described in CEN/TS may be integrated in a Quality Management system (e.g. ISO 9000 series) or it can be used on it's own to help the supplier in documenting fuel quality and creating adequate confidence between supplier and end-user.

Principle The proposal of a standard /TS covers quality assurance of the supply chain and information to be used in quality control of the product, so that traceability exists and confidence is given by demonstrating that all processes along the overall supply chain of solid biofuels up to the point of the delivery to the end-user are under control. The fuel shall be traceable to the first operator in the entire supply chain, exemplarily shown in Figure 6.

Packaged biofuels

Biofuels in bulk material

Biomassresourceowners

Biomassresourceowners

Biofuel producer/supplier

Biofuel retailer End-user

Biofuel producers

Biofuel producer=supplier

Biofuel supplier

End-user

Documentation of origin and source

Fuel QualityDeclaration

Fuel QualityDeclaration

End-user

Operator or owner

Operator

Operator

BIOMASS ORIGIN/SOURCE

SUPPLIER END-USER

CEN Quality Assurance for solid biofuels

Figure 6: Examples of the documentation of origin and source and fuel quality declaration in different biofuel supply chains



Quality Assurance and Quality Control measures Quality Assurance aims to provide confidence that a steady quality is continually achieved in accordance with customer requirements.

The methodology shall allow producers and suppliers of solid biofuels to design a fuel Quality Assurance system to ensure that traceability exists, requirements that influence the product quality is controlled and end-user can have confidence in the product quality.

Some documentation is mandatory while other documentation is voluntary. Mandatory documentation on quality assurance measures are:

• documentation of origin (traceability of raw material)

• steps in the process chain, Critical Control Points (CCP), criteria and methods to ensure appropriate control at CCPs, non conforming products (production requirements)

• description of transport, handling and storage

• quality declaration/labelling (final product specification)

The Quality Assurance measures distinguished between production and transportation requirements.

Quality Control considers the specification of origin and source, specification of traded forms, and determination of properties (testing and laboratories, sampling and sample handling, normative properties for different traded forms, informative properties and accuracy in determination of properties).

Annex A provides a general methodology for operators according to the instructions given in the guideline elaborated in Task IV.2. Annex B – D contains examples of documentation to fulfil the requirements set on Quality Assurance and Quality Control.


The review in Task IV.1 [2] has clearly shown the need of the adoption of a Quality Management system with emphasis on Quality Assurance and Quality Control. This should be specifically adapted to solid biofuels. Thereby aspects as defining the product quality, traceability, documentation, statistical control, testing and sampling are of importance.

The field trials in Task IV.2 have identified the demand on a general methodology applicable by each operator. It is thereby of great significance, that each operator applies such an adapted Quality Management system referring to his own needs and in accordance with the requirements and recommendations given in the guideline and in the proposal of a standard/TS for Quality Assurance.

The development of the proposal of a standard in Task IV.3 has shown the need of a supporting document, i.e. a guideline [3] as developed in Task IV.2. In such a document - besides instructions how to fulfil the requirements of the standard - general information is



needed about Quality Assurance and Quality Control explaining the approach and benefits of implementing an adapted Quality Management system.

6 Recommendations

It is recommended to apply both the proposal of a standard/Technical Specification for Quality Assurance and the Guideline in practice. Furthermore it is recommend to investigate aspects of Quality Planning and on interactions of the various quality related activities in the quality policy of a company.

7 Acknowledgements

The work package leader (IE) acknowledge the partners (GLR, TNO, EE, VTT, KCL) and host companies (Biowatti Oy, Bruins&Kwast, C&T, The National Forest and Nature Agency, EEE, MANN Naturenergie, NRGi Handel A/S, Skelleftea Kraft, SODEAN, Vapo Oy, Wood Flame) of WPIV of the BIONORM project for their cooperation and contribution.

8 References

1 prCEN/TS 14961: Solid biofuels – Fuel specifications and classes; draft Technical Specification, CEN/TC 335, July 2004

2 Koppejan, J.; Jansen, J.P.: BIONORM WPIV: Quality systems for solid biofuels: Task IV.1. Review of existing systems, WPIV BIONORM-project (ENK6-CT2001-00556), October 2002

3 Langheinrich, C. et al.: Solid biofuels - A guide for Quality Assurance System. BioNorm – WPIV BIONORM-project (ENK6-CT2001-00556), 2004

4 N111 Draft 10: Solid biofuels – Fuel Quality Assurance; draft Technical Specification, CEN/TC 335, November 2004



9 Glossary

Supply chain: Defined as the overall process of handling and processing raw materials to a point of the delivery to the end-user.

Operator: Body or enterprise, which is responsible for one or several activities in the solid biofuel supply chain. The operator can be for example a biofuel producer or a subcontractor to the biofuel supplier. The first operator is a body or an enterprise which operates at the beginning in the overall supply chain.

Customer: Next operator within the supply chain (thus not in each case the end user).

End user: Consumer (private person, enterprise, utility, etc) using the fuel for energy purposes.

Producer: Enterprise or a body responsible for the production of the fuel or any operation with the purpose of changing the biofuel properties. The producer can also be the supplier of the fuel.

Supplier: Enterprise responsible for supplying solid biofuels. One supplier may be in charge of fuel deliveries from several producers as well as delivery to the end user.

Point of delivery: Location specified in the delivery agreement, at which the proprietary rights over, and responsibility for a fuel batch is transferred from the supplier/producer to the end-user.

Critical Control Point: Point within or between processes at which relevant properties can be most readily assessed, and point that offer the greatest potential for Quality Improvement; abbreviation: CCP.

Manual: Process- or site-specific document reflecting all activities related to the Quality Assurance system implemented and applied in practise.

Fuel specification: Document, which specifies the quality class, properties as well as origin and material composition.

Delivery agreement: Contract for fuel trade, which specifies e.g. origin and source, quality and quantity of the fuel, as well as delivery terms.

Fuel quality declaration: Document dated and signed by the supplier, specifying origin, traded form and properties of delivered solid biofuels.

Part 2 Research exchange with NMS/NAS


Part 2.5 Research exchange with NMS/NAS

Final report prepared by: Franziska Müller-Langer1) Contributing co-authors: Radka Fikova2), Michaela Hein1), Marzena Hunder6), Martin Kaltschmitt1), Galina Kashkarova4), Bela Marosvölgyi7), Vladimir Neužil3), Nerijus Pedisius5), Marcin Pisarek6) Radi Radev2), Peteris Shipkovs4), Vaclav Vacek3), Andrea Vityi7), Stranislovas Vrubliauskas5), Grzegorz Wisniewski6)

1) Institute for Energy and Environment gGmbH, Germany 2) Central Laboratory of General Ecology – BAS, Bulgaria, 3) KONEKO marketing, s.r.o., Czech Republic 4) Latvian Academy of Science - Institute of Physical Energetics, Latvia, 5) Lithuanian Energy Institute, Lithuania, 6) EC Baltic Renewable Energy Centre, Poland 7) University of West Hungary (NYME) - Department of Energetics, Hungary

1 Summary

Work package VI (WP VI) of BioNorm dealt with the national conditions of New Member States (NMS) and Newly Associated States (NAS) respectively, and their research exchange with the countries of EU 15. This work package is aimed to primarily increase the information flow between NMS/NAS and the pre-normative work of BioNorm. Therefore, country reports were prepared by the partners from the NMS/NAS (i.e. Bulgaria, Czech Republic, Latvia, Lithuania, Poland and Hungary) with focus on solid biofuels as energy source, existing standards and guidelines, needs for standardisation as well as recommendations. Moreover, national platforms were established as an interface between project consortium and solid biofuel standardisation bodies and involved companies of the NMS/NAS. The results are briefly summarised in this final report.

Recent trends have shown a continuous increasing international market of solid biofuels. This emerged market also stimulates the domestic production of refined solid biofuels within the considered NMS/NAS. Nowadays, these countries already use biomass (predominantly for domestic heat provision) and posses of promising potentials of solid biofuels and for bioenergy utilisation in order to reach national RES targets. However, currently limited experience in utilisation of refined solid biofuels and missing R&D contributes to a lack of solid biofuels standards and quality assurance guidelines. In all of the NMS/NAS there are no specific biofuel standards and quality assurance implemented yet. Companies that produce refined solid biofuels for export currently apply national standards of the import countries, mainly the Austrian, German and Swedish pellets and briquettes standards. Referring to this, all NAS/NMS-partners clearly stated that common standards are urgently required for increasing the solid biofuel market. Coupled with this, the European standards currently being developed by CEN need to be quickly adopted. Moreover, further cooperation is recommended with the European countries, that already used biomass efficiently and their legislation and biofuels standards are harmonious developed.



2 Objectives

Especially in the New Member States (NMS) and Newly Associated States (NAS) solid biofuels can significantly contribute to reach directive targets of the European Union increasing the share of renewable energy sources within the primary energy demand.

The background of WP VI was to give NMS and NAS the chance to participate in the forthcoming international market of solid biofuels in an early stage and to implement a continuous information exchange in both ways. The partners (i.e. Bulgaria, Czech Republic, Latvia, Lithuania, Poland and Hungary) have served as an interface between the pre-normative work of BioNorm and the NMS/NAS. This has provided the NMS/NAS with the opportunity to actively take part in the development of European standards for solid biofuels. Consequently, it is ensured that their concerns as well as their requirements were considered within the process of standardisation. Furthermore, this also contributed to raise the acceptance of renewable energy sources within the NMS/NAS and lead to a stronger integration of these countries into the coming European biofuel market and the European Community altogether.

The aim of WP VI was achieved in two tasks.

• In the first Task VI.1 the BioNorm partners elaborated country reports referring to the situation of solid biofuels in their respective countries in general and the needs for standardisation in particular.

• The second Task VI.2 dealt with national platforms in order to establish a continuous information exchange between the project consortium and all affected bodies of the NMS/NAS that are involved in pre-normative and standardisation activities.


In WP VI, no fuels were investigated in terms of analysis of biogenious materials. According to the WP-tasks, the appraisal was focused on national situations of the NMS/NAS in the field of bioenergy.

Since estimations of present data were uncertain, the responsible partners calculated the consumption of bioenergy, the production of solid biofuels as well as the trade within Europe. Bioenergy potentials and productions were analysed for agriculture residues, forestry residues and by-products of wood-industry.


The main results achieved in this work package are presented as follows according to the two tasks. For more detailed information please also refer to the deliverables presented by the respective country reports.



4.1 Country Reports

As mentioned above the country reports reflect the national situation in the NMS/NAS with the emphasis on the use of bioenergy in the national energy system, legal framework conditions, as well as existing standards and needs for standardisation. Starting with a general overview of the market situation in the NMS/NAS (section 4.1.1), afterwards the respective country situations are considered in more detail (section 4.1.2 et seqq.).

4.1.1 Market Situation for Solid Biofuels in the NMS/NAS

Bioenergy potentials and production In the NMS/NAS available biomass resources, which are most important regarding the production of solid biofuels, are: (i) forestry residues (logging and thinning residues from young-forest clearings), (ii) agriculture residues (straw and far-less significant maize residues), and (iii) wood industry by-products, which provide as main raw materials for wood fuels production. Peat as solid biofuel is only a relevant in few specific regions.

With exception of the Baltic states, so far only a relatively small share of primary energy consumption in the NMS/NAS is covered by bioenergy. The highest share can be noticed in forestry-reach countries, consuming large volumes of firewood and wood industry by-products. From early 1990’s also forestry residues consumption as wood chips has started. Peat is extracted in Estonia, Latvia and Poland. However, the major part has been predominantly utilised for horticulture purposes. For energy purposes peat has only been utilised for domestic heating and export in Estonia and less significant in Latvia.

A summary of potential available for energetic applications as well as the current share of biomass on the total national primary energy demand and the primary bioenergy production is given in Table 1.

Table 1: Share of biomass, primary bioenergy production in selected NMS/NAS based on wood and straw fuels applied for households heating, DH and CHP (at the present time horizon) and potentials available for energy use [1, 3]

Primary bioenergy production

[PJ/a]

Potentials for energy purposes [PJ/a] Country

Share of biomass on total primary energy demand

[%] Wood Straw Forest residues

Wood industry by-products Straw

Czech Republic 0.5 - 18.8 15.4 Estonia 32.5 - 19.5 11 2.2 Hungary - 8.3 7.8 28.5 Latvia 29 12 0.014 9 9.8 8.2 Lithuania 8 27 0.12 20.4 4.7 5.8 Poland 4 162 1.45 23.4 74 108 Slovakia 0.2 12 - 3.7 9.4 7.7 Slovenia - 11 - 2.2 6.2 0 Romania - 115 - 4.1 5.2 13.5 Bulgaria - - - - 26 21.6 Total 91 173 - data are not available



Bioenergy policies and support To be in line with the directives of the European Commission, all NMS/NAS have included future targets to increase the share on renewable energy sources within the internal energy system as one of the main objectives in their national energy policies. Furthermore, objects of specific strategies are also (i) to extent environment protection obligations, (ii) to increase the independency of energy sources and the security of supply as well as (iii) to increase the efficiency of energy use. Besides this, also reforms in agriculture production and follow-up forestry management rules and changes are items of legislations. Bioenergy supporting instruments of the NMS/NAS are abstracted in the following Table 2.

Table 2: Summary of restrictions and policies supporting bioenergy production in the NMS/NAS [1]

Restrictions and supporting policies Country RES targeta

ISb TBc QSd FITe Czech Republic 4 % x x x Estonia - x x Hungary 7.2 % x x Latvia 12 % x x Lithuania - x x Poland 7 % x x x x Slovakia 4 % x Slovenia 12 % x x x Bulgaria - - - - - Romania - - - - - a share of RES on the total internal primary energy demand b IS – investment support c TB – tax break d QS – quota system e FIT – feed-in tariffs - data not available

Markets of solid biofuels Traditionally, domestic produced biomass has been used in the same geographical region. In recent years this pattern has changed and in most of the Eastern European countries the production of refined solid biofuels started primarily for export and the expected growing internal market.

It is noted, up to now there is a lack of reliable statistics on the field of productivity and trade flows of solid biofuels. This is in particular true for the NMS/NAS and thus do not allow comprehensive analyses and discussions of the market. In Table 3 production quantity of the solid biofuels pellets and briquettes of the NMS/NAS as well as those exports are summarised. According to this, currently largest amounts are produced in Poland, Latvia, Estonia and in the Czech Republic. In Lithuania an intensively increase of production could be recognised since 1999. For countries such as Hungary, Slovenia, Romania and Bulgaria production amounts were not available, although they are players on the European biomass market.



Table 3: Wood briquettes and pellets production in the NMS/NAS and export in 2003 [1]

Country Pellets and briquettes production [ktons/a]

Export to European countries [ktons/a]

Czech Republic 140 - 160 ktons 80 % (~ 120 ktons) are exported

Estonia 20 ktons of briquettes 120 ktons of pellets

most of briquettes and pellets are exported to Sweden and Denmark

Hungary - -

Latvia 172 ktons (35 companies) some part of briquettes is exported mainly to Scandinavian states

Lithuania 30 ktons briquettes 40 ktons pellets

export of 90 % briquettes (27 ktons) and almost 100 % pellets (~ 40 ktons) to Sweden and Germany

Poland 130 ktons pellets (16 companies) 64 ktons briquettes (15 companies)

90 % (175 ktons) of pellets and briquettes are exported to Sweden, Denmark, Germany as well as less-significant to Norway

Slovakia 27 ktons briquettes 5 ktons pellets

export of 75 % (24 ktons) to Austria, Denmark as well as less-significant to Germany, Italy and Sweden

Slovenia - - Romania - - Bulgaria - - Total ca. 760 ktons ca. 400 ktons - data not available

Bioenergy trade has gained rapidly with emphasis in Northern Europe where large-scale use of biomass fuels for energy purposes and enormous supply of recycled wood and forest residues can be noted. Most significant volumes of solid biofuels are traded from the Baltic states (like Estonia, Latvia, Lithuania, Poland) to the Scandinavian countries, especially to Sweden and Denmark. There is also a significant gross boarder trade between Germany and Austria as well as Poland, the Czech Republic and Hungary.

Solid biofuel quality assurance and standards So far quality assurance system and standards for the production of solid biofuels, handling and trade have not been implemented at national level in the NMS/NAS with the necessary scope and scale. Certain single standards exist e.g. referring to the wood processing industry, but not existing in the broader quality management system. Thus, it is assumed that the majority of solid biofuel production within the NMS/NAS still faces problems concerning the provision of quality biofuel products. This is due to less efficient processing technologies, low cost practices of by-products handling, storage and trade which decisive influence raw material quality and impurities control [1].

The lack of standards for biofuels creates problems on the biofuel market. Since the production of pellets and briquettes in the NMS/NAS and those exports to other European countries is staidly increasing, the compliance of solid biofuels quality properties are provided by the application of standards existing in the respective import country. For that the most common standards valid in various West European countries for pellets and briquettes property requirements are the Austrian Ö-Norm M7132, the German DIN 7135 as well as the Swedish SS187120-21.



4.1.2 Bulgaria

Biomass within the energy system Currently, in Bulgaria the interest of both manufacturers of solid fuels and of combustion installations in this project’s outcome is considerable. Based on the statistical data of the last years (2000 to 2002) a significant increase of total biomass consumption can be recognised. Also the production of briquettes made from biomass and the sale of biomass combustion facilities show an increase during the period of 2000 to 2004. With regard to this, the majority of biofuels is utilised in households. It was pointed out that there is a significant potential for the production of biofuels in the timber industry, agriculture and woodworking industry as at present only a part of it is used.

National energy policies A new national Energy Strategy (2002) lays down the basis for introducing market mechanisms and transforming the energy sector as well as improving Bulgaria's energy efficiency [4]. The Bulgarian parliament adopted two new legislative acts (named Energy Act in 2003 and Energy Efficient Act in 2004) that will boost the use of renewable energy sources, biomass included. Furthermore, National Programmes for Energy Saving and on Renewable Energy Sources have worked out. Those have to be implemented along with actions plans and in the period ending 2010. These involve an increase of biomass share for meeting energies requirements.

Biofuel quality and standardisation So far, there is no standardisation for solid biofuels in Bulgaria. In some of the briquette facilities standards of Austria (Ö-Norm) and implemented in the Czech Republic (see below) are used. Moreover, some of the coal standards are applied. The standardisation of solid biofuels will play important role for their more substantial involvement in Bulgaria’s economy.

The Bulgarian partner recommended that it is necessary to provide economic incentives for all participants along the entire chain, i.e. the production of biofuels, the utilisation of biogenious waste products as well as the trade with them and their consumption [5].

4.1.3 Czech Republic

Biomass within the energy system Similar to many other central European countries, the Czech Republic has a good supply of cheap coal based energy. So far, the use of biomass played only a minor role in the Czech Republic. The annual energy needs corresponded to only about 1.4 % from biomass (2.4 PJ). The majority of this demand is met by brown coal (62.1 %), by imported natural gas and oil (26 %) as well as by nuclear and hydroelectric energy (10.5 %). The total annual consumption of bioenergy is about 1.8 to 2 million tons. The mostly used kinds of biomass are wood residues and biodiesel oil. The annual production of standardised solid biofuels (i.e. pellets and briquettes) achieves 140 to 160 thousand tons, 80 % of them are exported to the European Union.

In the Czech Republic the potential for energetic utilisation of biomass is considerable. It has been estimated that about 80 PJ/a could realistically be utilised up to the time horizon



of 2010. Regarding this, for 2020 largest potentials are estimated for the electricity generation and for heat production from solid biofuels [3, 6]. Analysing trends of consumption, the State Energy Conception of the Czech Republic assume that for the biomass sector besides energy crops, supplemented by residues of forest and waste biomass will have to be supplied to the market for the production of about 2.2 TWh/a of electricity and about 50 PJ/a of heat [6].

National energy policies According to the passed commitment of Kyoto, the Czech Republic is prescribed to reduce the production of greenhouse gas emissions by 8 % compared to 1990 before 2010. The target of electricity based on RES to be achieved in 2010 is also 8 % [3]. However, at present most of the energy legislation is in preparation or amendment. This is e.g. true for a new Act on support of renewable energy sources within the implementation of the Directive 77/2001/EC and also for the Act on Energy Management within a National Programme. In terms of selected support of sustainable energy systems, the State Programme to Support Energy Savings and the Use of Renewable Energy Sources also give investment supports for common heat and power generation system based on biomass (particularly biogas). Furthermore, as of 2002 there is a minimum annually adjusted feed-in-tariff of about 8 €ct/kWhel for electricity from biomass (biofuels and biogas) CHP [3, 6].

Biofuel quality and standardisation There is a complex set of standards in the Czech Republic related to the quality of solid fuels. Although, these standards have been prepared solely for various kinds of coal, basically they cover all relevant aspects related to standards for biofuels. However, no Czech standards have been elaborated specifically for biofuels; only the separate quality standard exists being a part of the Czech “Ecolabel” for environmentally friendly products. Czech producers of biomass briquettes and pellets and foreign clients apply the various foreign national standards to manage quality of marketed solid biofuels. Most often Austrian, German and Swedish standards are used. Biofuels products are usually tested by independent and authorised laboratories.

Both Czech producers and importers of solid biofuels emphatically stated that standards of biofuel quality are urgently required for further enhance of their activities, preferably on an international (i.e. European) level. The development of Czech national standards (based on existing German and Austrian standards) has recently been terminated to comply with the internal rules of CEN. Furthermore, the Czech Standardisation Institute is ready to accept the resulting standards of the BioNorm project (CEN TC 335) [6].

4.1.4 Latvia

Biomass within the energy system Since Latvia is not rich in domestic primary energy sources the import of primary energy resources (primarily oil and natural gas) account about 65 to 70 %. This is also true to meet Latvia’s electricity demand. Up to 75 % of the electricity demand is provided predominantly by hydro power plants, to a small extent by CHP plants and marginally by wind generators. The rest is purchased from abroad (i.e. Russia, Estonia, Lithuania) reflected in a significant increasing share. On this background, to expand the use of local and renewable energy resources is of high importance for Latvia. Biomass and wind are



the most prioritised RES for use in electricity generation. Biomass residues from forestry and from wood industry are partly already used in Latvia, but rather of biomass is not utilised or employed in an inefficient way. Wood and peat are that domestic energy sources predominantly used as firewood for heat production in small and, as a rule, low-efficient boilers in private households. Wood briquettes are produced for a free internal market as well as for export to Scandinavia [3, 7].

Because forests cover about 46 % of the Latvian territory generally comparable high biomass potentials are available. The total potential biomass availability including firewood, forest residues, brushwood and wood processing residuals are estimated until the time horizon of 2020 to be about 6 million m³; the allowed cutting of the State Forest Service is assumed to be about 12 million m³. Technical potential of wood wastes for energy use was estimated as 6.9 to 10.8 PJ/a for the time period of 2005 to 2010 [7].

National energy policies The Energy Law (1998) regulates economic activities in Latvia’s energy sector according to European agreements. This law is focussed on ensuring all national energy demands are being met, improving efficiencies of energy production and minimising environmental impacts [7]. The RES electricity target of Latvia to be achieved in 2010 is about 49 % [3]. The amendment of the Energy Law (2001) phased out a double feed-in tariff by January 2003, regulations fixing the total capacity for installation and specific volumes for next year that are annually published. E.g. for biomass power plants this annual feed-in-tariff equals the double single average sales tariff for electricity. However, this tariff is only valid for generated surplus electricity (i.e. after consumption of own needs of the electricity producer) and restricted to eight years after plants start up. In 2004 the National Environmental Plan was worked out obligating to increase use of RES, in particular of biomass and defining the lack of more efficient biomass combustion technologies as a major problem.

Besides energy policies there are also a Law on Forest Management and Utilisation and the Forest Policy. Latvia’s forestry economy is obliged to achieve the target of a sustainable forest utilisation; i.e. to balance all the components of forestry output as well as to ensure an even consumption of forestry resources [7].

Biofuel quality and standardisation Although there are not specific biofuel standards in Latvia now, there exist already some standards concerning the wood-processing industry (e.g. quality classification of timber, durability of wood and wood-based products, criteria for the assessment of testing laboratories). However, wood pellets companies produced them according consumers quality criteria (i.e. size, moisture content, calorific heat value etc.). Dealing with related issues and for discussion of the biofuel standard, a national group involving representatives from Forest Department, Heat Producer Association, University as well as companies concerned was organised.

It is finally stated that the adaptation of EU solid biofuel standards is critical for Latvia. Accredited testing laboratories should be organised for scrutinising of solid biofuel quality. In the nearest years the legislation on energy sector should be improved and perfected, also focussed on biomass use [7].



4.1.5 Lithuania

Biomass within the energy system In connection with political and market change, the structure of the Lithuanian energy system has profoundly changed besides significant decrease in final energy consumption. Currently, households and transport are the largest energy consumers. The dramatic increase of fossil fuel prices and growing environmental requirements encouraged the increasing use of local and renewable energy resources. The share local resources, with solid biofuels as a large part of it, increased from 0.4 Mtoe in 1990 to 1.2 Mtoe in 2002, which is about 13.3 % of the total primary energy consumption [8].

E.g. wood chips and sawdust were started to use as a fuel for district heating boilers. Today total installed capacity of wood burning boilers makes up about 320 MWth. The total consumption of wood fuel in 2002 was equivalent to about 659 ktoe (27 PJ) with households as main consumer (78 %). Wood potentials are largely utilised, considering the total technical potential of wood fuel that is evaluated to be about 845 ktoe (35 PJ). The use of straw as biofuels also for district heating is relatively new in Lithuania. Straw burning boilers are installed with a total capacity of about 7 MWth, the fuel consumption in 2002 corresponded to about 3 ktoe (0.13 PJ). Different to the situation of wood fuel, generally the use of straw can be expanded because the technical potential of straw is estimated to be about 130 ktoe (5 PJ).

Similar to other East European countries, about 85 % of briquettes produced by local companies are exported to Sweden, Germany, Denmark and Norway. The remaining part is sold on a local market.

National energy policies One of the objectives in Lithuania’s National Energy Strategy (2002) is to increase the share of RES in the total primary energy balance of up to 12 % by 2010. The target of RES electricity is about 7 % in 2010. The new National Energy Efficiency Programme (2001) consist of five trends; one of the main goals of item “utilisation of local, renewable and secondary energy sources” deals especially with wood and straw fuels. Besides the laws on energy and electricity, there is also a Law on Biofuel (2000) to determine legal base for production and use of biofuels. The Forestry and Forest Industry Development Programme promote the use of wood based energy, closely in line with national and EU policies. Since February 2002 there are feed-in-tariffs (for power plants based on biomass: 5.8 €ct/kWhel in 2004) with no guaranteed time [3, 8].

Biofuel quality and standardisation For regulation of fuel quality no normative documents or standards are developed in Lithuania. Accounts between wood fuel (wood chips and sawdust) suppliers and consumers usually are performed related to volume regardless of fuel calorific heat value. This is also true for quality properties of straw.

Under the Ministry of Environment there is the Lithuanian Standards Board (LST) dealing with standard issues in the country. Following the suggestion of Lithuania’s national working group, since February 2004 a special technical committee (TC 71 “Solid Biofuels”) with the scope of “elaboration of standards in the area of quality testing methods for solid biofuels and solid recovered biofuels” was established. In the work of



TC 71 also translation and legitimation in Lithuania of CEN technical specification are included. Moreover, the Lithuanian partner clearly stated that such a classification system for solid biofuels should be as simple as possible and also applicable; i.e. not only laboratory test methods should be developed but on-site measurement methods also [8].

4.1.6 Poland

Biomass within the energy system The structure of primary energy production and consumption has been changing during last years. Today, it is still dominated by coal (63 % of total primary energy consumption) as well as increasing shares of both crude oil (21 %) and natural gas (12 %). The share of renewable sources is growing slowly though achieving 4 %. Biomass covers about 95 to 98 % of this balance and is considered to be the most promising renewable energy in Poland, for both electricity and thermal energy production. The total technical potential for biomass has been calculated at about 755 PJ/a at the time horizon of 2010, containing agricultural residues (i.e. straw and hay) at 195 PJ/a, forestry residues and wood at 42 PJ/a, industrial biogenious residues at 58 PJ/a as well as energy crops at 424 PJ/a. At present about 200 ha are used for the cultivation of energy crops; though estimations indicate that 1.5 million ha of arable land is potentially available. The utilisation of biogas will play a less important role than solid biofuels at about 34 PJ/a [3, 4, 9].

In Poland the production of wood briquettes started in early 1990’s, the production of wood pellets 2 to 3 years ago. At present there are 15 producers of briquettes and 16 producers of pellets on the Polish market. Pellets and briquettes are mostly exported into EU countries. So far, biomass is utilised in DHP and CHP systems, predominantly for heat provision.

National energy policies As recently at the end of the 1990s the political engagement in renewable energy development was increased. The Energy Law (1997) and, coupled therewith, the Energy Act marked a major milestone for the regulation of the Poland’s energy sector. It is aimed to achieve efficiency in the field of energy production, transmission, distribution and trade and establish a base for RES. In the Development Strategy for the Renewable Energy Sector (2000) Poland has established a target of 7.5 % of primary energy production from RES by 2010 and 14 % by 2020. For the Polish heat market the Thermal-Modernisation Act (1998) is an economic instrument that is focused on supporting the substitution of conventional energy sources by non-conventional sources, including RES. The amendments of the Energy Law (2000, 2003) resulted in new feed-in ordinance for electricity. The prices for electricity from RES are set by means of negotiations between an independent power producer and an energy utility. Regulated by open tenders the average prices which where paid in 2002 were about 3 €ct/kWhel for electricity from biomass and 4.9 €ct/kWhel for biogas [3, 4, 9]. Additionally, there is an Act on Biocomponents Use in Liquid Fuels and on Liquid Biofuels (2002) that share of bioethanol per litre of petrol should not exceed 5 vol.%.

Biofuel quality and standardisation In Poland only a single standard is available with reference to wood chips for energy purposes (Polish norm PN-91/D95009). It defines dimensions, possible decay and



impurities of wood chips. Similar to other NMS/NAS in Poland there is still a lack of methodology for classification of solid biofuels. Since there are no classes for biofuels on the Polish market, producers of solid biofuels has prepared their own classification or put other standards (e.g. German DIN, Austrian Ö-NORM) into practice depending on the customers. However, it is stated that one is observed that Polish qualitative expertises based on EU countries’ standards are dishonoured by some of the EU customers. With regard to measurements of fuels properties, in practice some Polish (PN), European (EN) as well as International (ISO) standards appropriate for municipal solid wastes or solid fuels (e.g. coal) are applied.

4.1.7 Hungary

Biomass within the energy system Hungary is net importer covering 70 % of the total energy demand of Hungary by import. 96.5 % of the national energy production originates from nuclear or fossil resources (i.e. primarily from natural gas and crude oil). The penetration of the RES in the Hungarian primary energy production is relatively small (3.5 %). Biomass accounts for the largest share of Hungary’s renewable energy consumption. Currently, more than 80 % of the heat energy production is provided by biomass, predominantly by fuel wood combustion. Forestry wastes and sawmill by-products are currently burnt in furnaces for heat provision to the forestry industry or applied for briquette production for sale. Nearly 40 % of the round wood production is utilised for energy purposes. Consumption of heat based on biomass (mainly solid biofuels) amounted about 302 Mtoe in 2001 [3, 10].

There would be potentially good opportunities for the use of biomass in Hungary. It is estimated that only 10 % of forestry wastes are currently being utilised for energy production [4]. In sum the technically potential is estimated to be about 30 to 38 million tons dominated by the shares of biogenious municipal waste as well as agriculture and forestry residues. The significant amount of forestry by-products could potentially be used to generate electricity on larger scale, or more completely for the heat provision with regard to residential and industrial needs. It is noted that there will be a great difference between the potentially usable amount and the actually used amount of solid biofuels. This is because those parts of the heat market, which would be essential for biomass, are almost totally covered by natural gas [10].

National energy policies The main objectives of the Hungarian energy policy are stately in the Energy Policy Principles and the Business Model of the Energy Sector (1999). The concept includes increasing the share of renewable energy sources in the primary energy balance to 5 to 6 %, almost doubling the current figure. The Conservation and Energy Efficiency Improvement Action Programme (1999) supports the establishment of RES-based power plants, heat plants and energy plantations by ensuring 100 % reimbursement and preferential credits. The National Strategy for Energy Saving and the Increase of Energy Efficiency up to 2010 (1999) and the National Climate Protection Strategy (2000) determine the RES trends. Besides, the National Renewable Energy Strategy is being in preparation.



The new Electric Power Act regulates the binding acceptance of electric power from renewables and waste. Introduced through that act, since January 2003 there is a guaranteed feed-in-tariff which is an indefinite term and independent of RES and technology specific. Current tariffs depend on demand (i.e. peak, off-peak, average) and are between 6.3 and 10 €ct/kWhel. The electricity act intends the feed-in-tariffs to be an intermediate solution, which should lead to a green certificate system [3, 4, 10].

Biofuel quality and standardisation As so far the utilisation of solid biofuels is at an initial stage in Hungary, there are no regulated trade and logistic system; and thus, no respecting quality management and official test methods for biofuels in practice. The Hungarian Standards Institution (HSI or MSZT) is employed in the nationalisation of the European standards of bio-fuels. There exists no standard related to solid biofuels yet. The only used internal biofuel standard concerning fuel specification is that of fire wood standard (MSZ 1220:1984). Besides this, standards of solid fuels (e.g. coal) are used. In case of international trade by means of export Hungarian briquettes and pellets producers apply the relevant standards of the importing country (i.e. Austrian Ö-NORM, German DIN).

Adequate amending of regulations, establishment of the missing conditions are especially important tasks for Hungary, because of mainly the integration of the domestic market and producing onto the European market. It is finally recommended that standards corresponding to the solid biofuels EU-regulation have to be integrated into the Hungarian law in a short time. The establishment of these standard and market methods is becoming a most important task for Hungary [10].

4.2 National Platforms

Accordingly to this task, national working groups that consist of all national bodies of importance in the field of solid biofuels have been established in the frame of the BioNorm project. Purposes of these groups were the detailed discussion of requirements and recommendations within the standardisation process. This was realised in three meetings, which were held in each of the partner countries.

As of initial intricacies, it has been emphasised that for obtaining the highest benefit for all participants involved the information exchange requires to occur in both directions; i.e., the information flow about the current biofuel situation as well as standardisation needs of the NMS/NAS should be implemented into the European standardisation process. However, simultaneously the standardisation bodies of the NMS/NAS need up-to-date details from the solid biofuel standardisation process currently taking place in Europe. This is due to assure that future standard requirements can be fully met from the very first and additionally to minimise national standardisation efforts. This also allows to match corresponding structures, methodologies as well as equipments to meet the requirements of the European standard.

Besides this, a major object of that national platforms was also to transmit and disseminate information from BioNorm project consortium and CEN TC 335 “Solid biofuels” to local organisations and institutions related with biofuels production, trade and utilisation.



5 Scientific conclusions and recommendations

The survey of the reported country situations has been revealed the following, which is true to a large extent for all the NMS/NAS considered within the BioNorm project:

• In general, it is required to boost the share of RES on the national primary energy consumption in order to increase the internal security of energy supply and be in line with respective European directives (e.g. climate change mitigation).

• Major promising potentials for the energetic application of biomass (particularly solid biofuels such as agriculture and wood residues) within the RES have been identified. This is primarily linked to local conditions (e.g. characteristics of land use, energy consumption structure and power supply services, level of technologies for biomass application and present consumption).

• Basically, increasing use of bioenergy is an item of national energy policies. However, for the broad biomass utilisation changes in national legislation systems are needed with regard to the harmonisation of laws and regularly documents with requirements of the EU (e.g. environment quality standards), the harmonisation of the taxes, prices and tariffs as well as requirements of external and domestic markets.

• The current limited experience in utilisation of refined solid biofuels and R&D contributes to a lack of solid biofuels standards concerning measurements fuel properties and quality assurance guidelines. In turn this contribute to a further lack of solid biofuels classification and thus of information about solid biofuels prices on the national market. Consequently, it seems to be quite more difficult to develop a diversified solid biofuels market.

• Albeit, it is assumed an emerged market (i.e. mainly abroad) for more developed solid biofuels will stimulated the production of wood briquettes and pellets in many of the NMS/NAS.

• Comprising, it was clearly stated by all NAS/NMS-partners as well as the members of the national working groups that common standards are urgently needed for increasing the market shares of solid biofuels in the NAS/NMS, particularly in terms of biofuels export. Moreover, the European standards currently being developed by CEN need to be quickly adopted by the NAS/NMS. This was expressed explicitly in all country reports. Today, several NAS/NMS used national standards of EU 15 countries, mainly the Swedish, German and Austrian pellets and briquettes standards. Finally, it was emphasised the importance to further cooperate with the European countries, where biomass is already used efficiently and the legislation as well as biofuels standards are harmonious developed.

6 Acknowledgements

The project team of the Institute for Energy and Environment is very grateful to the Polish Baltic Renewable Energy Centre, the Czech KONEKO marketing, the Lithuanian Energy Institute, the Latvian Institute of Physical Energetics as well as Hungarian NYME Department of Energetics and the Bulgarian Central Laboratory of General Ecology the for



their cooperation and support to a good success in the survey of this BioNorm work package and in particular for the respective country reports.

7 References

1 Pisarek, M.; Hunder, M.; Gańko, E.; Szklarek, M. & Śmilgiewicz, T. (2004): Markets for Solid Biofuels in the New Member States. Proceedings of the Conference “Standardisation of Solid Biofuels”, October 6-7, 2004, Leipzig, Germany

2 Pisarek, M.; Hunder, M.; Gańko, E.; Szklarek, M. & Śmilgiewicz, T. (2004): Markets for Solid Biofuels in the New Member States. Presentation at the Conference “Standardisation of Solid Biofuels”, October 6-7, 2004, Leipzig, Germany

3 Comission of the European Communities (2004): The share of renewable energy in the EU – Country Profiles Overview of Renewable Energy Sources in the Enlarged European Union. Commission Staff Working Document, May 2004

4 Austrian Energy Agency (2005): Country energy profiles information available from homepage at URL: http://www.eva.ac.at/enercee/bg/energypolicy.htm, Access in January 2005

5 Radev R. (2004): Country Report on solid biofuel quality situation in Bulgaria. BioNorm Deliverable Report WP IV, Central Laboratory of General Ecology – BAS, Sofia, Bulgaria, September 2004

6 Neužil, V. & Vacek, V. (2004): Country Report on conditions relating to solid biofuel quality in the Czech Republic. BioNorm Deliverable Report WP IV, KONEKO marketing, s.r.o., Prague, Czech Republic, December 2004

7 Shipkovs, P.; Kashkarova, G. & Lebedeva, K. (2004): Country Report on solid biofuels quality situation in Latvia. Institute of Physical Energetics Latvian Academy of Science, Latvia, October 2004

8 Vrubliauskas, S. & Pedisius, N. (2004): Country Report on solid biofuel quality situation in Lithuania. BioNorm Deliverable Report WP IV, Lithuanian Energy Institute, Kaunas, Lithuania, 2004

9 Hunder, M.; Rogulska, M.; Gańko, E.; Rutkowska-Filipczak, M. & Kunikowski, G. (2004): Country Report on solid biofuel quality situation in Poland. BioNorm Deliverable Report WP IV, EC Baltic Renewable Energy Centre, Institute for Building, Mechanisation and Electrification of Agriculture EC BREC/IBMER, Warsaw, Poland, 2004

10 Marosvölgyi, B. & Vitiyi, A. (2004): Country Report on solid biofuel quality situation in Hungary. BioNorm Deliverable Report WP IV, University of West Hungary (NYME) - Department of Energetics, Sopron, Hungary, September 2004



8 Glossary

CEN European Committee for Standardisation CHP combined heat and power (production) DHP district heat and power EU European Union MSZT Magyar Szabványügyi Testület (Hungarian Standards Institution) NAS Newly Associated States NMS New Member States RES renewable energy sources

Part 3 Management Report



Table of contents

1 Introduction and overview of BioNorm work plan .................................................. 1

2 List of deliverables....................................................................................................... 3

3 Dissemination and use of the results .......................................................................... 5 3.1 Functional analysis of project deliverables ........................................................... 5 3.2 Targeted audience for dissemination..................................................................... 5 3.3 Publications resulting from the BioNorm project ................................................. 6 3.4 Technological Implementation Plan.................................................................... 10

4 Management and co-ordination aspects .................................................................. 10 4.1 Performance of the consortium ........................................................................... 10 4.2 Comparison of initially planned activities and work actually accomplished ...... 11 4.3 Analysis of the project schedule.......................................................................... 12

5 Contact details of partners ....................................................................................... 13

6 References .................................................................................................................. 17

Annex - Technological Implementation Plan................................................................A-1 Executive summary ........................................................................................................A-1

Original research objectives ......................................................................................A-1 Expected deliverables ................................................................................................A-2 Project's actual outcome ............................................................................................A-3 Broad dissemination and use intention for the expected outputs ..............................A-4

Community added value and contribution to EU policies .............................................A-5 European dimension of the problem..........................................................................A-5 Contribution to developing S&T co-operation at international level. European added value ..........................................................................................................................A-5 Contribution to policy design or implementation......................................................A-5

Contribution to community social objectives ................................................................A-6 Improving the quality of life in the Community .......................................................A-6 Provision of appropriate incentives for monitoring and creating jobs in the Community ................................................................................................................A-6 Supporting sustainable development, preserving and/or enhancing the environment ...............................................................................................................A-6 Quantified Data on the dissemination and use of results...........................................A-7 Expected project impact ............................................................................................A-7

Overview of main project results ...................................................................................A-7

Part 3 Introduction and overview of BioNorm work plan


1 Introduction and overview of BioNorm work plan

Starting with an overview of the project work plan, within this management report the final status of BioNorm is considered with focus regarding to obtained deliverables (chapter 2), coordination aspects (chapter 3) as well as results dissemination and use (chapter 3).

The results within the BioNorm project represented in Part 1 (Synthesis Report) and Part 2 (Detailed Final Reports) were achieved by the elaboration of the BioNorm work plan. This plan is subdivided into six work packages (WP), which consisted of the following interrelated tasks:

• Work Package (WP) I: Sampling and sample reduction

∗ Task I.1 Investigation of methods for sampling biofuels

∗ Task I.2 Investigation of methods for sample reduction of biofuels

• Work Package (WP) II: Physical / mechanical tests

∗ Task II.1 Moisture content and bulk density

∗ Task II.2 Ash melting behaviour

∗ Task II.3 Particle size distribution and dimension

∗ Task II.4 Durability and raw density of pellets and briquettes

• Work Package (WP) III: Chemical tests

∗ Task III.1 Sulphur, chlorine and nitrogen content

∗ Task III.2 Major and minor elements

• Work Package (WP) IV: Fuel quality assurance

∗ Task IV.1 Review

∗ Task IV.2 Implementation

∗ Task IV.3 Draft Standard

• Work Package (WP) V: Advisory Board

∗ Task V.1 Advisory Board

• Work Package (WP) VI: Research exchange with NMS/NAS

∗ Task VI.1 Country report

∗ Task VI.2 National platforms

According to the goals of this pre-normative BioNorm project, there was a close cooperation with various working groups of CEN TC 335 “Solid Biofuels”. Within the following working groups a number of Technical Specifications have been elaborated:

Part 3 Introduction and overview of BioNorm work plan


• Working Group (WG) I: Terms and definitions

• Working Group (WG) II: Fuel classification and Quality Assurance

• Working Group (WG) III: Sampling and sample reduction

• Working Group (WG) IV: Physical/mechanical tests

• Working Group (WG) V: Chemical tests

With regard to the obtained BioNorm results the information exchange was well realised as many partners of the BioNorm project were at the same time members in the respective WG.

Part 3 List of deliverables


2 List of deliverables

The list of deliverables completed during the duration of the project, and which constitute contractual deliverables, is provided below (Table 1). Generally, no major deviations from the work plan, which affect the overall project progress, occurred in any of the work packages and/or tasks. The main part of the activities has been dedicated to elaborate the deliverable reports due within the respective reporting periods (please see below). In some cases the deliverables were postponed.

Table 1: List of deliverables of the individual tasks, which do not form integral parts of overall project reports (to be continued next page)

Delivery date

No. Deliverable title sched-uled

(DoW)actually

Dissem-ination levela

Remarks

WP I Sampling and sample reduction

I.1.D4 Task Report on the results of sampling trials - 36 PU

I.2.D4 Task Report on the results of sample reduction trials - 36 PU

no deliverable scheduled, results provided as report of annexes A-U

WP II Physical/mechanical tests

II.1.D1 Report on reference test methods for moisture content 18 28 RE

II.1.D2 Report on rapid on-site test method 24 30 RE II.1.D3 Report on bulk density test method 30 30 RE

II.1.D4 Guidelines for moisture content and bulk density testing 36 36 PU consists of 2

separate parts II.2.D1 Report on new methods for ash melting behaviour 24 26 RE II.2.D2 Report on improving of existing standard 24 23 RE

II.2.D3 Report on draft standards for ash melting behaviour 36 36 PU

II.3.D1 Report on size classification equipment 28 27 RE

II.3.D2 Report on sensitivity of size distribution testing to influencing factors 16 28 RE

II.3.D3 Report on testing equipment for particle dimensions 27 30 RE

II.3.D4 Report on interactions and fundamental findings for size distribution 28 36 PU

II.3.D5 Best practice guidelines and draft standard for size classification 36 36 RE

II.4.D1 Report on equipment for durability testing of pellets and briquettes 20 20 RE

II.4.D2 Report on equipment for density testing of pellets and briquettes 22 23 RE

II.4.D3 Report on interaction and fundamental findings in testing durability and density of pellets and briquettes

28 36 RE

II.4.D4 Best practice guidelines and draft standard for testing durability and density of pellets and briquettes

36 36 PU

Part 3 List of deliverables


Delivery date

No. Deliverable title sched-uled

(DoW)actually

Dissem-ination levela

Remarks

WP III Chemical tests

III.1.D3 Task Report describing the method evaluation, trials and the results obtained for sulphur, chlorine and nitrogen, including best practice guidelines

36 36 PU combined with

deliverable III.2.D5

III.2.D3 Best practice guideline for the production of homogeneous biomass samples 22 25 PU

III.2.D4

Report on the results of method comparison and method validation for major and minor element detection in biofuels including statistical evaluation

30 32 PU

III.2.D5 Best practice guidelines regarding validated and statistically proved methods for the detection of major and minor elements in biofuels

36 36 PU combined with

deliverable III.1.D3

WP IV Solid biofuels: Fuel Quality Assurance

IV.1.D1 Task report on pros and cons of Quality Assurance Systems 14 14 PU

IV.2.D4 Manual for implementation of quality systems 30 36 PU

IV.3.D3 Proposal for a standard on “Solid Biofuels: Fuel Quality Assurance” 34 36 PU

WP V Advisory Board (AB)

V.1.D3 Protocols of the Meetings of the Advisory Board to be written by the secretariat and distributed between the different members

2, 9, 18, 27, 36 - RE no AB meetings

held

WP VI Research Exchange with NMS/NAS

VI.1.D1 Country reports on biofuel quality situation in NMS/NAS 32 33 PU consists of 6

country reports a PU public, RE restricted (group specific by EC)

Part 3 Dissemination and use of the results


3 Dissemination and use of the results

3.1 Functional analysis of project deliverables

The main deliverables in form of reports and best practice guidelines obtained during the project lifetime are based on the method developments and improvements as well as on results and experiences. Since they provide a basis for the writing of standards by the different working groups of CEN TC 335 "Solid Biofuels" their direct application is guaranteed. Furthermore, best practice guidelines are designed for the end-users.

The development of an overall Quality Assurance system has a major function for the project. The Quality Assurance is the prerequisite for expanding the market. Currently, one of the major problems for creating a dynamic and sustainable market is still the extremely variance of biofuel quality among the producers. Thus, users are often reluctant to purchase biofuels which quality and composition cannot be guaranteed. This is also true for manufacturers of equipment for using solid biofuels.

With the country reports developed by the partners of the NMS/NAS, special emphasis is given to the current conditions of these countries. Due to their high specific biofuel potentials they have a strong interest to participate in the European biofuel market.

3.2 Targeted audience for dissemination

The work of BioNorm was of pre-normative nature. Since many partners were at the same time members of the different working groups of CEN TC 335 "Solid Biofuels", a close link between the project and the European standardisation committee was established, i.e. the first recipients of the project outcome are the members of CEN TC 335. Thus, the results to be obtained by this pre-normative research were and are directly contributed to the writing of accurate standards applicable in practice. These standards are seen as a major key to open the national biofuel markets and to support a European biofuel trade, thus leading to an increased use of solid biofuels. By extending the project with the research exchange with the NAS (meanwhile commonly NMS), it is ensured that these standards also comply with the needs of the NMS/NAS. In particular these countries dispose of high specific biofuel potentials. Another target group for the project results are producers, traders and users as well as companies upgrading solid biofuels. By now, contacts with this group have taken place within various tasks. Moreover, by implementing a guideline on fuel Quality Assurance and Quality Control within companies operating in the solid biofuel field, additional contacts to users have been developed during the project lifetime.

The sectors involved indirectly into an energy provision from biomass (i.e. local authorities, emission reduction agencies, consultancies, producer of combustion units as well as fuel provision devices) benefit from the results of BioNorm. Furthermore, the presentation of project results to a wider public was carried out at international conferences such as the 2nd Biomass World Conference, Rome, May 2004 and the conference of Standardisation on solid biofuels, Leipzig, October 2004. In addition, to transfer the



information gathered to parties interested scientific papers as well as further conference contributions have been written by the project partners.

3.3 Publications resulting from the BioNorm project

The publications and contributions listed as follows were reported by the partners during the project lifetime. According to the progress of the pre-normative project, most of them were released in the third project year 2004.

Alakangas, E. (2004): The European pellets standardisation. Proceedings of the European Pellets Conference 2004, 3- 4 March 2004, Wels Austria

Alakangas, E. & Lehtoranta, T. (2004): Forest residues production and utilisation chain for Forssa CHP plant in Finland, Field study - Vapo Ltd, (ENK6-CT-2001-00556), VTT Report PRO2/P6024/04, April 2004.

Alakangas, E., Levlin, J.-E. & Valtanen, J. (2004): Classification, Specification and Quality Assurance for Solid biofuels. 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10 – 14 May 2004 Rome, Italy, (poster presentation)

Alakangas, E. & Lehtoranta, T. (2004): Vapo Oy – Summary of company specific manual, WP4 BIONORM-project (ENK6-CT-2001-00556), VTT Report PRO2/P6027/04, October 2004

Alakangas, E.; Halonen, P.; Kuusisto, K.; Jäkälä, M. & Hirvikoski, T. (2004): Quality Assurance system manual for wood fuel entrepreneurs in Finland- Model quality manual; VTT Processes & Finnmetko , Finland, October 2004.

Bärnthaler, G., Zischka, M., Arich, A. & Obernberger, I. (2003): Survey of the state-of-the-art for the determination of major and minor elements in solid biofuels. Biomass & Bioenergy (submitted for publication)

Bärnthaler, G., Zischka, M. & Haraldsson, C. (2004): Determination of major and minor elements in solid biofuels. 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Bodlund, B. & Sjöberg, L. (2004): Standardisation of Solid Biofuels. 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Böhm, T. et. al. (2004): Guidelines for bulk density determination. Deliverable D4, Part 2, EU-Project "BioNorm" (NNE5-2001-00158), Task II.1

Englisch, M., Haraldson, C., Bakker, F., Westborg, S., Thomsen, E., Niebergall, K., Versterinen, R., Carrasco, J. & Agrifiotis, C. (2004): Determination of chemical key parameters in biofuels - Nitrogen, Sulfur, Chlorine. 2nd World Conference and Technology



Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Hartmann, H.; Böhm, T. & Bock, M. (2002): Measuring Bulk Density of Solid Biofuels. Proceedings of the 12th European Conference and Exhibition on Biomass for Energy, Industry and Climate Protection, 17-21 June 2002, Amsterdam, The Netherlands

Hartmann, H., Böhm T. & M. Bock (2002): Measuring size distribution of wood chips. Proceedings 12th European Conference on Biomass for Energy, Industry and Climate Protection, 17-21 June 2002, Amsterdam, The Netherlands

Hartmann, H., Böhm, T., Jensen, P.D., Temmerman, M., Rabier, F., Jirjis, R., Hersener, J.-L. & Rathbauer, J. (2004): Methods for bulk density determination of solid biofuels. 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Hartmann, H., Böhm, T., Jensen, P.D., Temmerman, M., Rabier, F., Golser, M. & Herzog, P. (2004): Methods for size classification of wood fuels. 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Hein, M. & Kaltschmitt, M. (2004): Pre-normative work on sampling and testing of solid biofuels for the development of quality assurance systems (BioNorm). 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Jensen, P.D., Hartmann, H., Böhm, T., Temmerman, M., Rabier, F., Jirjis, R., Burvall, J., Hersener, J.-L., Rathbauer, J., Calzoni, J.A., Boavida, D.H., Lecourt, M., Hudson, B. & Pisarek, M. (2004): Methods for Rapid On-Site Moisture Determination of Solid Biofuels. 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Kaltschmitt, M. & Hein, M. (2003): Pre-Normative Work on Sampling and Testing of Solid Biofuels for the Development of Quality Assurance Systems. EC workshop "Bio-energy enlarged perspectives". 16-17 October 2003, Budapest, Hungary

Koppejan, J.; Jansen, J.P. & Langheinrich, C. (2004): Design of a Guideline for Quality Assurance and Quality Control; draft document, WP4 BIONORM-project (ENK6-CT2001-00556), March 2003

Koppejan, J. (2004): Interim report about the progress of the elaboration of site-specific manuals, TNO; WP4 BIONORM-project (ENK6-CT2001-00556), Task 4.2 Field-trials, March 2004

Langheinrich, C. (2004): Interim report about the progress of the elaboration of site-specific manuals. IE; WP4 BIONORM-project (ENK6-CT2001-00556), Task 4.2 Field-trials, March 2004

Langheinrich, C. & Kaltschmitt, M. (2004): Development and implementation of quality assurance systems for solid biofuels. 2nd World Conference and Technology Exhibition on



Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Langheinrich, C. & Döring, S. (2004): MANN Naturenergie/IE – Company specific manual, WP4 BIONORM-project (ENK6-CT-2001-00556), August 2004

Larfeldt, J., Hjuler, K., Koukios, E.G., Laitinen, R. & Hofbauer, H. (2004): Evaluation and Identification of the Best Appropiate Methods for the Determination of Ash Melting Behaviour. 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Levlin, J.-E. & Alakangas, E. (2004): Classification, specifications and quality assurance for solid biofuels. 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Marosvölgyi, B. & Vityi, A.(2003): Biobrikett-Production aus Energiegas und aus Energie-Schilf / Experiments on briquettizing and firing energy plants 4. International Symposium "Materials from Renewable Resources", 11-12 September 2003, MesseCongressCenter Erfurt, Germany

Marosvölgyi, B., Vityi, A.(2004): The European standardization process on the biomass utilization for energy purposes, within the frame of BioNorm Project (A biomassza energetikai hasznosításához kapcsolódó szabványosítási tevékenység az EU-ban, a BioNorm projekt keretében ). MTA-AMB XXVIII. Research and Develpoment Symposium, January 2004, Gödöllő, Hungary (paper accepted)

Neužil, V., Šafařík, M., Vacek, V., Kopáč, P. (2004): Solid biofuel quality situation in the Czech Republic. 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Nitschke, M.(2004): Interim report about the progress of the elaboration of site-specific manuals, ELSAM Eng.; WP4 BIONORM-project (ENK6-CT2001-00556), Task 4.2 Field-trials, March 2004

Pike, D.C. & Langheinrich, C. (2003): Fourth draft, Task 4.2 Field-trials – draft manual for hosts. WP4 BIONORM-project (ENK6-CT2001-00556), July 2003

Pike, D.C. (2004): Interim report about the progress of the elaboration of site-specific manuals. GLR; WP4 BIONORM-project (ENK6-CT2001-00556), Task 4.2 Field-trials, March 2004

Shipkovs, P., Kashkarova, G., Lebedeva, K. (2003): Biomass for sustainable development in Latvia. RIO-3 - World Climate & Energy Event, Brazil, 2003. pp. 311-316

Shipkovs, P., Kashkarova, G., Rubina, M. (2003): Domestic fuels – fuel wood and other biomass for heat supply. UNO & UNDP, Riga, 2003. 50 p. (in Latvian). Accepted for publication



Shipkovs, P., Kashkarova, G., Rubina, M. (2003): Boiler houses conversion for domestic fuels using – fuel wood and other biomass. UNO & UNDP, Riga, 2003. 57 p. (in Latvian). Accepted for publication

Temmerman, M., Rabier, F., Jensen, P.D., Hartmann, H., Böhm, T., Golser, M., Tuomi, S. (2004): Comparison between two methods for wood pellets durability testing. European pellets conference, 3-4 March 2004, Wels, Austria. (paper accepted)

Temmerman, M., Rabier, F., Jensen, P.D., Hartmann, H., Böhm, T., Rathbauer & J., Carrasco, J. (2004): Methods for durability and particle density determination of pellets and briquettes. 2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

Valtanen, J. (2004): Interim report about the progress of the elaboration of site-specific manuals, KCL; WP4 BIONORM-project (ENK6-CT2001-00556), Task 4.2 Field-trials, March 2004

Vityi, A. (2002): Hiánypótló irodalom a biomassza hulladékok hasznosításáról (Suppletory scientific literature on the biomass utilization for energy purposes ). Agrárinfó – Megújuló energiaforrások, December 2002., VII. vol. 2. p.

Vityi, A. & Marosvölgyi, B. (2002): A biomassza hasznosítás gazdasági és technológiai elemzése – harkai tapasztalatok (Economical and technical analysis of the utilization of biomass - experiments at Harka ). XLIV. Georgikon Napok – Stability and Institutional system in the agro-economy - Scientific Conference, 26-27 September 2002, Keszthely, Hungary

Vityi, A., Marosvölgyi, B. (2002): Bio-briquettizing as a Way for Energy Use of Wood Chips and Other Waste Matters. Logistics of Wood Technical production in the Carpatian Mountains, 9-10 September 2002, Technical University, Zvolen, Hungary

Vityi, A., Marosvölgyi, B. (2002): Wood utilization for energy purposes in Hungary. FORNET FP-modul, 17 October 2002, NYME, Sopron, Hungary

Vityi, A., Marosvölgyi, B., Ivelics, R. (2003): A biobrikettgyártás alapanyagbázisának fejlesztése (Development of the stock-basis of bio-briquette production). XXI. R&D Symposium, 2002, Gödöllő, Hungary

Vityi, A., Marosvölgyi, B., Ivelics, R. (2003): Szűcs-Szabó László: A biobrikett-gyártás alapanyagbázisának bővítése újabb alapanyagok bevonásával (Development of the bio-briquette production by drawing new materials into production). XXI. R&D Symposium, 2002, Gödöllő, Hungary

Vrubliauskas, S., Kavaliauskas, A. (2004): Wood fuel quality problems. Heat Energetics and Technologies, 4-5 February 2004, Kaunas University of Technology, Kaunas, Lithuania

Vrubliauskas, S., Pedisius, N. (2004): Development of Solid Biofuels Usage in Lithuania and Quality Assurance Problems. 2nd World Conference and Technology Exhibition on



Biomass for Energy, Industry and Climate Protection, 10-14 May 2004, Rome, Italy (paper accepted)

3.4 Technological Implementation Plan

With the BioNorm Mid-Term Report a draft Technological Implementation Plan (TIP) has been submitted. Compared to this TIP, changes in project progress and actual outcome have been occurred. Thus, adjustments and revisions are deemed as necessary. For the detailed TIP please refer to the Annex below.

4 Management and co-ordination aspects

4.1 Performance of the consortium

Discussion and data exchange between partners as well as development of the BioNorm project web page were actively coordinated and/or supported by the BioNorm coordination throughout the whole project period.

In the following, meetings of importance for the ongoing BioNorm project are briefly summarised. Between these meetings within the reporting period, in all work packages and/or tasks meetings of the respective work package or task were held to discuss the results obtained as well as the further proceeding.

Kick-off meeting – February 2002 in Berlin For projects kick-off, all work package and task leaders participated in the management meeting in Berlin, Germany, on 26th and 27th February 2002. During this meeting, all relevant administrative aspects of the project as well as details of the work plan (e.g. selection of biofuels, methodical approaches, analysis methods for sampling and testing) were discussed. Agreement was obtained on further protocol harmonisation as well as periodic reports in terms of progress of the BioNorm project. Subjects of this meeting were also the flow of information that was maintained through meetings and reports on task- and WP-level. Already at that time, a strong need was seen to continue the work after the results of the BioNorm project are available.

Plenary meeting – October 2002 in Vienna In cooperation with the CEN TC 335 “Solid Biofuels” that held a meeting on 2nd October 2002, under participation of the majority of project partners a plenary meeting was held in Vienna, Austria, on 3rd October 2002. The work package leaders of the respective work packages presented the specific aims, the biofuels to be investigated and the integration into the CEN/TC 335 work. The task leaders presented the objectives of their tasks and subtasks, the work plans, partners involved, results achieved and the specific plans for the next months. Furthermore, this meeting dealt with the project schedule (i.e. planned and actual schedule) as well as financial and management issues.

Mid-term meeting – September 2003 in Brussels Almost all partners participated in the BioNorm mid-term meeting, which took place in Brussels, Belgium, on 25th and 26th September 2003, again in connection with the CEN TC



335 "Solid Biofuels" plenary meeting on 24th September 2003. The BioNorm mid-term meeting was organised by the BioNorm coordination and the CEN meeting centre in Brussels. The BioNorm coordination discussed general aspects of project management and coordination as well as individual financial aspects.

Final conference – October 2004 in Leipzig The BioNorm final conference took place within the scope of the international conference “Standardisation of solid biofuels – Status of the ongoing standardisation process and results of the supporting research activities (BioNorm)” in Leipzig, Germany, on 6th and 7th October 2004. This was performed again in cooperation with CEN TC 335 "Solid Biofuels", which had its plenary meeting on 5th October 2004. This conference provides an update of the latest developments within the standardisation process of CEN TC 335 “Solid Biofuels”. The focus was put on the presentations of the contents of the different standards under development at this time and on the results of the BioNorm project. The presented results of this conference can be gathered from the redacted proceedings.

Besides the scheduled meeting BioNorm was also presented at other meeting, such as:

• “Bio-Energy, Enlarged Perspectives” workshop organised by the European Commission in Budapest, Hungary, on 16th and 17th October 2003

• “2nd World Conference and Technology Exhibition on Biomass for Energy, Industry and Climate Protection” in Rome, Italy, on 10th to 14th May 2004

The BioNorm coordination, assisted by all partners, has implemented a homepage, which contains continuously updated information on the BioNorm project. For the internal use of the web site a confidential chapter has been implemented and furnished with a login and password. This confidential chapter contains information on meetings and reports on task, work package and project level. The URL of the web site is: http://www.ie-leipzig.de/BioNorm/Standardisation.htm.

For an updated list of contact persons of all partner institutions including complete addresses, please refer to chapter 5.

4.2 Comparison of initially planned activities and work actually accomplished

Initially, the BioNorm project consisted of the first five work packages (WP I to WP V) named above (cf. chapter 1). During the end of the first progress year this BioNorm scope was supplemented by a “NAS-BioNorm” by extending the work packages dealing with physical/mechanical (WP II, Task II.1) and chemical test methods (WP III, Task III.2) and by adding a new work package WP VI “research exchange with NAS“ (meanwhile mostly NMS). This extension aimed to raise the number of involved European countries and thus to ensure that the effects of the BioNorm project are of greater impact on the future European biofuel market. Coupled with this, costs will be reduced and energy sustainability will be improved by increasing supply and demand of standardised solid biofuels.



Besides these work packages, at the beginning of the project a further work package (WP V - “Contribution to IEA Task 28 and Advisory Board”) was intended. This WP V should ensure a good exchange on information and results of BioNorm project with the CEN TC 335 "Solid Biofuels" consisting of representatives of the European Commission and the IEA Task 28 by installing an advisory board. However, in agreement with the members of the advisory board, no meeting was scheduled since there was no necessity during the project. As mentioned above, the information exchange was well realised as many partners of the BioNorm project are at the same time members in various working groups of CEN TC 335. Due to the lack of the required matched funding by the IEA, which was the prerequisite for this task the execution of IEA Task 28 was terminated.

4.3 Analysis of the project schedule

In general, during BioNorm projects lifetime there were no major deviations from the project schedule affecting the overall project progress in any of the tasks. However, in some tasks the original schedule was required to be revised. The main part of the activities has been dedicated to elaborate the deliverable reports. In several cases the deliverables needed to be postponed. These adjustments were in consequence of the following reasons such as:

• extension of number of tested biofuels and investigated parameters

• extra round robin trials due to accomplished newly defined CEN TC 335

• additional tests to sample preparation (e.g. for chemical tests)

• unexpected complexity of field trials

Part 3 Contact details of partners


5 Contact details of partners

# Acronym Institution/Organisation Street name and number

Town/City Post Code Country Code

Name Telephone No Fax No e-mail

1 IE Institut für Energetik und Umwelt gGmbH

Torgauer Straße 116 Leipzig D-04347 D Martin Kaltschmitt

+49 341 2434 113 +49 341 2434 133

[email protected]


Torgauer Straße 116 Leipzig D-04347 D Michaela Hein +49-3412434 432 +49 341 2434 433 [email protected]


Torgauer Straße 116 Leipzig D-04347 D Christian Langheinrich

+49 341 2434 437 +49 341 2434 433 [email protected]

2 GLR Green Land Reclamation Ltd 30, Strawberry Vale Twickenham TW1 4RU UK Andrew Limbrick

+44 20 88 91 41 78

+44 20 88 91 40 14

[email protected]

2 GLR Green Land Reclamation Ltd Porthiddy Farm West, Berea

Haverford-west, Pembroke-shire

SA62 6DR UK David Pike +44 13 48 83 70 33

+44 13 48 83 75 88

[email protected]

3 CTI Comitato Termotecnico Italiano, Energia e Ambiente

Via Pacini 11 Milano I-20131 I Julio Calzoni +39 02 26 62 6522

+39 02 26 62 65 50

[email protected]

3 CTI Dibiaga – University of Ancona Via B. Bianche Ancona I-60128 I Giuseppe Toscano

+39 07 12 20 49 17

+39 07 12 20 48 58

[email protected]

4 LTW Technologie- und Förderzentrum Nachwachsende Rohstoffe

Vöttinger Strasse 36 Freising D-85354 D Hans Hartmann +49 81 61 71 38 97

+49 81 61 71 40 48

[email protected]

4 LTW Technologie- und Förderzentrum Nachwachsende Rohstoffe

Vöttinger Strasse 36 Freising D-85354 D Thorsten Böhm +49 81 61 71 5319

+49 81 61 71 40 48

[email protected]

5 INETI Instituto Nacional de Engenharia e Tecnologia Industrial, Departamento de Engenharia Energética e Controlo Ambiental

Estrada do Paço do Lumiar 22

Lisbon P-1649-038 P Dulce Boavida +35 121 716 51 41

+35 121 716 65 69

[email protected]

6 SFN Signalsfromnoise.com Ltd Buckingham-shire HP14 3XZ UK Roger Sym +44 118 942 77 03

+44 14 91 63 86 60

[email protected]

7 TNO TNO Environment, Energy & Process Innovation

Business Park ETV, Laan van Westenenk 501, P.O. Box 342

Apeldoorn NL-7300 AH

NL Peter Jansen +31 55 549 30 39 +31 55 549 32 87 [email protected]

7 TNO TNO Environment, Energy & Process Innovation

Business Park ETV, Laan van Westenenk 501, P.O. Box 342

Apeldoorn NL-7300 AH

NL Jaap Koppejan +31 55 549 31 67 +31 55 549 32 87 [email protected]

8 SLU Swedish University of Agricultural Sciences, Biomass Technology & Chemistry

Robacksdalen, P.O. Box 4097

Umeå SE-90403 S Jan Burvall +46 90 78 69 491 +46 90 78 69 404 [email protected]

8 SLU Swedish University of Agricultural Sciences, Department of Bioenergy

P.O. Box 7060 Uppsala SE-75007 S Raida Jirjis +46 18 67 25 24 +46 18 67 38 00 [email protected]

9 BLT Bundesanstalt für Landtechnik Rottenhauser Strasse 1 Wieselburg A-3250 A Josef Rathbauer +43 74 16 521 75 43

+43 74 16 521 75 45

[email protected]






10 KCL Finnish Pulp & Paper Research Institute

P.O. Box 70 Espoo FIN-02151 FIN Jouni Valtanen +35 89 43 71 456 +35 89 46 43 05 [email protected]

11 USTUTT Universität Stuttgart, Institut für Verfahrenstechnik und Dampfkesselwesen

Pfaffenwaldring 23 Stuttgart D-70569 D Sven Unterberger

+49 711 68 53 572

+49 711 68 53 491

[email protected]

12 OFI Österreichisches Forschungsinstitut für Chemie & Technik

Franz-Grill-Strasse 5, Arsenal, Objekt 213

Wien A-1030 A Martin Englisch +43 17 98 16 01 490

+43 17 98 16 01 480

[email protected]

12 OFI Österreichisches Forschungsinstitut für Chemie & Technik

Franz-Grill-Strasse 5, Arsenal, Objekt 213

Wien A-1030 A Philipp Koskarti +43 17 98 16 01 970

+43 17 98 16 01 480

[email protected]

13 TPS TPS Termiska Processer AB Studsvik Nyköping S-611 82 S Niklas Berge +46 155 22 13 79 +46 155 26 30 52 [email protected]

14 TUV TU Wien, Institut für Chemieingenieurwesen

Getreidemarkt 9/159 Wien A-1060 A Hermann Hofbauer

+43 1 58 801 15 970

+43 1 58 801 15 999

[email protected]

14 TUV TU Wien, Institut für Chemieingenieurwesen

Getreidemarkt 9/166 Wien A-1060 A Emmanuel Padouvas

+43 1 58 801 15 971

+43 1 58 801 15 999

[email protected]

15 NTUA National Technical University of Athens, Department of Chemical Engineering

Zografou Campus Athens GR-15700 EL Emmanuel Koukios

+30 210 77 23 191

+30 210 77 23 163

[email protected]

15 NTUA National Technical University of Athens, Department of Chemical Engineering

Zografou Campus Athens GR-15700 EL Stamatis Georgopoulos

+30 210 77 23 286

+30 210 77 23 163

[email protected]

16 FORCE FORCE Technology (former dk-TEKNIK )

Hjortekærsvej 99 Lyngby DK-2800 DK Klaus Hjuler +45 39 55 59 99 +45 39 69 60 02 [email protected]

16 FORCE FORCE Technology (former dk-TEKNIK )

Hjortekærsvej 99 Lyngby DK-2800 DK Susanne Westborg

+45 39 55 59 99 +45 39 69 60 02 [email protected]

17 UOULU University of Oulu, Department of Chemistry

PO Box 3000 Oulu FIN-90014 FIN Rsito Laitinen +35 88 553 1611 +35 88 553 1608 [email protected]


PO Box 3000 Oulu FIN-90014 FIN Heikki Ollila +35 88 553 1672 +35 88 553 1608 [email protected]


PO Box 3000 Oulu FIN-90014 FIN Minna Tiainen +35 88 553 1672 +35 88 553 1608 [email protected]

18 Ciemat Centro de investigaciones energéti-cas, medioambientales y tecnológi-cas, Departamento de Energías Renovables

Avda. Complutense, 22 Madrid E-28040 E Juan Carrasco +34 91 34 66 682 +34 91 34 66 037 [email protected]

19 ECN Energy Research Centre of the Netherlands

P.O. Box 1, Westerduinweg 3

Petten NL-1755 ZG

NL Frits Bakker +31 224 56 41 25 +31 224 56 84 88 [email protected]

20 CRA Agricultural Research Centre of Gembloux

Chaussée de Namur, 146 Gembloux B-5030 B Michaël Temmerman

+32 81 627 157 +32 81 61 58 47 [email protected]

20 CRA Agricultural Research Centre of Gembloux

Chaussée de Namur, 146 Gembloux B-5030 B Fabienne Rabier +32 81 627 169 +32 81 61 58 47 [email protected]






21 DFLRI Danish Centre for Forest, Landscape & Planning

Hoersholm Kongevej 11 Vejle DK-7100 DK Niels Heding + 45 45 76 32 00 +45 45 76 32 33 [email protected]

22 HFA Holzforschung Austria Franz Grill Strasse 7 Wien A-1030 A Michael Golser +43 1 798 26 23 62

+43 1 798 26 23 50

[email protected]

22 HFA Holzforschung Austria Franz Grill Strasse 7 Wien A-1030 A Paul Herzog +43-1 798 26 23 54

+43 1 798 26 23 50

[email protected]

23 FCA Forestry Contracting Association Ltd. Dalfling, Blairdaff, Inverurie

Aberdeenshire AB51 5LA UK Barrie Hudson +44 1467 65 13 68

+44 1467 65 15 95

[email protected]

24 FAT Eidgenössische Forschungsanstalt für Agrarwirtschaft und Landtechnik Tänikon

Tänikon Ettenhausen CH-8356 CH Jean-Louis Hersener

+41 52 33 82 525 +41 52 33 82 528 [email protected]

25 AFOCEL Association Forêt Cellulose, Laboratoire Bois-Process

Domaine de l’Etancon Nangis F-77370 F André Themelin +33 160 67 02 53 +33 160 67 02 56 [email protected]

25 AFOCEL Association Forêt Cellulose, Laboratoire Bois-Process

Domaine de l’Etancon Nangis F-77370 F Michaël Lecourt +33 160 67 02 49 +33 160 67 02 56 [email protected]

25 AFOCEL Association Forêt Cellulose, Station Territoriale Nord-Est

Route de Bonnencontre Charrey sur Saone

F-21170 F Nathalie Mionetto

+33 380 36 36 20 +33 380 36 36 44 [email protected]

26 TW Elsam Engineering A/S Kraftvaerksvej 53 Fredericia DK-7000 DK Helle Junker +45 79 23 30 93 +45 75 56 44 77 [email protected]

26 TW Elsam Engineering A/S Kraftvaerksvej 53 Fredericia DK-7000 DK Max Nitschke +45 70 27 28 76 +45 70 27 28 71 [email protected]

27 SKAB Skellefteå Kraft AB Kanalgatan 71 Skelleftea S-93180 S Seved Lycksell +46 910 77 29 53 +46 910 77 28 78 [email protected]

27 SKAB Skellefteå Kraft AB Kanalgatan 71 Skelleftea S-93180 S Anna-Karin Nilsson

+46 910 77 29 66 +46 910 77 28 78 [email protected]

28 VT-TUG TU Graz, Institut für Ressourcenschonende und Nachhaltige Systeme

Inffeldgasse 25 Graz A-8010 A Thomas Brunner

+43 316 48 13 00 13

+43 316 48 13 00 04

[email protected]


Inffeldgasse 25 Graz A-8010 A Georg Bärnthaler

+43 316 48 13 00 74

+43 316 48 13 00 04

[email protected]


Inffeldgasse 25 Graz A-8010 A Ingwald Obernberger

+43 316 48 13 00 12

+43 316 48 13 00 04

[email protected]

29 CPERI Centre for Research & Technology Hellas, Chemical Process Engineering Research Institute

6th km, Charilaou-Thermi Road

Thermi-Thessaloniki

EL-57001 EL Alexandros Konstandopoulos

+30 31 49 81 92 +30 31 49 81 90 [email protected]

30 VTT Technical Research Centre of Finland, VTT Processes

P.O. Box 1603, Koivurannantie 1

Jyvaeskylae FIN-40101 FIN Eija Alakangas +35 814 67 25 50 +35 814 67 25 98 [email protected]

30 VTT Technical Research Centre of Finland, VTT Processes

P.O. Box 1603, Koivurannantie 1

Jyvaeskylae FIN-40101 FIN Raili Vesterinen +35 814 67 25 74 +35 814 67 25 98 [email protected]






31 SP Swedish National Testing & Research Institute, Chemistry & Materials Technology

P.O. Box 857, Brinellgatan 4

Borås S-50115 S Conny Haraldsson

+46 33 16 56 65 +46 33 12 37 49 [email protected]

31 SP Swedish National Testing & Research Institute, Chemistry & Materials Technology

P.O. Box 857, Brinellgatan 4

Borås S-50115 S Benny Lyven +46 33 16 52 68 +46 33 12 37 49 [email protected]

32 IEC Institut für Energieverfahrenstechnik & Chemieingenieurwesen

Reiche Zeche Freiberg D-09596 D Bernd Meyer +49 37 31 39 45 11

+49 37 31 39 45 55

[email protected]

32 IEC Institut für Energieverfahrenstechnik & Chemieingenieurwesen

Reiche Zeche Freiberg D-09596 D Steffen Krzack +49 37 31 39 45 24

+49 37 31 39 45 55

[email protected]

33 IFE-A IFE Analytik GmbH Torgauer Straße 116 Leipzig 04347 D Michael Hanrieder

+49 341 24 34 616

+49 341 24 34 633

[email protected]

33 IFE-A IFE Analytik GmbH Torgauer Straße 116 Leipzig 04347 D Knut Niebergall +49 341 24 34 611

+49 341 24 34 633

[email protected]

34 CLGE Bulgarian Academy of Sciences, Central laboratory of General Ecology

Gagarin Street 2 Sofia BG-1113 BG Radi Radev +359 2 705 379 +359 2 705 379 [email protected]

34 CLGE Bulgarian Academy of Sciences, Central laboratory of General Ecology

Gagarin Street 2 Sofia BG-1113 BG Radka Fikova +359 2 73 61 37 +359 2 705 498 [email protected]

35 KONEKO KONEKO marketing Ltd Sojovická 2 Prague 9 – Kbely

197 00 CZ Vladimir Neuzil +420 2 6603 2471 +420 2 6603 2471 [email protected]

35 KONEKO KONEKO marketing Ltd Štichova 654 / 54 Prague 4 – Ha'je 149 00 CZ Vaclav Vacek +420 22 7292 5986

+420 22 7292 5986

[email protected]

36 IPE LAS Latvian Academy of Sciences, Institute of Physical Energetics

Aizkraukles Str. 21 Riga LV-1006, LV Peteris Shipkovs +371 75 53 537 +371 75 53 537 [email protected]

36 IPE LAS Latvian Academy of Sciences, Institute of Physical Energetics

Aizkraukles Str. 21 Riga LV-1006, LV Galina Kashkarova

+371 75 53 537 +371 75 53 537 [email protected]

37 LEI Lithuanian Energy Institute Breslaujos 3 Kaunas LT-3035 LT Stanislovas Vrubliauskas

+370 37 40 18 43 +370 37 35 12 71 [email protected]

37 LEI Lithuanian Energy Institute Breslaujos 3 Kaunas LT-3035 LT Nerijus Pedisius +370 37 40 18 64 +370 37 35 12 71 [email protected]

38 BREC EC Baltic Renewable Energy Centre ul. Rakowiecka 32 Warszawa PL-02-532 PL Grzegorz Wisniewski

+48 22 8484832 112

+48 22 8484832 113

[email protected]

38 BREC EC Baltic Renewable Energy Centre ul. Rakowiecka 32 Warszawa PL-02-532 PL Marcin Pisarek +48 22 8484832 127

+48 22 8484832 113

[email protected]

38 BREC EC Baltic Renewable Energy Centre ul. Rakowiecka 32 Warszawa 02-532 PL Marzena Hunder

+48 22 8484832 128

+48 22 8484832 113

[email protected]

39 NYME University of West Hungary, Department of Energetics

Ady Endre út. 5. Sopron H-9400 HU Bela Marosvölgyi

+36 99 518 188 +36 99 518 188 [email protected]

39 NYME University of West Hungary, Department of Energetics

Ady Endre út. 5. Sopron H-9400 HU Andrea Vityi +36 99 518 188 +36 99 518 188 [email protected]

Part 3 References


6 References

1 Kaltschmitt, M. & Thrän, D. (2001): Description of Work (DoW) - Pre-normative work on sampling and testing of solid biofuels for the development of quality assurance systems (BioNorm). Technical Annex of a Shared-Cost research project for the Fifth Framework Programme, October 2001

2 Kaltschmitt, M. & Thrän, D. (2002): Description of Work (DoW) - NAS Extension to Contract No. ENK6-CT-2001-00556, Pre-normative work on sampling and testing of solid biofuels for the development of quality assurance systems (NAS-BioNorm). Technical Annex, October 2001

3 Kaltschmitt, M. & Hein, M. (2003): Mid-term Report of the BioNorm project. August 2003

4 Kaltschmitt, M. & Hein, M. (2004): Second Progress Report of the BioNorm project. February 2004

Part 3 Annex - Technological Implementation Plan

BioNorm Project - Final Technical Report A-1

Annex - Technological Implementation Plan

Executive summary

Original research objectives Investigations have shown that there are methods for sampling and testing as well as for Quality Assurance applied for solid fossil fuels (e.g. coal). However, experience has shown that these methods are only partly applicable to solid biofuels. CEN TC 335 "Solid Biofuels" continued to develop standards on the basis of the best information currently available. Thus, there is a strong need for further research in some areas to improve the reliability of selected sampling and testing methods, and the strong need for the development of an overall Quality Assurance system for solid biofuels taking practical needs into consideration.

Therefore, the aim of this project was to carry out research to improve sampling and sample reduction procedures as well as physical/mechanical and chemical tests, and to integrate these results into a fuel Quality Assurance system covering the overall supply chain. Besides this, the results of the different tasks should be fed directly into the ongoing work of CEN TC 335 "Solid Biofuels" to accelerate the development of improved standards considering requirements in practice. Hence, the conveners of the different Working Groups of CEN TC 335 were deeply involved within this project.

According to the objectives named above, the project work plan of BioNorm has consisted of interrelated work packages (WP).

Sampling and sample reduction (WP I). Sampling and sample reduction variabilities are often much more significant than testing variabilities. Hence, investigations of methods for sampling as well as for sample reduction should be carried out to provide a representative bulk sample.

Physical tests (WP II). Testing of physical properties of solid biofuels is commonplace but the precision and reproducibility of the results is often very poor. Physical characteristics are often technically limited properties in production, transportation and end-use processes (predominantly combustion). Therefore, the need for rapid methods for either fuel acceptance or rejection has been expressed in practice. Thus, the accuracy of existing procedures had to be improved, respectively, new procedures had to be developed for moisture content and bulk density with a focus on rapid on-site test methods, ash melting behaviour, particle size distribution and dimension as well as durability and raw density of pellets and briquettes.

Chemical tests (WP III). Test procedures for chemical fuel characteristics are derived primarily from the analysis of coal. Hence, for determining the accuracy of existing procedures had to be improved respectively new procedures had to be developed for sulphur, chlorine and nitrogen content as well as for major and minor elements.



Fuel Quality Assurance (WP IV). Currently no Quality Assurance system exists which takes the whole supply chain of solid biofuels into account. To develop such an overall Quality Assurance system initially a review had to be made of existing quality systems. Then, a guideline providing a methodology on how to develop end implement a company specific quality system had to be improved during field trials carried out to gain experience from practice. As final result a proposal for a standard for Quality Assurance systems for solid biofuels was required.

Contribution to IEA Task 28 and Advisory Board (WP V). This work package should ensure a good exchange on information and results of BioNorm project with the CEN TC 335 "Solid Biofuels" consisting of representatives of the European Commission and the IEA Task 28 by arranging an advisory board.

Research exchange with NMS/NAS (WP VI). This work package was aimed to increase the flow of information between NMS/NAS and the pre-normative work of BioNorm. Therefore, country reports had to be elaborated describing the respective national situation in detail with focus on solid biofuels and present existing standards. National platforms had to be established serving as interface between the project consortium, the NMS/NAS standardisation bodies and companies of the biofuel field.

This pre-normative project had to be provided verified tests and procedures with high precision for sampling and sample reduction as well as the determination of physical-mechanical and chemical properties, which are strongly needed for the successful development of the bioenergy markets throughout Europe. Furthermore, BioNorm should lead to the development of a quality system for solid biofuels to ensure the availability of appropriate biofuel quality at reasonable costs at the plant gate.

Resulting from BioNorm, widely accepted standards contribute also to create a market for sampling and testing devices for solid biofuels. Moreover, the end use conversion technologies (e.g. combustion units) can be further optimised with regard to the different fuel qualities. The same applies also to machinery and devices needed for the production and provision of solid biofuels in order to overcome the distance e.g. between the forest and the combustion plant; i.e. logistical issues also have to be taken into account.

Expected deliverables BioNorm provided reports and best practice guidelines, which are based on sampling and testing method developments and improvements as well as on results and experience obtained during the project lifetime. Their direct application is guaranteed as the different working groups of CEN TC 335 „Solid Biofuels“ used them as basis for the writing of European standards. Moreover, the best practice guidelines were directly designed for the end-users.

Closely linked to this, Quality Assurance is one of the prerequisite for expanding the market of solid biofuels. One of the major problems for creating a dynamic and sustainable market is still the extreme quality variation of the traded biofuel among the producers.



Thus, users are often reluctant to purchase fuels of unknown quality. Therefore, the development of an overall Quality Assurance system plays a key role for extending the European biofuel market.

A review of national conditions within the NMS/NAS is given by respective country reports. Special emphasis is given to these nations, which have a strong interest in participating in the European biofuel market due to their high specific biofuel potentials.

Project's actual outcome The outcome of the pre-normative work of the BioNorm project includes among methods for sampling and sample reduction, improved and new developed methods and procedures for the determination of physical-mechanical and chemical biofuel properties as well as the development and implementation of a company specific Quality Assurance system. Coupled with this, basic recommendations are given to apply both the proposal of a standard/TS and the guideline in practice.

For sampling and sample reduction (WP I) of woody biofuels (i.e. chips, sawdust and pellets) and straw bales, the outcomes of the work includes recommendations for methods for the several combinations of sampling point, material and proposed test method to be selected.

In physical/mechanical testing (WP II), respective parameters (e.g. moisture content and bulk density, ash melting behaviour, particle size distribution, durability and particle density) were investigated for various kinds of solid biofuels covering a wide scope of fuels using selected test methods. Existing methods and equipment (e.g. applied for fossil solid fuels) were adapted and improved with regard to their accuracy, reproducibility and repeatability for solid biofuels. Moreover, new methods were developed. Both involves - if applicable - the development of suitable reference and rapid on-site methods. Practice guidelines have been elaborated for the best appropriate determination of moisture content, bulk density, particle size distribution, durability and raw density.

For the chemical tests (WP III), subsequent to sample preparation and homogeneity testing, a broad variety of available procedures for the determination of sulphur, chlorine and nitrogen content as well as on major and minor elements were carried out to identify the most promising methods suitable for solid biofuels. Further test series have been evaluated referring to method validation and sensitivity analysis. Best appropriate test methods have been involved in respective best practice guidelines and draft standards.

Within the work on fuel Quality Assurance (WP IV), the review has clearly shown the need of the adoption of Quality Management systems; aspects as defining product quality, traceability, documentation, statistical control, sampling and testing are of importance. The field trials have identified the demand on a general methodology applicable by each operator. Based on this, requirements and recommendation for a company specific Quality Assurance systems were demonstrated in the elaborated guideline and in the proposal of a standard/TS.

In agreement with the members of the advisory board, no meeting was scheduled since there was no necessity during the project. Due to the lack of the required matched funding by the IEA the execution of IEA Task 28 (WP V) was terminated. As mentioned above, the



information exchange was well realised as many partners of the BioNorm project are at the same time members in various working groups of CEN TC 335.

The basis for the research exchange with the New Member States (NMS) and Newly Associated States (NAS) has successfully been established (WP VI). For Bulgaria, Czech Republic, Latvia, Lithuania, Poland and Hungary country reports were elaborated with focus on solid biofuels as energy source (i.e. potential and use), existing standards and guidelines as well as needs for standardisation. Moreover, national platforms were established as an interface between project consortium and solid biofuel standardisation bodies and involved companies of the NMS/NAS.

The milestones and deliverables specified in the BioNorm work plan were mostly achieved in time, in some tasks minor delays which do not affect the overall project progress occurred. A good cooperation between the partners was given. In all work packages the results obtained have provided useful information required for the standardisation process of CEN TC 335 "Solid Biofuels".

Broad dissemination and use intention for the expected outputs The work of BioNorm is of pre-normative nature. Since many partners are at the same time members of the different working groups of CEN TC 335 "Solid Biofuels", a close link between the project and the European standardisation committee was established. Thus, the results to be obtained by this pre-normative research were directly contributed to the development of accurate standards. These standards are seen as a major key to unlock the national fuel markets and support a European fuel trade, thus leading to an increased use of solid biofuels and in turn contributing to address European policies concerning renewable sources.

By extending the project with the research exchange with the NMS/NAS, countries with high specific biofuel potentials, it was ensured that the ongoing standardisation process also applies to their needs. This extension further aimed to raise the number of involved European countries and thus to ensure that the effects of the BioNorm project are of greater impact on the future European biofuel market.

Another target group for the project results are producers, traders and users. By now, contacts with this group have taken place within various tasks. Moreover, by implementing a guideline on fuel Quality Assurance and Quality Control within companies operating in the solid biofuel field, additional contacts to users were established during the project lifetime.

The sectors involved indirectly into an energy provision from biomass (i.e. local authorities, emission reduction agencies, consultancies, producer of combustion units as well as fuel provision devices) are recipients that use the outcome of this project, too. Moreover, the results of this project are disseminated to a wider public at international conferences. In addition, scientific papers as well as conference contributions were written by the project partners to transfer the information gathered to parties interested.



Community added value and contribution to EU policies

European dimension of the problem The declared aim of the European Commission is to substantially increase the use of renewable energy sources and thus to reduce greenhouse gas emissions within the years to come. Because of its advantages concerning climate change and security of supply, energy from biomass has the potential to play a key role in reaching these aims. Developing highly sophisticated and widely accepted European standards is regarded as major step towards the market development and cross-border trade and thus the increased use of solid biofuels.

Since the project results are directly disseminated to CEN TC 335 "Solid Biofuels", BioNorm significantly support the European standardisation process. The research carried out within BioNorm provides appropriate methods for sampling and testing of biofuel properties and characteristics as well as methods for the development and implementation of company specific Quality Assurance systems. The integration of different countries from the NMS/NAS ensures that their standardisation requirements are taken into account.

Contribution to developing S&T co-operation at international level. European added value The pre-normative work on BioNorm was performed by the creation of a new consortium consisting of 39 partners from 20 European and NMS/NAS countries, which are experts with intensive experiences in the different fields of sampling and testing as well as Quality Management. This international partnership in BioNorm combines and increases knowledge with regard to the various aspects of research; thus improving the European added value. Moreover, the NMS/NAS partners established national platforms in their countries and contributed to the ongoing standardisation work with their specific needs and opportunities. A strong linkage to CEN TC 335 "Solid Biofuels" was guaranteed by the directly provision of the project results and as some partners are delegated to the CEN working groups, which aims to draft standards for solid biofuels. These standards - fitting to all circumstances given in Europe and thus are acceptable within the EU and the NMS/NAS - are a prerequisite for the development of fast growing biofuel markets throughout Europe.

Contribution to policy design or implementation BioNorm significantly contributes to EU policies by addressing the White paper (COM(97)599). Energy from biomass currently contributes to about 3.7 % (approx. 2.3 EJ) of the gross inland consumption in Europe. This share will almost have to be doubled within the forthcoming years to fulfil the White paper target of 8.53 % energy from biomass in 2010.

With regard to sampling and testing as well as Quality Assurance the results of BioNorm serve to promote the trade and use of solid biofuels within Europe and thus contribute to reach the set goals mentioned above. This will also support the EU in fulfilling their commitments in the post-Kyoto process. Additionally, BioNorm contributes to Key Action 5.1.1 "Cleaner fuels by substitution and treatment" as it aims to substitute the energy use based on fossil fuels and Key Action 5.3.3 "Improving the acceptability of renewables" by reducing the so called non-technical barriers of solid biofuel use.



Contribution to community social objectives

Improving the quality of life in the Community The sampling and testing procedures as well as the Quality Assurance within the whole supply chain for solid biofuels developed in BioNorm will allow to improve the provision and management of the solid biofuels in a more predictable and reliable, i.e. a more environmentally sound way. This will help to reduce the health risks of the employees involved in production, provision, trading and use of solid biofuels. A reliable solid biofuel Quality Assurance improves process control and thus contributes to increase end-conversion efficiencies, that in turn contribute to decrease air emissions. It will help to identify the best appropriate utilisation of different biofuels with respect to protect the environment. The users will be able to purchase fuels of a quality needed to meet the requirements of the equipment in use. In conclusion, the project makes a significant contribution in reducing pollutants toxic to humans and ecosystems throughout the overall supply chain. Besides, quality of life is also improved by the positive socio-economics effects.

Provision of appropriate incentives for monitoring and creating jobs in the Community The results of BioNorm will be used as basis for the standards to be developed within the different working groups of CEN TC 335 "Solid biofuels". These standards are essential to promote a market for solid biofuels. Resulting from this market development, new jobs in the business sectors of engineering, manufacturing, energy production, energy distribution and consultancy can be generated. Also, by stimulating biofuel production, more jobs will be created in forestry, agriculture and horticulture. Building up biomass using energy plants will create employment, especially in rural areas, since the implementation of small-scale biomass technology is labour intensive compared to the larger-scale energy production using fossil or nuclear power.

Supporting sustainable development, preserving and/or enhancing the environment Based on improved sampling and testing procedures as well as Quality Assurance of solid biofuels, BioNorm can provide the information needed to develop accurate and widely accepted standards. The defined fuel properties and an overall Quality Assurance system will allow the environmentally sound use of solid biofuels. By these measures the confidence in biofuels will increase and using biofuels as energy carriers will become more attractive, more easy and more predictable. The substitution of fossil fuel energy by energy from biomass will help to decrease emissions of greenhouse gases (i.e. environmental goals) and contribute to the security and diversification of energy supply (i.e. improvement of the energy supply security in Europe).



Quantified Data on the dissemination and use of results

Currently achieved quantity Estimated future quantity

Product innovations - - Process innovations 1 1 New services (commercial) - - New services (public) - - New methods - 5 Scientific breakthrough - - Technical standards to which this project has contributed 8 17

EU regulations/directives to which this project has contributed - -

International regulations to which this project has contributed - -

PhDs generated by the project - 4 Grantees/trainees including transnational exchange of personnel - 2

Expected project impact

EU policy goals scale of expected impact over the next 10 years

other other

-1 to 3* not applicable to project

project impact too difficult to estimate

improved sustainable economic development and growth, competitiveness

2

improved employment 2 improved quality of life and health and safety 2

improved education x improved preservation and enhancement of the environment 2

improved scientific and technological quality 2

regulatory and legislative environment 3

other * legend: 1 negative impact, 0 no impact, 1..3 small.. medium.. strong positive impact

Overview of main project results

This pre-normative research project was aimed to improve and develop sampling and sample reduction procedures as well as physical-mechanical and chemical tests, and based on these results to support the development of Quality Assurance systems covering the overall supply chain. Besides this, the results of the different tasks were fed directly into the ongoing work of CEN TC 335 "Solid Biofuels" to accelerate the development of improved standards taking practical needs into consideration.



Result Category Owner of results The research carried out in BioNorm contributes substantially to the development of standards to be compiled within CEN TC 335 "Solid Biofuels"

A (usable outside the consortium) All (whole consortium)

Publication of individual group results in scientific journals, magazines and other print or electronic media

A (usable outside the consortium) All (whole consortium)

Documents

Pre-Normative Work on Sampling and Testing of Solid ...virtualmaze.co.in/sample/Biofuels Info/Rhombus Power - biodiesel... · 11 USTUTT Institute of Process Engineering and ... 12