EUROPEAN WASTE WATER MANAGEMENT - AquaEnviro · PRACTICAL APPLICATION OF MODULAR OFF-SITE BUILD: A COMMISSIONING PERSPECTIVE 65 Baird, A., WPL Ltd, UK LIVERPOOL WWTW SBR CARBONACEOUS

TECHNICAL SESSIONS

• Nutrient Removal and Recovery

• Innovation

• New Technologies

• Control

• Preliminary Treatment and FOG

• Low-TOTEX solutions to meet future growth and quality drivers

• Priority Substances

3 - 4 October, The Royal Armouries, Leeds, UK

EUROPEAN WASTE WATER MANAGEMENT CONFERENCE PROCEEDINGS

www.ewwmconference.com www.aquaenviro.co.uk twitter.com/aquaenviro #EWWM

Supporting Organisations

Media Partners

Sponsor

www.ewwmconference.com

PUBLISHED BY:

Aqua Enviro,

8 Appleton Court

Calder Park

Wakefield

West Yorkshire, WF2 7AR

T. +44 (0)1924 257891

E. [email protected]

W. www.aquaenviro.co.uk

This electronic book contains the proceedings of the 11th European Waste Water Management

Conference. The contents of these are not necessarily the views of Aqua Enviro or those of the

Organisations who the Authors represent.

Unauthorised copying of this documents is not permissible.

The papers contained in these proceedings are for the dissemination of scientific and engineering

information and not for any other commercial or other purpose.

Authors of scientific and engineering technical papers may quote these papers using the following

citation

Name of Author (s): Title of Paper: Proceedings of the 11th European Waste Water Management

Conference, Aqua Enviro, October, 2017


11th European Waste Water Management Conference

3rd – 4th October 2017, Leeds, UK

2

TABLE OF CONTENTS

PROVIDING PHOSPHORUS REMOVAL FOR RURAL TREATMENT WORKS 3

Bowman, B. and Aboobakar, A., United Utilities, UK

GPS-X MODELLING TO OPTIMISE NITRIFICATION AND RISK ASSESS DESIGN PROPOSALS -

CASE STUDY OF BROCKHAMPTON SEWAGE TREATMENT WORKS 20

Ruswa, E.1, Chadha, M.1 and Copp, J.2, 1Severn Trent Water, UK, 2Primodal, UK

HARNESSING THE POWER IN NITRIFYING SAND FILTERS 41

Chan, T. F. and Koodie, T., Black & Veatch, UK

NUTRIENT REMOVAL WITH MICROALGAE – REDUCTION OF THE EFFLUENT CONCENTRATION FROM

WASTEWATER TREATMENT PLANT 50

Wawilow, T.1, Hasport, N.1, Theilen, U.1 and Thomsen, C.2, 1THM - University of Applied Sciences, Germany, 2Phytolutions GmbH, Germany

1-STEP® FILTER: THE SOLUTION FOR COST-EFFECTIVE REMOVAL OF PHOSPHOROUS AND OTHER

PRIORITY CHEMICALS FROM WWTP EFFLUENT 56

Kramer J.F.1, Menkveld H.W.H.2, Cunliffe, T.2, Merks, C1, 1Witteveen+Bos, the Netherlands, 2 Nijhuis Industries

UK and Ireland, UK

PRACTICAL APPLICATION OF MODULAR OFF-SITE BUILD: A COMMISSIONING PERSPECTIVE 65

Baird, A., WPL Ltd, UK

LIVERPOOL WWTW SBR CARBONACEOUS TRIAL 75

Akinola, O., Black, J., Sherwood, A., and Hornsby, J., United Utilities, UK

MODULAR UPGRADE OF AN ASP TO MEET RAPID POPULATION UPSURGE, NEW WASTE STREAMS

AND TOUGHER CONSENTS IN THE UK 89

Bassey, B.O.1, Njunbemere, N.1 and Ogarekpe, N.2, 1Coventry University, UK, 2Cross River University of

Technology, Nigeria

KINGSPAN CALLS FOR FOOD SERVICE INDUSTRY TO CONSIDER FOG AS FUEL 98

Curran, J., Kingspan Environmental, UK

TECHNICAL AND COMMERCIAL CONSIDERATIONS IN THE REMOVAL OF PRIORITY SUBSTANCES AS

SPECIFIED WITHIN THE EU WATER FRAMEWORK DIRECTIVE FROM TREATED DOMESTIC SEWAGE:

CASE STUDY FROM A 16,000 PE NON-RURAL WWTP 106

Parocki, D., Nkrumah-Amoako, K., and Campen, A., Arvia Technology, UK

ADVANCED OXIDATION PROCESSES AND NON-THERMAL PLASMA FOR THE REMOVAL OF

EMERGING CONTAMINANTS IN WATER 115

Tizaoui, C., and Ni, Y., Swansea University, UK

EARLY LIFE PERFORMANCE OF STONEYFORD INTEGRATED CONSTRUCTED WETLAND 121

Hall, L.1; Woodward, D. 1; McDermott, R. 1; McCurdy, D. 2; Griffin, J. 2; Crabbe, D.2 1Ulster University, 2 Northern Ireland Water, Northern Ireland

ON SITE LANDFILL LEACHATE TREATMENT: INVESTIGATIONS INTO ECONOMICAL AND

ENVIRONMENTAL SUSTAINABLE SYSTEMS FOR NORTHERN IRELAND 137

Devlin, Y.1, Nicholl, G.1, McRoberts, C.1, Johnston, C.1, Rosinqvist, D.2, Svensson, B.M.3, Mårtensson, L.3, 1Agri-

Food and Biosciences Institute UK, 2Laqua Treatment AB Sweden, 3Kristianstad University, Sweden



3

PROVIDING PHOSPHORUS REMOVAL FOR RURAL TREATMENT WORKS

Bowman, B.1 and Aboobakar, A.1 1United Utilities, UK

Corresponding Author Email [email protected]

Abstract

There is an ongoing phosphorus challenge to achieve “good ecological status” in waterbodies across North

West England. The size range of sites required to meet tight regulatory permits is increasing, in some cases

with phosphorus targets below 0.5mg/l. In recent years, much of the focus has been on technological

developments appropriate for large works, leaving a knowledge gap in how phosphorus removal can be

achieved sustainably in small, rural works.

For works with population equivalents under 1000, there are added complications of limited workforce, power

supply and access routes. Conventional approaches of chemical or biological nutrient removal may not be

appropriate for these situations and alternatives are not readily available. Utilising the concepts of a circular

economy we aim to deliver low maintenance, robust, sustainable phosphorus removal for small rural works. The

role of reactive media for phosphorus removal and regeneration or reutilisation routes may be critical to

delivering this goal.

Keywords

Phosphorus; Reactive media; Rural; Sustainable; WFD; WwTW

Introduction

The Water Framework Directive (WFD) has set an objective for wastewater treatment to achieve phosphorus

(P) removal for the prevention of eutrophication in waterbodies. Modelling suggests that in order to meet ‘good’

ecological status in river catchments within the United Utilities region there will be a requirement to provide P

removal at a large number of small treatment facilities (<5000PE). In addition, the permit requirements are

becoming tighter, in some cases below 0.5mg/l.

Conventional P removal at small treatment facilities has typically been achieved through chemical processes.

These processes contribute to significant operating costs and carbon emissions as well as increased customer

impact due to chemical deliveries. This leads to a disproportional cost/benefit for the intervention. Chemical P

removal operates through the generation of ferric or aluminium phosphate salts; these are largely unreactive

precipitates from which P recovery is not viable. In the interest of future resilience this route does not meet our

strategic vision.

Developments in technology to reach total P permit conditions less than 1mg/l have exploited economies of

scale and intensive processes suitable for large, urbanised treatment facilities. These have limited suitability to

rural catchments where a combination of catchment and end-of-pipe solutions has been shown to be the best

approach to driving overall water quality improvements. A new type of approach is needed to suit the

requirements for rural catchments which could also be an enabler to the recovery of phosphorus.

The Integrated Catchment approach

The basis of the Integrated Catchment (IC) approach is to consider a river as a complete system;

incorporatingthe water and wastewater catchments, full range of river users, point discharges and disperse

pollution. This systems thinking model has been applied to the River Petteril in Cumbria to form a case study to

roll out the approach across the region.

mailto:[email protected]



4

The River Petteril is a tributary of the much larger River Eden (which is a designated Site of Special Scientific

Interest), located in Cumbria, in the Solway-Tweed River Basin area of the North West Region. The length of

the river flows from Penrith in the south up to Carlisle in the north (confluence with the River Eden), parallel to

the M6 motorway. For much of the length of the river, the Petteril catchment is rural, consisting mostly of

agricultural activities, with Carlisle being the urban area of the catchment.

The current WFD ecological classifications of the Petteril and its tributaries range between “good” ecological

status in the southern parts of the catchment and “moderate” for most of the river up to

Carlisle with one of the tributaries, Blackrack Beck, currently classified as “poor”

Figure 1).

Figure 1: Map of the River Petteril catchment, with WwTWs, ecological classifications and

designated drinking water nitrate vulnerable zones

Within the Petteril catchment United Utilities has 10 small WwTWs which serve a combined population

equivalent of just over 2000. Four of these WwTWs were identified in the National Environment Programme

(NEP5) as requiring a phosphorus permit in order to meet our “fair share” reduction towards achieving good

ecological status in water quality by 2027. The cost to achieve these permits though conventional means was

disproportionate making this a good candidate for demonstrating the Integrated Catchment approach.



5

Figure 2: Applying the integrated catchment principles to develop an alternative plan for the

Petteril catchment

The IC approach is summarised in Figure 2; in essence this is an innovative approach which aims to provide

value for money for customers and targets interventions to the ‘right’ areas within a catchment. The approach

involves:

• Collaboration and partnership working with customers and stakeholders within the catchment

• Evidence gathering to build a better understanding of the inputs to the river allowing modelling

analysis and scenario planning

• Investigation and application of catchment interventions including those with additional benefits such

as that mitigation of flooding

• Asset solutions that are tailor made to small works.

Catchment sampling

Limited information was available regarding the sources of phosphorus within the catchment; particularly the

contributions from diffuse pollution within the tributaries and discharges from our WwTWs. This drove us to

develop a programme of investigations, sampling and monitoring key parameters such as: phosphorus, nitrates,

ammonia, BOD, solids and flow across the catchment.

Our newly gathered evidence allowed us to:

• Run and compare different scenarios (using the EA’s SIMCAT model) to achieve water quality

improvements.

• Identify key issues and risks at catchment level, and to target and prioritise solutions accordingly.



6

• Engage with the EA to understand the differences between their cost benefit assessment and our

required capital investment.

• Influence the EA’s technical review of the Water Industry National Environment Programme (WINEP)

and to agree the right permitting requirements for our WwWTs in the Petteril catchment.

For United Utilities this evidence-based approach set the precedent for using smart catchment strategic

principles in decision-making, engaging with key stakeholders and driving the most affordable solutions for

customers and the best outcomes for the environment.

The result for the Petteril catchment was a revision of permits to more accurately reflect the contributions

throughout the river system and a flexible approach to integrate catchment level interventions with treatment

works solutions to provide a greater overall benefit to river quality. To enable this approach a new way to

approaching P removal for small, rural WwTW was required.

Reactive media for phosphorus removal

Reactive media

In recent years a number of studies (Nilsson, et al., 2013) (Vohla, et al., 2011) (Molle, et al., 2005) have

investigated the possibility of removing P from wastewater using materials with P-retentive characteristics; these

materials are termed reactive media. There are a wide variety of reactive media products on the market, each

of which appear to operate in subtlety different ways; however the main reaction pathway involves the

generation of a phosphate precipitate utilising reaction sites on the media surface or through the dissolution of

parts of the media to form new phosphorus-containing compounds (Molle, et al., 2005) (Nilsson, et al., 2013)

(Vohla, et al., 2011).

These studies led to the theory that reactive media could be used as a tertiary treatment stage on small, rural

treatment facilities in order to meet the required P targets. The technology aligns with existing treatment

processes on these sites; it is not energy intensive, nor does it require frequent chemical deliveries or

intervention to optimise performance.

As time passes, the media sorption capacity reduces to the point that treatment is not being achieved and the

media needs to be replaced (Vohla, et al., 2011). The point at which this occurs is dependent on the reactive

media in use and the volume of media installed for a given flow rate, the influence of phosphorus concentration

is dependent on the reaction mechanism within the media. Spent media is rich in slow release, bioavailable

phosphorus which is considered of value to agriculture in countries where this technology is already in use

(Vohla, et al., 2011).

Pilot scale investigations

Design

Laboratory scale investigations were used to provide an indication of the possibilities of a range of reactive

media materials. However, these tests are not able to mimic real wastewater complexity as well as flow and

quality variations. For this reason, two commercial reactive media materials: Polonite and Phosclean (a

formulation of apatite) were selected to take forward to pilot scale on a small, rural wastewater treatment facility.

An additional media was selected to trial at pilot scale due to promising laboratory results. This media is still in

development and details are confidential, the media will be referred to as Media C for the rest of this paper.

The site chosen for the trials; Calthwaite WwTW has a population equivalent (PE) of 222, is situated in Cumbria

within the River Petteril catchment and has been given an AMP6 phosphorus permit. Calthwaite WwTW has a



7

pumped inlet with existing treatment as depicted in Figure 3. Flows passing through the pilot plant therefore

returning the flows upstream of the take-off point was considered to have a negligible impact on concentrations.

Figure 3: Calthwaite WwTW existing treatment processes and pilot trial arrangement

Previous studies have typically seen reactive media used within a constructed wetland system. This has been

considered to exacerbate potential issues when removing and replacing the media once it becomes saturated.

One potential mitigation for this is to place reactive media into a distinct unit which can be more easily replaced

(Vohla, et al., 2011). This is the principle used in the design of the pilot plant.

A sub-section of final effluent has been diverted from the final effluent chamber passes through one of three

filter arrangements each containing a different reactive media, collected and returned to the main flow at the

outlet of the tertiary SAF. The pilot trial arrangement is depicted in Figure 3. The media for each stream is held

within two up-flow filters to ensure the media is continuously saturated with wastewater and improve plug-flow

characteristics. Tertiary SAF effluent was sampled at the buffer tank and the outlet from each individual media

stream, a spot sample was taken every day for the trial duration. All samples have been analysed on site using

test kits with a sub-set of samples being sent to central labs for corroboration of on-site analysis.

Prior to commencement of the trial retention tests were carried out to determine the preferred hydraulic retention

times. For the first phase of the trial a continuous flow was maintained across the filters to replicate the hydraulic

retention time in average conditions. For the second phase of the trial the flow rate has been varied to more

accurately mimic a diurnal profile as well as replicating dry and wet weather conditions to ascertain the impact

of varying flow rates on the performance of the media.

Retention tests

A batch test was carried out to identify the preferred hydraulic retention time in average conditions. The filter

vessel was filled with media, then tertiary SAF effluent, a sample was taken every hour for 24 hours and

analysed for total P. This was repeated twice for each media (Figure 4, Figure 5 and Figure 6)

Figure 4: Phosclean retention test results Figure 5: Polonite retention test results



8

Figure 6: Media C retention test results

The retention tests showed that in all cases a target of <1mg/l total P was achieved in less than 24 hours. This

was then used as the basis for the pilot trial flow rates.

Phosclean results

Effluent has been treated by Phosclean for 27 days: in this time the media has not performed as expected

(Table 1: Phosclean performance under continuous fixed flow conditionsTable 1). Various changes

have been made to improve performance, however none have been successful to date. Investigations are on-

going to understand why the performance of this reactive media is not reflecting lab scale tests or batch test

results which have also been carried out with tertiary SAF effluent from Calthwaite WwTW. The media is known

to still be under development by the supplier and it is thought that this may be a cause of poor performance.

Table 1: Phosclean performance under continuous fixed flow conditions

Sample Total P (mg/l) Soluble reactive P (mg/l)

average range average range

Reactive media influent 4.95 1.94 – 6.83 4.72 1.97 – 6.63

Reactive media effluent 4.24 2.28 – 6.57 4.10 2.15 – 6.14

Polonite: Continuous fixed flow rate

A filter containing Polonite reactive media has been in operation for 39 days with a continuous fixed flow rate

simulating average conditions. In this time the tertiary SAF effluent total phosphorus concentration has ranged

from 2.15mg/l to 8.99mg/l, of this the soluble reactive phosphorus fraction represents almost the entirety of total

phosphorus present. Effluent following treatment showed a decrease in soluble reactive phosphorus (SRP) to

0.71mg/l (average) whilst the average total phosphorus concentration decreased to 1.93mg/l (Table 2 and

Figure 7). The disparity between total P and SRP is believed to be due to the release of calcium phosphate

precipitate which could be captured by a final solids capture process which also aligns with the vision of

sustainable wastewater treatment. The efficacy of this process is currently under investigation, it is expected

that with solids capture the treated effluent P concentration will be somewhere between the total P and SRP

concentrations presented here.

Table 2: Polonite performance with continuous fixed flow conditions




Reactive media effluent 1.93 0.93 – 4.24 0.71 <0.5 – 2.88



9

Figure 7: Polonite performance under continuous fixed flow conditions

The tertiary SAF effluent data captured as part of the trial has emphasised how variable the concentration of

phosphorus passing through the works has been. In mid-July an operational issue occurred with the tertiary

SAF which we believe contributed to the very high concentrations observed and the sudden drop between the

21st July and the 23rd July. However, the variability is frequently seen and could be a feature of this particular

influent flow. More will be understood on this as the trial continues.

Polonite: Variable flow rate

The feed flow rate has now been modified to replicate a more representative flow pattern including wet and dry

conditions. This phase of the trial is still in progress and will continue for a number of weeks. Initial results (Table

3 and Figure 8) have indicated that this has resulted in an improvement in performance, potentially due to the

introduction of ‘rest periods’ which the supplier suggests will improve the reactivity of the media.

Table 3: Polonite performance under variable flow rate conditions







10

Figure 8: Polonite performance under variable flow rate conditions

A function of the reactions undertaken by Polonite is a significant increase in pH to alkaline conditions of up to

pH11.5. This would require correction prior to release of effluent to the water course; investigations are ongoing

into sustainable options for pH correction.

A wide range of parameters have been tested as part of the analysis. Initial indication suggest that

concentrations following treatment by Polonite are less than or equal to the concentrations currently present in

tertiary SAF effluent.

Media C: Continuous fixed flow rate

A filter containing Media C has been in operation for 36 days with a continuous fixed flow rate simulating average

conditions. In this time the tertiary SAF effluent total phosphorus concentration has ranged from 2.15mg/l to

8.99mg/l, of this the soluble reactive phosphorus fraction represents almost the entirety of total phosphorus

present. Effluent following treatment showed a decrease in soluble reactive phosphorus (SRP) to 0.48mg/l

(average) whilst the average total phosphorus concentration decreased to 0.53mg/l (Table 4 and Figure 9).

Table 4: Media C performance with continuous fixed flow conditions




Reactive media effluent 0.53 <0.50 – 0.97 0.50 <0.50 – 0.94



11

Figure 9: Media C performance under continuous fixed flow conditions

Media C: Variable flow rate

The feed flow rate has now been modified in the same way as the Polonite filter to replicate a more

representative flow pattern including wet and dry conditions. This phase of the trial is still in progress and will

continue for a number of weeks. Initial results (Table 5 and Figure 10) have indicated that this has not resulted

in a detrimental change in performance.

Table 5: Media C performance under variable flow rate conditions







12

Figure 10: Media C Performance under variable flow rate conditions

Initially using Media C with variable flows resulted in greater variability in the results, this has not continued as

the trial has progressed. The cumulative effect of variable flow rates is yet to be determined; more information

on this will be collected as the trial progresses.

Effluent treated by Media C has also been analysed for a wide range of parameters Initial indications suggest

that concentrations of metals following treatment are less than or equal to the concentrations currently present

in tertiary SAF effluent.

Discussion of pilot trial results to date

The trial results so far indicate that reactive media could provide effective phosphorus treatment at small, rural

treatment works. The laboratory and batch results using Phosclean do not appear to scale up to a continuous

flow system. This may be due to the developmental status of the media being investigated or may be due to

other reaction characteristics which have not been fully identified. Results using Polonite have been positive;

however they also demonstrate the need for tertiary solids capture downstream and the increasing pH

associated with this media. Media C has consistently produced an effluent with an average total phosphorus

concentration less than 1mg/l. This product is in development and a number of steps need to be taken to ensure

that it is fit for purpose prior to rolling out as a phosphorus treatment technology.

The trial is continuing over the next few months, in this time we aim to provide more confidence over treatment

capability under variable flow rate conditions and complete the analysis into the likelihood of release of other

compounds as a result of using reactive media.

Conclusion



13

Development of an integrated catchment approach for the River Petteril and investigation into sustainable P

removal technologies has led to the following conclusions:

1. The objective to reduce nutrient levels in rivers in order to meet the WFD has resulted in increasing

numbers of P permits within rural catchments

2. A systems thinking approach has been applied to provide a greater benefit to the customer; applying

a more relaxed permit at a treatment facility can enable catchment interventions leading to a greater

overall benefit to river quality

3. An evidence-based approach has been used alongside collaboration with key stakeholders within the

catchment to ensure that interventions are targeted where they are needed and all parties remain

engaged in the overarching purpose

4. In order to provide P treatment at small WwTWs, typically technologies developed for large works

have been scaled down. Consequently, the cost to provide treatment is disproportionate to the

number of customers and the environmental benefit in many cases. There is a gap in the market for

sustainable P treatment tailored to the needs of small WwTWs

5. A retention time of 24 hours was considered sufficient in batch tests. This is a reasonable volume to

be suitable for small WwTWs

6. The reactive media Phosclean has so far not been found to be effective at removing phosphorus in

this situation. Efforts are ongoing to understand why this is

7. The reactive media Polonite has shown consistent removal of phosphorus although there is a

disparity between the total phosphorus in the treated effluent and the soluble reactive phosphorus

implying that the media is reacting and releasing a phosphorus precipitate. If this can be captured in a

final treatment stage then the potential of this product increases

8. Additional treatment stages may be required to provide solids capture and pH correction

9. Media C is providing very promising results with respect to the provision of sustainable phosphorus

removal and has produced an effluent with an average total phosphorus concentration less than 1mg/l

10. There are more questions to be answered. Trial results to date have shown that there is potential in

this approach to provide sustainable phosphorus removal which is tailored to small WwTWs and rural

river catchments.

Acknowledgements

The work reported in this paper has been supported by the United Utilities Process Technology Team, Cumbria

Wastewater Teams and Lancaster University. The reactive media has kindly been provided by Ecofiltration and

ARM.

References

Molle, P. et al., 2005. Apatite as an interesting seed to remove phosphorus from wastewater in constructed

wetlands. Water Science Technology, 51(9), pp. 193-203.

Nilsson, C. et al., 2013. Effect of organic load on phosphorus and bacteria removal from wastewater using

alkaline filter materials. Water Research, Volume 47, pp. 6289-6297.

Vohla, C. et al., 2011. Filter materials for phosphorus removal from wastewater in treatment wetlands - A

review. Ecological Engineering, Volume 37, pp. 70-89.



14

TERTIARY DAF COMBINED WITH I-DOSE FOR MAXIMUM REMOVAL OF

PHOSPHORUS FROM MUNICIPAL WASTEWATER EFFLUENT

Menkveld, H.W.H.1, Broeders, E.1, Mansell, L.2, Fox, E.2, Tolman, M.3

1Nijhuis Industries, Netherlands, 2United Utilities, UK, 3Nijhuis H2OK Ltd, UK

Corresponding Author Tel +31(0)314-749000 Email [email protected]

Abstract

United Utilities undertook a phosphorus removal trial as part of UKWIR’s Chemical Investigations Programme

Phase 2 (CIP2) Innovation Fund trials using Tertiary-Dissolved Air Flotation (T-DAF) technology from Nijhuis

Industries. The T-DAF with intelligent real-time control ‘i-DOSE phosphorus’ was constructed for modular

installation and trialled for 6 months under the supervision of United Utilities at Macclesfield WwTW located near

Manchester in the United Kingdom. Phosphorus removal is achieved by chemical precipitation and flocculation

of phosphates prior to dissolved air flotation. The system utilised Nijhuis Industries’ intelligent chemical dosing

system (i-DOSE phosphorus) which controls the system in real-time based upon the incoming phosphorus load

to ensure efficient phosphorus removal after the clarifiers is achieved for the lowest possible OPEX cost.

The T-DAF in combination with the ‘i-DOSE phosphorus’ is an efficient and robust technology for effluent

polishing to remove phosphorus and has the potential to reach an average effluent concentration for total

phosphorus (TP) of 0.4 mg/l without exceeding the discharge limits for iron. Also BOD, COD and TSS are greatly

removed in the process. The produced sludge is expected to have an average DS% of 4%, which would save

sludge processing costs downstream.

The presentation will provide an overview of the process, key design features, treated effluent quality results

and operational observations.

Keywords

Tertiary Dissolved Air Flotation, Phosphorus Removal, i-DOSE, UKWIR, CIP2, United Utilities

Introduction

The Chemical Investigations Programme Phase 2 (CIP2) is co-ordinated by UKWIR in response to the

challenges of the environmental quality standards (EQS) set out in the Water Framework Directive. In order to

address the specific challenge of meeting the EQS for phosphorus, the UK Water Industry (under the co-

ordination of UKWIR and in collaboration with the Environment Agency) initiated an evaluation of the likelihood

of the effluents from innovative technologies to contribute to a reduction in phosphorus concentrations in

wastewaters.

Phosphorus is currently the limiting factor preventing good chemical status under the Water Framework

Directive for UK inland surface waters. Typically, aerobic wastewater treatment does not substantially reduce

phosphorus concentrations and WwTW’s can become a point source discharge unless additional treatment is

applied. Diffuse sources are often also present, although these are often less controlled (for example through

fertiliser runoff).

The removal of phosphorus from point source discharges into the water catchment can create real ecological

benefits within waterways. This has driven new discharge consents and as a result technologies and methods

have been developed to reduce effluent TP concentrations to sub 0.5 mg/l levels.

[email protected]



15

In 2009 preliminary pilot testing with a T-DAF was conducted for United Utilities at Grasmere WwTW. The

results showed an average concentration for total phosphorus of 0.25 mg/l and an average sludge

concentration of 4.1% DS can be reached using a T-DAF [Kluit, A, ter Horst, C, Crewdson, C, Crewdson, J.,

pilot unit test report WwTW Grasmere]. Since 2009 many innovations have taken place to improve the

performance of the T-DAF as a separation technology and the first proof of principle for the Nijhuis Industries’

intelligent chemical dosing system ‘i-DOSE phosphorus’ was conducted on raw influent of WwTW Eindhoven

the Netherlands in 2014 [STOWA 2014]. Combined with intelligent real-time control ‘i-DOSE phosphorus’ a new

T-DAF pilot plant was constructed for modular installation and trialled for 6 months under the supervision of

United Utilities at the Macclesfield WwTW located near Manchester in the United Kingdom.

Material and methods

Phosphorus removal is achieved by precipitating phosphates with a chemical coagulant and subsequently the

precipitated particles are flocculated together with other small particles using a cationic polymer. After

coagulation and flocculation suitable flocs are formed that can be removed efficiently using dissolved air

flotation. During the pilot trials at Grasmere WwTW the coagulant dosage was calculated based on grab

samples on site that were analysed for phosphate (PO4-P). Naturally the phosphate concentration varies over

time due to changes in the influent characteristics and due to dilution during rainy weather. To prevent over (or

under) dosage of the coagulant and to guarantee a good performance for phosphorus removal, an online dosing

control was deemed a necessity.

The system at Macclesfield WwTW utilised i-DOSE phosphorus which controls the system in real-time upon

both feed-forward (based on the incoming phosphate load) and feed-back (based on phosphate load of the

treated wastewater) and by using a unique algorithm the chemical dosage was optimised at all times. The

phosphate in the influent and effluent of the T-DAF was measured by using a Hach Phosphax analyser and a

Filtrax sample filtration system. The site analyses for TP and ortho-phosphate (PO4-P) were measured by using

Hach cuvette tests (LCK349/348/350).

The feed-forward phosphate control ensures that the coagulant dosage is appropriate to remove most of the

incoming phosphate load. This is required to deal with fluctuations of phosphate in the wastewater. The feed-

back control is a very slow control mechanism that adjusts the coagulant dosage to reach the desired effluent

concentrations of phosphate. This double control strategy ensures an efficient final effluent phosphorus removal

is achieved for the lowest possible OPEX cost.

Table 1: Design features pilot plant TDAF

Design feature Value Unit

Flow 5 m3/h

HRT coagulation 3-4 min

Type of coagulant Fe2(SO4)3

Concentration 12.5% Fe

Maximum Fe:PO4-P ratio 4.5

HRT flocculator 1 min

Type of polymer

Concentration at dosing point 0.1 %

Aeration pressure 6 bar

PO4-P influent 4 mg/l

Goal TP effluent < 0.5 mg/l

Zetag 9014 (cationic)



16

Figure 1 shows a schematic of the T-DAF pilot and Table 1 displays the key design features of the T-DAF pilot.

The T-DAF comprises a reaction tank for coagulation with a hydraulic retention time of 3-4 minutes followed by

a plug flow reactor where (in this specific case) a medium charged high molecular weight cationic polymer was

added to flocculate the coagulated particles. The coagulant to phosphate molar ratio was limited to a maximum

of 4.5 to prevent the total iron content in the effluent of the T-DAF increasing above the requested trial limit of

4.0 mg/l (95 percentile grab sample). The flocs are removed in the T-DAF by using dissolved air floatation

resulting in a relatively low solids effluent (15 mg/l 95 percentile grab sample). The effluent is partially

recirculated and aerated under pressure (approximately 6 bar) to create the microbubbles which are the driving

force of dissolved air floatation. This so called white water is injected in both the flocculator and the T-DAF

system in order to achieve the highest possible solid separation. Sludge is thickened and skimmed of the top

and can be further processed by dewatering or digestion.

Figure 1: Schematic overview of the T-DAF

Results and Discussion

During July and August the trial encountered issues with the activity of both the polymer and the ferric sulphate

(a few weeks after 50-50 dilution of the ferric sulphate the activity degraded to less than 20% of the original

activity). For these reasons a representative period of 5 weeks of the T-DAF trial was chosen for the graphic

display. Table 2 represents the validated data over the trial period excluding the periods where the performance

is not considered representative due to operational issues mostly due to a reduced activity of the chemicals.

Phosphorus removal

Figure 2 shows a graph of the online PO4-P measurement of both the influent and effluent of the T-DAF,

combined with lab measurements for PO4-P and TP over a period of 5 weeks. The graph over this 5 week period

shows that the T-DAF can remove TP to an average of 0.32 mg/l. The online influent PO4-P measurement

deviates approximately 0.5 mg/l compared to the site analyses and is usually lower than the actual value. The

online PO4-P measurement in the effluent is slightly higher than the site analyses. In this case the set point for

PO4-P of 0.15 mg/l is close to the detection limit (0.05 mg/l) of the analyser.



17

Figure 2: Online PO4-P and site lab measurements of PO4-P and TP of influent and effluent T-DAF

Iron content in effluent

Figure 3 shows that the iron content of the T-DAF effluent increases due to the coagulation with ferric sulphate

in most cases. The average influent iron concentration in Figure 3 was 1.47 mg/L and the average effluent of

the T-DAF contained an iron concentration of 2.16 mg/L. The 95 percentile value for the total iron concentration

in the effluent of the T-DAF is 3.6 mg/L. This is compliant with the request of United Utilities of a 4 mg/L 95

percentile limit on iron in the effluent of the T-DAF. In multiple measuring days the iron content does not

increase. Further optimisation can be realised to reduce the iron content in the effluent to prevent peaks of iron

above 4.0 mg/L in the effluent. During dry weather a lower metal (ferric) – phosphate (Me:PO4-P) ratio [Szabó

et al. 2006] can be used compared to rain weather conditions when the phosphorus concentration is much lower

due to dilution with rain water.

Figure 3: Total iron and dissolved iron concentration of influent and effluent T-DAF

DS content of sludge after treatment

One of the major benefits of applying a T-DAF for phosphorus removal is that sludge can be concentrated to an

average of 4% DS (results P trial Grasmere WwTW). This saves costs for sludge processing (less volume) and

transportation. The reasons for the lower sludge dry solids content than expected were two-fold. Firstly during

the pilot trials long piping (approximately 150 m) has been installed for sludge discharge from the sludge hopper

combined with the sludge and sand drains on the bottom of the T-DAF installation. These long pipelines caused

the water level in the T-DAF to vary because the sludge and sand drain could not discharge sufficient solids

due to clogging and resistance in the pipe lines. Because of the changing water level in the T-DAF the scraper

cannot be set to remove the sludge at a high DS%. Secondly the in order to fit the unit within the container the

design of the slope of the sludge hopper was adjusted which prevented sludge freely flowing. To prevent sludge

build-up in the hopper the water level in the T-DAF has been increased in order to scrape more water off the



18

top resulting in a lower DS%. Both of these effects resulted in varying sludge concentrations of less than 1%

and up to 3% DS.

Optimisation of the plant has been carried out in order to improve the dry solid content at the end of the trial.

Figure 4 shows the sludge on top of the T-DAF and after sampling following this optimisation. The sludge DS%

will be closely monitored during the final stage of the pilot trials to prove the results for DS% (average 4%) of

the trials at Grasmere WwTW can be replicated.

Figure 4: Sludge layer on top of the T-DAF and after sampling

Removal efficiencies of all components

Table 2 represents the validated data (both site analyses for TP and PO4-P as regulatory analyses for BOD, COD

and TSS throughout the trial period.

Alongside TP other components such as BOD, COD and TSS have been monitored throughout the trials and

are displayed in Table 2 below. Next to a high removal rate for TP of 93%, BOD, COD and TSS are greatly

reduced by (78%, 67% and 74%). The ammonium concentration before and after the TDAF have been

monitored, but as to be expected since ammonium will not be coagulated, there is no removal of ammonium.

Table 2: Removal and removal efficiency for TP, PO4-P, BOD, COD and TSS

Operational features

The T-DAF in combination with the ‘i-DOSE phosphorus’ is an efficient and robust technology for effluent

polishing to remove phosphorus. During the 6 month period some mechanical problems have occurred of which

none would cause issues for full-scale application. Most issues that occurred were related to a low dosing flow

of the coagulant, causing air locks in the piping and sludge not flowing down properly in the sludge hopper

caused by the limitation of the slope of the sludge hopper due to placement in a container.

Conclusions

Utilising the T-DAF technology with the i-DOSE phosphorus led to an average effluent concentration of

approximately 0.36 mg/L TP in the validated data set presented. After addition of ferric (sulphate or chloride)

TP [mg/L] 5.09 [2.2 - 8.1] 0.36 [0.09 - 0.99] 93%

PO4-P [mg/L] 4.2 [1.9 - 6.3] 0.10 [0.03 - 0.34] 98%

BOD [mg/L] 14.3 [7.5 - 33.2] 3.1 [0.9 - 8.2] 78%

COD [mg/L] 75 [32-106] 25 [2 - 40] 67%

TSS [mg/L] 29 [20 - 55] 7.5 [3 - 16] 74%

Parameter UnitInfluent T-DAF

average [min-max]

Effluent T-DAF

average [min-max]

Removal

Efficiency



19

the iron concentration naturally increases in the effluent. However by applying an intelligent dosing control the

iron concentration was within the 4.0 mg/L 95% percentile requirement.

Other components such as BOD, COD and TSS (78%, 67% and 74%) are significantly removed in the same

process. Due to the low TSS content of WwTW effluent, the T-DAF is designed for hydraulic load and not for

solid load. Application of a T-DAF therefore has the advantage that an extra safeguard is in place in case of

sludge washout after the secondary clarifiers.

Another benefit when applying a T-DAF system is the DS% of the sludge. Because of pilot scale issues the

sludge DS% varied <1% up to 7.5% has been monitored), the sludge in a full-scale application is estimated to

be an average of 4% DS after treatment as shown in the pilot trails at WwTW Grasmere in 2009. Reducing the

sludge volume has the potential to save OPEX for transportation and further processing of the sludge,

depending on the local conditions of the WwTW

References

Szabó, A., Takács, I. (2006), The importance of slow kinetic reactions in simultaneous chemical P removal.

Proceedings of the Water Environment Federation, WEFTEC 2006: Session 61 through Session 70, pp. 4864-

4872(9)

STOWA (2014) Toepassing van Dissolved Air Flotation als voorbehandeling van communaal afvalwater,

STOWA, rapport 2014-03 / ISBN 978.90.5773.648.3

http://www.ingentaconnect.com/content/wef/wefproc;jsessionid=1xnhwgihgogc6.x-ic-live-03



20

GPS-X MODELLING TO OPTIMISE NITRIFICATION AND RISK ASSESS DESIGN

PROPOSALS - CASE STUDY OF BROCKHAMPTON SEWAGE TREATMENT WORKS

Ruswa, E.1 and Copp, J.2 Chadha, M3., 1Severn Trent Water Process Design Engineering (UK): [email protected]

2Primodal (Canada), [email protected], www.primodal.com 3Severn Trent Water Asset Creation (UK), [email protected]

Abstract

Brockhampton STW is currently being re-permitted to a tighter ammonia limit of 1.2mgN/l from 5mgN/l.

Compounding the issue was the fact that the works was experiencing unexplained daily ammonia spikes that

were risking compliance and increasing OPEX requirements. To diagnose the issue and investigate future

options, Severn-Trent Process Design Engineering commissioned Primodal to develop a GPS-X process

model. Model development showed the measured effluent ammonia data was well represented if a significant

delay was included in the aeration system sub-model. This assumption was confirmed when the Brockhampton

(15,000PE) aeration control code was found to contain a hard-coded 30-minute delay and found to be an exact

copy of the Minworth STW (1,500,000PE) control code. The modelling work predicted that eliminating the delay

would eliminate the ammonia spikes. The delay was removed on-site and the spikes were eliminated. The

model has subsequently been used to risk assess future scenarios and this has negated the requirement for

extensive civil works in the current capital scheme.

Keywords

GPS-X, Modelling, ASP, Anoxic, Aeration, Nitrification, Risk assess

Introduction

In AMP6, a number of sewage discharge permits are likely to be tightened as a result of the Water Framework

Directive. River modelling undertaken by Severn Trent Water has identified that Brockhampton STW is subject

to an ammonia permit of 1.2 mg/l (95 percentile).

Table 1: Design Basis

Current (2013) Future (2028)

Population Equivalent 12150 16025

Permit [BOD/SS/NH3/P/Fe/Al] (mg/l) 15/25/5 10/25/1.2

Permitted DWF (m3/d & l/s) 3360 & 38.9 4220 & 48.8

Permitted FFT (m3/d & l/s) 8208 & 95 9072 & 105

Measured DWF (Q80) (m3/d & l/s) 3333 & 38.6 4018 & 46.5

Measured average flow (m3/d & l/s) 4752 & 55.0 5435 & 62.9

The present works consists of an inlet pumping station (Archimedes screw, no overflow), primary treatment (4

Dortmund tanks) followed by a 6 pocket ASP, and four Finals Settlement Tanks. There was a history of

unexplained daily NH3 spikes in the final effluent at Brockhampton which have significantly increased the

operator visits to the site over the past few years.



http://www.primodal.com/




21

Domestic Growth

New housing developments in the catchment are expected to increase the PE by 3,875 (31%) by the 2028

design horizon (2).

Cheltenham Race-course Loads

The Cheltenham Racecourse has a dedicated sewer which discharges directly into two storage tanks at the

works (4232m3).

There are several race meetings over the course of the year but the most significant is the Gold Cup meeting

in March. This 4 day festival typically sends 3021m3 of waste to the storage tanks. This equates to approximately

1963 kgBOD and 560 kgNH3-N. When returned to the ASP over 30 days, the race course returns are expected

to be approximately;

• 65.4 kgBOD/d

• 18.7 kgNH3-N/d

• 100.7 m3/d

Leachate Loads

Additionally, a nearby tip is permitted to discharge leachate to the works through a dedicated sewer. Measured

tip leachate loads appear to fluctuate between 20 to 50 kgNH3-N/d.

Design Basis

The design envelope gives future settled loads of 730.4 kgBOD/d and 213.4 kgNH3-N/d which is inclusive of

domestic, racecourse and trade waste.

In summary the works has been designed on two premises;

• No race (domestic + tip leachate + trade): 665kgBOD/d and 194.8kgNH3-N/d

• With Cheltenham Race returns: 730.4kgBOD/d and 213.4kgNH3-N/d



22

Figure 11: Birds-eye-view of Brockhampton STW

ASP

The existing ASP which consists of 2 unaerated and 5 aerobic pockets1 is currently operating at an F:M of

approximately 0.06 including the anoxic zone (NH3 F:M of 0.016) for periods excluding the race meetings. As

the crude sewage comprises 3 main constituents, domestic, tip leachate and race course returns, the future

1 length/width: 9.4m and depth 5m. Total anoxic volume: 442m3. Aerobic Volume: 2209m3



23

works capacity has been process modelled using GPS-X, outputs of which have been used to derive the process

options for the site.

Modelling Brockhampton

Model Objective

The model developed as part of this work was based on an earlier model developed by Primodal in 2008 to

assess the risk associated with a planned upgrade at that time. Following that modelling work in 2008, two

additional ASP pockets were constructed and the aeration system was upgraded at the site. The plant design

was revisited in 2014, in part, because contrary to the 2008 modelling results even with the additional ASP

capacity, the plant was unable to adequately treat the incoming ammonia load on a daily basis. As a result, a

new modelling project was initiated to investigate the current state of the plant and new upgrades options for

the plant.

The main objectives of the modelling work were:

• to calibrate a model of the present works,

• to investigate possible causes for observed effluent ammonia spikes,

• to investigate process control/optimisation options to address the effluent ammonia issue.

Plant Design

The Brockhampton wastewater treatment facility receives wastewater that can be primarily characterised into 4

categories: i) domestic wastewater (including a normal daily industrial load); ii) leachate; iii) storm water; and,

iv) intermittent high-strength wastewater from the Cheltenham racecourse that occurs only during race

gatherings, a couple of times per year. The plant itself has seven ASP pockets with the first two being unaerated.

No internal recycles are currently operating. Off-line, the plant has two storage facilities: i) a storm tank (~420m3)

which is meant to receive flow when the incoming flow exceeds 95 L/s (FFT) and a 4232 m3 storage facility that

is used for racecourse waste during race meetings. Stored storm waste is fed back into the system at 25 L/s,

when the incoming flow drops below 70L/s. The racecourse waste, which is a much higher strength, is fed back

into the system at approximately 1.2 L/s with an aim to empty the storage tanks over a 30-day period.

Plant Design

The Brockhampton wastewater treatment facility receives wastewater that can be primarily characterised into 4

categories: i) domestic wastewater (including a normal daily industrial load); ii) leachate; iii) storm water; and,

iv) intermittent high-strength wastewater from the Cheltenham racecourse that occurs only during race

gatherings, a couple of times per year. The plant itself has seven ASP pockets with the first two being unaerated.

No internal recycles are currently operating. Off-line, the plant has two storage facilities: i) a storm tank (~420m3)

which is meant to receive flow when the incoming flow exceeds 95 L/s (FFT) and a 4232 m3 storage facility that

is used for racecourse waste during race meetings. Stored storm waste is fed back into the system at 25 L/s,

when the incoming flow drops below 70L/s. The racecourse waste, which is a much higher strength, is fed back

into the system at approximately 1.2 L/s with an aim to empty the storage tanks over a 30-day period.

Data Analysis

Data analysis and reconciliation is a crucial part of any model development project. This involves accumulating

all available data including high frequency, composite and grab sample data over several years and analysing

that data for trends and systematic errors through various techniques like mass balancing around clarifiers. All

too often, this data analysis exercise reveals a problem with the data or a process issue that was previously

unknown (i.e. flow splits not as expected, problems with sampling locations, sampling units, sensor or lab

procedures that are leading to systemic errors in the recorded data…).




Current Process Schematic

Figure 12: Process schematic





Figure 3 highlights a typical problem discovered through a simple mass balancing exercise performed

during another project. The issue here was that the design engineer was basing their design and

calculations on the recorded data which looked intuitively reasonable (all flows and ratios within the design

specification) but was clearly incorrect when examined more rigorously as part of the subsequent model

development.

Figure 13: Clarifier mass balancing for data reconciliation.

Flow splits are another example of typical problems as actual flow splits are rarely equal and unequal flow

splits can lead to unexpected process behaviour that is not readily identifiable from the data alone. The

purpose of this exercise is to ensure that the model development is based on a dataset that is accurate and

appropriate, and that the data does not contradict the owner’s understanding and operation of the process.

Once this is done, it is possible to more closely assess the reasons for differences between measured and

modelled results. If the measured data has been reconciled and the model cannot reproduce the observed

behaviour, then it is necessary for the modeller to determine what is wrong with the model and not simply

start changing model parameters to curve fit the model to the observed data. A thorough investigation of

these differences in this case led to the discovery of a problematic – and not previously realised – issue

with the control and operation of the plant that was causing the plant to spike ammonia on a daily basis.

Influent Characterisation

In this case, all of the 2008 data was available, but as part of the more recent model update, the flows to

Brockhampton were revisited to identify a reasonable starting point for the simulations. An analysis of the

analogue flow data was performed and showed that typical flows to Brockhampton had not changed since

2008 (Table 2). Similarly, an analysis of the daily flow patterns revealed that the measured 2008 and 2014

patterns were similar suggesting little change in the catchment behaviour (Figure 14).




26

Table 2: Brockhampton flow analysis.

Units 2008 Measured Flows 2014 Measured Flows (2009-2014)

Estimated dry weather flow L/s 41.2 41.2

Average daily flow L/s 52.5 (2007) 52.6

Figure 14: Comparison of the measured daily flow pattern in 2008 (2008) vs the average profile

measured since then (2014).

This result contrasted the more recent influent ammonia profile data which showed a considerable deviation

from the 2008 profile. The impact of this profile change is significant as the ammonia concentration peak

corresponded with the highest daily flow peak. These two things when combined result in a significant peak

in ammonia load during that time of day. It is interesting to note that the ammonia concentration spike also

corresponded with the period of time that leachate is assumed to be arriving at the plant.



27

Figure 15: Daily ammonia profiles in the influent to Brockhampton measured in 2008 and 2014.

The behaviour of the model is highly dependent on the fractionation of the incoming COD into its component

parts. This relatively simple concept is complicated by the fact that modelling fractions typically cannot be

measured directly. Rather, influent measurements taken on-site are confounded by practical issues like the

definition of soluble material. For example, the amount of soluble material measured in the lab is a function

of the filter paper pore size and yet the model assumes that all degradable soluble material is readily

biodegradable. Therefore, depending on the lab procedures, the same sample could potentially give very

different soluble measurements. It is therefore important that the modeller understand how to interpret the

measured and modelled results to best estimate the various COD fractions correctly. A pictorial

representation of the influent COD fractionation for Brockhampton is shown in Figure 16 and it is interesting

to note that there was no obvious evidence in the model behaviour or measured data to suggest a change

to the COD fractions from those determined in 2008.



28

Figure 16: Modelled domestic influent COD fractions for Brockhampton.

In addition, to more realistically model the variations in the influent and better estimate the impact of storm

flows, a simplified sewer model was developed for the storm water input. This sewer model included

deposition and resuspension of material in the sewer based on the flow measured at the plant. This resulted

in first flush events (i.e. added load) at the beginning of storm events if the modelled sewer had deposited

material and dilution of the domestic load if the modelled sewer was empty of deposited material (i.e. after

a sustained wet weather period). This very simple approach, was capable of generating first flush and

dilution events and in the case of Brockhampton, the sewer model was tuned to give a variable influent that

was consistent with the influent concentration data measured at the plant so was deemed a reasonable

approximation of the real situation.

Model Layout

The model (Figure 7) layout was created based on information supplied by STW including the geometry of

all the unit processes involved and their operation such that recycled activated sludge (RAS) and influent

flows were input as measured on-site or inferred from operational documentation. Surplus activated sludge

was estimated to be 100 m3/d based on the SAS pump data (4.9 L/s) and the logs which suggested an

average running time of 6-7 hours per day. Settling behaviour was approximated using a SVI correlation.



29

Figure 17: GPS-X model of Brockhampton facility.

Simulation Results

When calibrated appropriately, municipal wastewater treatment plant models can simulate the behaviour of

a works with a high degree of accuracy. So it was somewhat surprising to learn that the plant was struggling

to treat the incoming ammonia given that the 2008 model had been used to confirm the additional ASP

capacity prior to construction and that the influent characterisation had not changed significantly over the

intervening years.

Initial runs with the updated 2008 model (two new aeration pockets, new aeration taper, new ammonia

profile) demonstrated that that model gave a good dynamic fit to the available data, except it did not predict

the daily ammonia spikes that the plant was experiencing (Figure 18) which suggested one of three things:

a) that the influent loads being used were in error; b) that an important part of the process was unaccounted

for; or c) the model was not properly simulating an operational issue. Even though the easiest solution

would have been to attribute the differences to the uncertainty in the incoming ammonia load and simply

increase that load to produce a fit, a more detailed data request suggested another cause.

The first two options were quickly eliminated as the likely cause when the model behaviour and the data

from the blowers was examined. Interestingly, and not unexpectedly, the effluent ammonia was affected by

the DO levels in the various pockets. It was also noticed that the dissolved oxygen control was not ideal as

the measured dissolved oxygen (DO) levels varied considerably throughout the day. The initial model runs

predicted a smooth DO profile and smooth airflow dynamic, but the actual blower data did not agree. Rather



30

the recorded blower behaviour was erratic and the measured DO profiles indicated that the aeration

equipment was unable to track the setpoint.

Figure 18: Modelled (line) and measured (dots) effluent ammonia over a 10-day period in the

winter when simulated with ideal DO control.

These results indicated that the model was not accurately reproducing the aeration behaviour and that

some combination of the control logic, blower response, pipe work and/or valve control was causing the

real system to oscillate around the setpoint. Several different possible explanations were investigated, but

the model quite accurately predicted the erratic blower behaviour and the oscillating DO values when the

DO sampling interval was decreased from every 10 minutes to every 25-30 minutes (effectively indicating

that there is a 30-minute delay between the probe calling for a change in air, and the aeration system as a

whole actually delivering that air). Modelling the aeration system with this delayed response provided results

that quite closely mimicked the behaviour in the plant (Figure 19 & Figure 20) in both cells that had DO

probes.



31

Figure 19: Modelled (line) and measured (dots) effluent ammonia over a 10-day period in the

winter.

Figure 20: Measured (dots) and modelled (line) dissolved oxygen in tank 7.

The process model strongly suggested that the spikes in effluent ammonia occurring at the plant were in

part being caused by the sluggish response of the aeration control system. The model demonstrated that

the slow response of the control system was preventing the ramping up of oxygen delivery in a timely

manner meaning that the DO was dropping to unsuitably low levels when high ammonia loads were entering

the system and this was allowing ammonia to pass through the plant untreated. Similarly, when the loads



32

dropped, the model was showing that the aeration system turn down was too slow to prevent the DO from

reaching unnecessarily high levels (Figure 20). Figure 21 shows the modelling results with and without the

aeration delay.

Figure 21: Measured (dots) and predicted (line) DO for the 26-day period showing the change

in DO variability after changing the aeration logic on day 7.

Using a model estimated delay of 25-30 minutes resulted in swings in DO consistent with the measured

data and predicted effluent ammonia spikes also consistent with the measured data. Without that delay,

the model predicted lower DO fluctuations and more consistent blower operation. The modelling work led

to the conclusion that the aeration control system delays were causing severe oscillations in the dissolved

oxygen concentrations which were leading to the effluent ammonia issues.

A review of the control logic was undertaken and a previously unknown 30-minute delay was found hard-

coded into the aeration control logic. Based on these modelling results, this hard-coded delay, which had

been included in the logic to dampen the aeration response, was removed from the control logic and the

aeration system immediately responded. The daily effluent ammonia spikes were eliminated in April 2015

as shown in Figure 22 below.



33

Figure 22: ESCADA trend of the final effluent NH3 before and after delays in the aeration system

were rectified

Modelling Conclusion

The modelling work done as part of this project highlighted several things including that: a) well-calibrated

models need very little maintenance over time and can be used years later with minimal additional effort;

b) daily spikes in effluent ammonia does not necessarily mean that the tank and/or aeration capacity is

limiting the ammonia treatment; c) poorly implemented control logic can have a profound impact on the

process and that rectifying control issues can negate the need for costly capital expenditures in plants that

have already sufficient capacity; and, d) planned and/or implemented control logic should be modelled by

an experienced professional so that problems and inefficiencies can be minimised so that more robust and

cheaper treatment can be realised.

The model clearly suggested that the aeration system control was contributing to the effluent ammonia

issue. Data from the plant showed large swings in the measured DO and an inability to track the DO

setpoint. The model correctly identified that poorly implemented control logic was the cause of the problem

and the model was used to correct that control logic. Subsequent changes to the logic on site resulted in

the elimination of the effluent ammonia spikes.

Process Solution

The aforementioned calibrated process model was then used to investigate design options for the works.

To do this, the simulated load to the plant was increased in line with the design, the planned aeration

setpoints were input, and the tanks were made deeper by increasing the depth of the plant using the

available hydraulic head.

In total 23, one year long scenarios were modelled assessing

• final effluent ammonia, nitrate, nitrite

• Winter performance



34

• Wet weather performance

• An appropriate anoxic volume

GPS-X modelling results were run at a 1.2hr communication interval and this revealed that the overall

summer dry weather 95% percentile performance of all the 23 modelled scenarios were well below the

1.2mg/l future permit. The model clearly indicated that the works would robustly meet the future permit in

the summer at dry weather flows; however winter temperatures (8-10ºC as shown in Figure 23) and storm

flows significantly affected the overall performance and hence dictated the overall ASP volume.

Figure 23: Modelled Temperature profile

Table 3 summarises some of the modelled scenarios starting from the existing plant. To assess the risks

of all the options, the number actions limit breaches were accounted for. In the case of Brockhampton, an

action limit breach is any modelled sample that exceeded 0.75mg/l NH3-N.

Optioneering

Simulating the existing plant with the future loads, for the majority of the time the plant would meet the

future permit with an overall 95%ile NH3-N of 0.59mg/l. However winter temperatures and storm weather

performance caused 253 action limit breaches. As each data point represents a 1.2hr communication step

in the model, this would equate to a combined 13 days in the year where NH3-N compliance may be

compromised.

Reducing the nominal anoxic retention to under 1 hr (scenario 4) with the existing plant, reduced the action

limit breaches down to 44, in the modelled year.



35

The final effluent ammonia from Scenario 4 (F:M: 0.091) and Scenario 5 (F:M: 0.081) are shown in Figure

24 and Figure 25 respectively. These figures demonstrate the relative risk of each option as each plot gives

a pictorial view of the plant performance over the course of a year.

Using this approach, Scenario 5 (321m3 anoxic) which has five modelled action limit breaches, was selected

as the design point for the site even though Scenarios 6 and 7, which do not have any action limit breaches,

appear to be more robust2.The more robust Scenario 6 did not have a sufficient anoxic volume and scenario

7 would require a further extension of the ASP basin and was dismissed on cost.

Therefore scenario 5 was taken forward on an acceptable risk basis and an anoxic volume optimisation

exercise was conducted (scenarios 5) to enhance denitrification levels.

2.




Table 3: Summary of the Modelled Scenarios

Scenario Additional

Pockets

Total

Settled

BOD

Load

(kgBOD/d)

F:M NH3,

F:M

ASP

Orientation

Anoxic:Aerobic

Annotation

95%ile Final

Effluent

NH3 (mg/l)

Number of

Action Limit

Breaches

>0.75mg/l

NH3-N

Current

Works

Existing Works 5m

Depth 730.4 0.092 0.027

0.59 253

4

Existing Works 5m

Depth (smaller

anoxic)

730.4 0.092 0.027

Failed on Ammonia. 0.28 44

5

246.5anoxic

1hr

Deeper ASP basin

5.7m Depth 730.4 0.081 0.024

0.15 1

5

321m3

anoxic

Deeper ASP

basin 5.7m Depth

(optimised

anoxic)

730.4 0.081 0.024

0.18 5

6 Deeper ASP basin

5.7m Depth 730.4 0.081 0.024

Passes on NH3. Appears

good on storm Anoxic

Volume might not be

enough Nitrate about 5-

6mg/l

0.13 0

7 Deeper ASP basin

6m Depth 730.4 0.077 0.022

Passes on NH3. Appears

good on storm. Anoxic

Volume might not be

enough Nitrate about 5-

6mg/l

0.11 0





Organised by Aqua Enviro

Figure 24: Scenario 4 final effluent Ammonia (F:M: 0.091)

Figure 25: Scenario 5 final effluent Ammonia (F:M: 0.081)

http://www.ewwmconference.com/





Anoxic Volume Modelling

The calibrated model was also used to size the anoxic volume by assessing the total aeration

requirements for the whole plant. Nominal anoxic retention ranging for 0.85 – 1.72 hrs was modelled.

Increasing the nominal anoxic retention from 0.85 to 1.42 hrs resulted in an approximate 5.5% reduction

in the overall aeration requirements due to better denitrification. Increasing the anoxic volume further

from 1.42 to 1.72hr did not show a discernible benefit as shown in Figure 26.

Figure 26: GPS-X Anoxic Zone Optimisation

Conclusion

Using the calibrated model to help design the future plant for Brockhampton STW has not only allowed

the project team the option to risk assess the proposed design, but it has reduced the size of the future

plant by approximately 15%.

The winter and wet weather performance dictated final size of the ASP on as all the modelled scenarios

suggested excellent summertime dry weather flow performance.

The correction of the delay in the aeration control logic has significantly reduced emergency operator

call outs to the site over the past 2 years and the improved site performance has underpinned the design

of the plant to meet the future ammonia permit.






The planned modification to the anoxic volume will reduce the future power consumption of the plant

and provide the added benefit of maintaining nitrate levels in the final settlement tank at acceptable

levels.

It is important to conclude that reliance on the Brockhampton process model was only possible because

the model was well-calibrated. Although significant effort was put into the model calibration through an

external consultant, data accumulation, process understanding and data reconciliation, the benefit to

Severn Trent Water from using the model to assess the risks has more than off-set that effort and those

costs.

Acknowledgements

I would to express my gratitude to Severn Trent Cheltenham area operational teams with their

assistance in gathering the operational data from site. Manjit Chadha (Severn Trent Asset Creation)

has been instrumental in taking forward the design proposals and is delivering the capital scheme for

the site. It was especially nice to see the works performance improve from the moment Manjit removed

the delay in the aeration and his realisation that the aeration control for Brockhampton had simply been

copied from Minworth, a site which serves a population of 1.5 million, was sobering.






HARNESSING THE POWER IN NITRIFYING SAND FILTERS

Chan, T.F.*; Koodie, T.; Sloper, M.J. and Wiggam, R.W.

Black and Veatch, 60 High Street, Redhill, Surrey RH1 1HS

*Corresponding Author Email [email protected]

Abstract

Black and Veatch installed and commissioned continuous blackwash nitrifying sand filters as the tertiary

treatment stage at 10 sewage treatment works. The filters were installed to enhance the effluent quality

of the existing works in order to meet the more stringent ammonia consents. Nitrifying sand filters have

a number of advantages including their compactness and the simplicity of their operation. This paper

has been written to present the key lessons learned and to provide guidance on achieving process

optimisation during commissioning and operation.

Keywords

Continuous backwash filter, nitrifying sand filter, process commissioning, optimisation, backwash, air

flow, dissolved oxygen profile, nutrient.

Introduction

Upflow continuous backwash filter technology was first developed in the 1970s for suspended solids

removal. The filter operates in an upflow mode through the filter media, typically sand. The feed

wastewater enters the lower section of the filter. The filter bed is moved continuously downwards,

countercurrent to the upward feed wastewater flow, by means of an air lift created using compressed

air at the base of the unit. The dirty sand/solids mixture is then passed, with added washwasher,

through the sand washer and into a wash box. The flow out of the wash box, and hence the washwater

flow, is controlled by an outlet weir which is always lower than the filtrate outlet. In the wash box, the

denser cleaned sand is separated from the lighter solids and returned to the top of the filter bed. The

solids laden washwater is diverted to the inlet works for treatment.

The key advantages the continuous backwash filter offers include compact installation, continuous

treatment and continuous dirty backwash water production. The process configuration has been further

modified to incorporate tertiary nitrification or denitrification. In a nitrifying sand filter (NSF), air is added

to the filter bed through aerators. The schematic of a typical continuous backwash nitrifying sand filter

is shown in Figure 1. For a denitrifying sand filter, a carbon source is added to the incoming wastewater.

These process configurations have since been widely applied in wastewater treatment worldwide

(Feldthusen, 2004).







Figure 1: Schematic of a nitrifying sand filter (courtesy of Hydro International).

Between 2010 and 2015, the fifth Asset Management Plan (AMP5) period, nitrifying sand filters were

installed at 10 sewage treatment works in England. A summary of the scope of supply is presented in

Table 1.

Table 1: Summary of scope of NSF supply

Parameter Range

Feed Humus tank effluent from trickling filter works

Flow 2.1 – 118 l/s

NH3-N discharge consent 3 - 6 mg/l (95th percentile)

Design NH3-N load 6.7 -53 kg/day

No of NSFs installed per site 2 – 8 units

Range of NSF bed depth 2 – 5 m

Range of NSF filter surface area 5 – 7 m2/ unit, 10 – 56 m2/ work

Range of NSF filter volume 10 – 35 m3/unit, 20 – 140 m3/work

The objective of this paper is to present the key lessons learned and to provide guidance on achieving

optimum treatment performance.

Commissioning and Operational Experience

Air Distribution System Modifications

Significant variations in the bubble patterns across some of the NSF units were observed during

commissioning at a number of works. An example of this variation is shown in Figure 2.






Figure 2: Even air distribution vs. uneven air distribution.

The larger surface area (7m2/unit) NSF units with rubber air hose distributors were installed at these

works. When one of the units was emptied, it was found that the air hose in the outermost ring was

stretched, as indicated by the movement of the fixing clamps away from the hose support, and showed

sign of folding (Figure 3). The deformation of the hose may have occurred during sand media loading

or due to the weight and continuous movement of the sand during operation. As a result of the

deformation, the air supply to part of the filter was cut off.

Figure 3: Deformed hose in the outer air distribution system.

The air distribution systems of the affected works were subsequently modified (Figure 4). A stainless

steel support was provided below the outer ring of hose to each unit. More fixing clamps were added

to limit air hose movement. An additional line was also introduced to supply air to both ends of the air

distribution hose and remove the single point of failure. The bubble pattern returned to normal

immediately after the modification and has remained so since that point.

Clamps have

moved inwards






Figure 4: Modified air distribution system.

Solids Accumulation and Remediation

After the air distribution systems had been modified, air patterns in the NSF units at a few of the works

soon deteriorated again. Upon detailed inspection, it was noted that the air lift in some of the units were

not performing sufficiently and resulted in stalled sand bed movement. Attempts were made to restart

the air lift by pulsing air through the system. When the air lift was restarted, the initial dirty washwater

contained a high concentration of solid sludge mass. After an extended period of high rate washing,

the solids concentration in the washwater decreased and the air distribution pattern returned to normal.

At one of the affected works, air lifts in the NSF units could not be restarted by air pulsing and other

attempts. When the units were drained for inspection, sludge mass as well as foreign objects were

found, having accumulated at the base of the units affected (Figure 5).

Figure 5: Sludge mass at the base of a NSF unit.

Once the sand was cleaned and replaced, normal operations could then be resumed. Therefore,

regular monitoring of sand movement, washwater flow rate and washwater quality would ensure the

backwash system is performing effectively.

Although 6mm auto-backwash filters have been provided upstream of the NSF units to prevent blockage

of the air lift system and the sand bed by coarse solids, it is necessary to ensure the solids concentration

and solid load in the feed are within the design limits. For example, more operator attention will be

required when ferric dosing is used for phosphorous (P) removal upstream. The resulting solid material

in the NSF feed can be more dense and can potentially accumulate at the base of the unit. In some






cases, this solid material may need to be removed by increasing the air lift or by occasional pulsing of

the air lift. More importantly, steps must be taken to prevent the humus tank sludge blanket from

spilling downstream and clogging the NSF units.

Effect of Air Flow on Treatment Performance

Figure 6 shows the percentage NH3-N removal during several sampling periods for one of the NSF

units. Both lab-tested and in-situ test kit results are included. It can be seen that the change in process

air flow rate from 260 l/min to 400 l/min prompted a step improvement in NH3-N removal efficiency. This

increase in the removal percentage is clear in both composite and spot samples. When the air flow

was reduced, the removal efficiency dropped.

Figure 6: Time-line showing process air to NSF and ammonia-N removal efficiency.

Figure 7: Ammonia-N removal percentage versus process air flow rate.

0

50

100

150

200

250

300

350

400

450

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Jan Feb Mar Apr May Jun Jul

Air

Flo

w (

l/m

in)

NH

3-N

% R

emo

val

NH3-N Removal % (Test Kit Spot) NH3-N Removal % (Lab Spot)NH3-N Removal % (Test Kit Comp) NH3-N Removal % (Lab Comp)Process Air

20%

30%

40%

50%

60%

70%

80%

200 220 240 260 280 300 320 340 360 380 400

NH

3-N

% R

emo

val

Process Air Flow Rate (l/min)

200l/min 230l/min 260l/min 400l/min






Figure 7 shows the information in Figure 6 by plotting percentage NH3-N removal against process air

flow rate. At the design air flow rate of 260 l/min, NH3-N removal efficiencies between 39% and 65%

were achieved. Increasing the air flow rate appears to show a trend in increased removal efficiency.

At the highest air flow tested, 400 l/min, the load removal was higher varying between 61% and 76%.

These results demonstrate the impact of increased air flow rate on the treatment of ammonia. While

higher air flows do not guarantee greater removal of ammonia, an improvement in removal performance

is indicated.

Dissolved Oxygen Profile

Dissolved Oxygen (DO) concentrations were measured within the same NSF unit through the use of a

modified solid-state DO probe. The probe was encased in a perforated protective shell and attached

to an aluminium pole. The shell prevented the ingress of sand, while allowing fluid to move freely to

the probe. DO measurements were taken during operation by lowering the modified probe into the

filter, stopping at incremental depths within the sand bed to allow the DO reading to stabilise.

Measurements were taken in four regions around the NSF unit at North, South, East and West,

approximately halfway along the filter radius from the centre.






Figure 8: DO profile at four locations around the NSF at different process air flow rate.

Figure 8 shows the DO profiles through the filter bed at different process air flow rates and in the four

regions. At -0.5 m, the results show the DO concentration within the liquid above the sand bed. 0.0 m

indicates the upper surface of the sand bed. Subsequent measurements are at the indicated depths

below the surface of the bed.

Prior to this test, the standard procedure for ensuring that adequate DO is supplied for nitrification has

been to measure the concentration in the liquid at the top of the filter, i.e. the filtrate. Barter and Smith

(2007) indicated that a DO concentration of 6 mg/l in the filtrate is indicative of sufficient oxygen being

present throughout the bed. The results presented in Error! Reference source not found. 8 have

demonstrated that the DO measurements at the top of filter or in the filtrate do not necessarily represent

the DO concentration within the filter bed.

DO concentration is known to be limiting to the growth of nitrifying bacteria below 2-4 mg/l Halling-

Sørensen and Jorgensen (1993). The results in Figure 8 show that under conditions of lower process

air flow (below 300 l/m) certain sections of the filter were very low in dissolved oxygen. These zones

of low DO concentration were likely to have lower nitrifying activity thus to contribute to a lower removal

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0 5 10 15

De

pth

in S

and

Bed

(m

)Dissolved Oxygen Concentration (mg/l)

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0 5 10 15

De

pth

in S

and

Bed

(m

)

Dissolved Oxygen Concentration (mg/l)

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0 5 10 15

Dep

th in

San

d B

ed (

m)

Dissolved Oxygen Concentration (mg/l)

300l/min

250l/min

200l/min

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0 5 10 15D

epth

in S

and

Bed

(m

)

Dissolved Oxygen Concentration (mg/l)EAST NORTH

SOUTH WEST






efficiency. This concurs with the lower NH3-N removal efficiency presented in Figure 6 and Figure 7

when the process air flow rate was at 200 and 230 l/min.

The zones of lower DO concentration were more prominent in the “West” regions of the filter bed when

the air flow rate was at 200 and 250 l/min. Other sections maintained higher DO concentrations even

at process air flow rates below the design level. This indicates that the supply of process air was less

evenly distributed at lower air flow rates.

Alkalinity and Phosphorous as Nutrient

At works where chemicals such as ferric compounds are added to enhance primary treatment

performance or for phosphorous (P) removal, it is necessary to ensure sufficient residual alkalinity is

available in the NSF feed to meet nitrification requirements.

When the P consent becomes more stringent, there is a drive to dose more chemical to ensure the

effluent quality meets the discharge consent. Chemical addition can potentially be too efficient in

removing P from the wastewater and the residual P concentration can become limiting for tertiary

nitrification processes. Nordeidet, et al. (1994) reported tertiary nitrification in a biofilm reactor was

limited when influent P concentration was lower than 0.15 mg PO4-P/l. Therefore, the chemical dosing

regime should be reviewed against the actual diurnal P loading pattern and the influent P level should

be monitored regularly to ensure there is sufficient P available in the NSF feed water.

Recommendations

Based on our commissioning and operation experience, NSF performance can be optimised by the

following:

• Ensure the air distribution system has been designed and constructed to suit the configuration

of the unit and operating environment.

• Gross solids in excess of the design load, such as humus tank blanket spill, should be

prevented from entering the NSF units.

• Effectiveness of the sand washing system should be checked by regular monitoring of sand

movement, washwater flow rate and washwater quality.

• Increased air flow can improve the uniformity of air distribution across sand bed, while dissolved

oxygen concentrations in the filtrate do not necessarily represent the DO concentration within

the bed.

• Increasing air flow may improve treatment performance and sufficient air must be provided to

the NSF which may be above the theoretical aeration rate to optimise treatment performance.

• At works where chemicals are dosed to enhance primary treatment performance and/or P

removal it would be prudent to ensure alkalinity and nutrients (P in particular) are available for

nitrification.

Acknowledgement(s)

The authors would like to thank the project team for their contributions, observations and suggestions

during the operations of the unit.

References

Barter, P., Smith, J. (2007). Using Tertiary Aerated Sand Filters for Ammonia Removal. In: European

Water & Wastewater Management Conference. Newcastle upon Tyne, UK. 24-26 Sep 2007.






Feldthusen, F. (2004). Continuous Sand Filters - Tertiary WWT and Other Applications. In: SAWEA

Workshop. Dammam, KSA. 22 March 2004.

Halling-Sørensen, B., Jorgensen, S.E. (1993). The Removal of Nitrogen Compounds from Wastewater.

Elsevier Science.

Nordeidet, B., Rusten, B. & Ødegaard, H. (1994). Phosphorus Requirements for Tertiary Nitrification in

a Biofilm. Wat. Sci. Tech., Volume 10-11, pp. 77-82.






NUTRIENT REMOVAL WITH MICROALGAE – REDUCTION OF THE EFFLUENT

CONCENTRATION FROM WASTEWATER TREATMENT PLANT

Wawilow, T.1; Hasport, N.1, Theilen, U1 and Thomsen C.2 1THM - University of Applied Sciences - ZEuUS, Germany, 2Phytolutions GmbH, Germany


Abstract

An alternative, environmentally sustainable method to remove nutrients from wastewater is to integrate

an algae-mediated wastewater treatment to reduce nutrient loads to preserve water bodies from

eutrophication and generate effective biomass. Compared with conventional treatment methods the

generated microalgae biomass is more energetic and rich in content with phosphorous (P) and nitrogen

(N). Therefore, a tertiary biotreatment coupled with the production of potentially valuable biomass, which

can be used for energetic or material purposes, is an efficient alternative to avoid using chemicals for

the removal of phosphorus via precipitation and flocculation. Algae species Scenedesmus was applied

for wastewater treatment and had proven abilities of removing nitrogen and phosphorous in retention

time of 24 hours. In this study, a photobioreactor (PBR) was implemented for large-scale research to

treat the wastewater treatment plant (WWTP) effluent while microalgae growth rate, nutrition removal

as well as operational and external conditions were evaluated. Moreover, the biomass was separated

and methane potential tests were conducted using microalgae as substrate.

Keywords

Effluent concentration, microalgae, nutrient removal, photobioreactor, wastewater treatment

Introduction

Recent national and European regulations decrease the discharge limits of contaminants from

wastewater treatment plants (WWTP) and in particular the regulation focused on nutrient removal (N

and P) in order to reduce the eutrophication/contamination in water bodies.

Usually techniques in wastewater treatment to reduce N or P are biological or physico-chemical

methods. To avoid using chemicals for the phosphorous removal a cost-effective alternative and

environmental treatment to reach more stringent limits is to integrate a PBR for algae-mediated

wastewater treatment as last step of biological treatment. In the study at the WWTP in Rotenburg-

Braach, Germany, a part of effluent from plant was treated by microalgae in PBR from Phytolutions

company, to demonstrate the alternative process and improve to avoid using chemicals for

phosphorous removal.

Compared to biological process in WWTP the microalgae depend on sunlight for higher growth rates

and for faster elimination rates because microalgae are like plants and require sunlight, CO2 and water

(nutrients) to growth. During the photosynthesis microalgae convert the solar energy into chemical

energy and produce oxygen and biomass. Furthermore, microalgae require nutrients to generate

biomass particularly nitrogen and phosphorous and many other micronutrients that are involved in

wastewater streams. The objective of this study was to use microalgae to remove nutrients from WWTP

effluent in order to prevent eutrophication. Algae species Scenedesmus was applied for wastewater

treatment and had proven abilities of removing N and P in retention time of 24 hours. The green

microalgae Scenedesmus has shown extraordinary vitality in urban wastewaters and these freshwater







algae tolerate a wide range of temperature and pH. The fast growing was limited by sunlight. The

Hessian Ministry of the Environment and the Stadtwerke Rotenburg are funding the ongoing research

project to implement a photobioreactor for large-scale research to treat the WWTP effluent with

microalgae. Moreover, the biomass was separated and methane potential tests were conducted using

microalgae as substrate.

Materials and Methods

Phytoplant – PBR System

In order to remove nutrients from the wastewater with microalgae, the following operating scheme

(Figure 1) was used in the research project in Rotenburg (a.d. Fulda, Germany). The WWTP of

Rotenburg is a conventional plant with nitrification and denitrification, additionally applying chemicals.

The pilot plant was built on an area of 200 m² at the WWTP area in Rotenburg. The plant volume is

10,000 l.

Operating Process

A temperature control-system utilized waste heat or cooling water to create a continuous optimal

environmental condition for algae growth. The Phytoplant-process operates as an outdoor fed batch

reactor, and was monitored and controlled for one seasonal cycle. The exchange volume was 10 or 20

% of the whole plant volume once a day.

PBR Phytobagsystem

The pilot plant consists of a photobioreactor from Phytolutions-Phytobagsystem, semi-closed flat-panel

photobioreactors produced from multi-layer plastic, 100 cm high and 7 cm wide. The Phytobags are

Figure 1: Phytoplant operating scheme






provided with a micro bubble airlift system that generates the turbulence that facilitates the access to

nutrients, CO2 and light to the microalgae (Phytolutions, 2016).

Characteristic wastewater effluent

The algae treatment process should remove the nutrients after conventional biological wastewater

treatment. Therefore, the treated water in the algae PBR was the effluent from the WWTP in

Rotenburg-Braach. The WWTP treats mainly urban wastewater with no special industrial influence.

Table 1 presents the yearly average value concentration for WWTP outflow and the limit value for the

effluent that must be strictly adhered.

Table 1: Yearly and limit value for WWTP effluent in Rotenburg

Parameter

Yearly average value 2016 Limit value

[mg/l] [mg/l]

BOD5 2.75 20.0

COD 18.50 55.0

Ptot 0.57 2.0

NH4-N 1.3 10.0

The concentration of phosphorous in the outflow of the WWTP is very low, because the plant uses

chemicals to reduce the concentration of P. To simulate the outflow from the WWTP without physico-

chemical treatment, the nutrient concentration in the PBR was partly raised up with artificial nutrient salt

(particularly regarding N and P). This way it was possible to simulate the elimination rate for higher

concentrations in wastewater streams.

To increase the biomass concentration in the system, the separated biomass was returned to the

Phytobags.

Analysis and monitoring

For the determining organic and inorganic parameters, the samples were taken from the INFLOW an

OUTFLOW reservoir before or after every exchange cycle every day.

The determination of COD, NH4-N, NO3-N, total nitrogen Ntot, total phosphorous Ptot, PO4-P and total

organic carbon (TOC) was carried out by using photometric cuvette tests (Hach-Lange, Duesseldorf,

Germany). The OUTFLOW samples of PBR were filtered with a 0.45 µm cellulose acetate filter before

analyzing.

In this study, the determining of the biomass was carried out with two different methods: optical density

at 750 nm and the determination of Total Syspended Solids (TSS). The microalgae strains were

identified based on morphological differences using a microscope. The number of cells was counted to

characterize the biomass mixture.

Temperature, pH and OD 750 were measured online during the whole time from January until October.

Spectrophotometric analyses were used for measurement of biological activity.







Microalgae strains

The microalgae used in the study were obtained from Phytolutions GmbH, Bremen. Before starting the

treatment process the algae species should acclimate in wastewater as new medium step by step to

avoid algae dying

The results of monitoring on algae strains during warm summer shows a change of species. The

reason for this change could be the high temperature or the input of different algae from the

wastewater, which can cope better with the conditions at hand and therefore grow faster than the first

species (Figure 2).

Nutrient removal

The following figures show the concentration of in- and outflow of NO3-N (Figure 3), NH4-N (Figure 4),

Ptot (Figure 5) and PO4-P (Figure 6). In the summertime, the inflow nutrient concentration was raised

up with artificial nutrients to simulate higher outflow concentrations from WWTP without precipitation.

The nitrogen removal rate is very high and the outflow concentration from the PBR is < 1 mg/l.

Figure 3: NO3-N concentration Figure 4: NH4-N concentration

0

5

10

15

20

25

30

35

40

45

50

J F M A M J J A S O

Concentr

ation [m

g/l]

IN NO3-N [mg/l] OUT NO3-N [mg/l]

0

1

2

3

4

5

J F M A M J J A S O

Concentr

ation [m

g/l]

IN NH4-N [mg/l] OUT NH4-N [mg/l]

Figure 2: Abundance of different microalgae (Phytolutions, 2016)






Similar to the nitrogen concentration figures 5 and 6 show the concentration of phosphorous. The

phosphorous concentration decreased from 3 - 11mg/l to < 0.5mg/l. At the end of the process time, the

removal slowed. The nutrient removal is influenced by microalgae growth and algae species.

Figure 5: Ptot concentration Figure 6: PO4-P concentration

Table 2 summarizes the results as average concentrations from the strains and nutrient removal rate.

The microalgae showed high ammonium and nitrogen removal rate. The removal rate of phosphorous

slowed down during the last phase of the project.

Nitrogen was continuously removed thoroughly and the removal rate increased over 80 %. The removal

rate for phosphorous is depending on microalgae strains and was between 55 % and 97 %.

Parameter Outflow

WWTP

Outflow WWTP +

N and P addition

Outflow PBR Removal rates

NH4-N [mg/l] 0.3 - 3.1 < 5 0.02 - 0.16 85 - 98 %

NO3-N [mg/l] 1.2 - 5.3 ca. 15 - 35 0.23 - 0.90 82 - 97 %

PO4-P [mg/l] 0.4 - 0.6 3.0 - 7.0 0.04 - 0.38 72 - 99 %

Ptot [mg/l] 0.5 - 1.1 3.8 - 8.5 0.05 - 0.48 55 - 97 %

Conclusions

The results from the study demonstrated the feasibility of nutrient removal with microalgae. This

technology presented a unique opportunity to achieve nutrient removal and production of algal biomass

for energy recovery. On a large scale the PBR was integrated into the WTTP process and operated

with Scenedesmus sp. and Pandorina Sp. placed in Rotenburg a. d. Fulda, Germany, achieving a

removal of more than 82 % N and 55 % P. The results reported in this study describe the way to increase

the nutrient removal performance of microalgae and biotechnology in wastewater treatment.

The next step is to provide a guide for the use of microalgae for nutrient elimination. Furthermore, the

economic boundary should be determined on the one hand by the necessary investments and operating

costs, on the other hand by the saved precipitant costs as well as the energy yield from use of the algae

biomass for methane yield.

However, there is still a large scope for the improvement and optimization of the process.

0

2

4

6

8

10

12

J F M A M J J A S O

Concentr

atio

n [

mg/l]

IN Ptot [mg/l] OUT Ptot [mg/l]

0

2

4

6

8

10

12

J F M A M J J A S O

Concentr

ation [

mg/l]

IN PO4-P [mg/l] OUT PO4-P [mg/l]






Aknowledgements

This research project is funded by The Hessian Ministry of the Environment and the Stadtwerke

Rotenburg a. d. Fulda.

References

2000/60/EC, (2000). European Parliament and of the Council (2000/60/EC); 2000 Directive

2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a

framework for Community action in the field of water policy [Online]. Available at http://eur-

lex.europa.eu/legal-content/EN/TXT/?qid=1487946213674&uri=CELEX:32000L0060.

Barbot, Y. N., Thomsen, L., & Benz, R. (2015). Thermo-Acidic Pretreatment of Beach Macroalgae

from Rügen to Optimize Biomethane Production—Double Benefit with Simul-taneous Bioenergy

Production and Improvement of Local Beach and Waste Management. Marine Drugs, 13(9), 5681-

5705.

Hasport, (2017). Hasport, N. Nährstoffelimination im Ablauf einer kommunalen Kläranlage mit

Mikroalgen, Masterthesis, Fachbereich Bauwesen der Technischen Hochschule Mittelhessen,

Gießen, 2017, unpublished.

Phytolutions, (2016). Phytolutions Systems [Online]. Available at www.phytolutions.de (Accessed 22

February 2017).






1-STEP® FILTER: THE SOLUTION FOR COST-EFFECTIVE REMOVAL OF

PHOSPHOROUS AND OTHER PRIORITY CHEMICALS FROM WWTP EFFLUENT

Kramer, J.F.1, Menkveld, H.W.H.2, Bechger, M.3, Cunliffe, T.2, Nieuwenhuijzen, A.F. van1 1 Witteveen+Bos Consulting Engineers, NL, 2 Nijhuis Industries UK and Ireland, UK, 3 Waternet, NL


Abstract

To ensure a cleaner and protected environment and human health conditions for the future, both UK

and EU environmental legislation require further improvement of surface water quality by advanced

wastewater treatment. Ultra-low phosphorous standards (< 0.25 mg P/L) altogether with low nitrogen

and total suspended solids standards are on the way. As a consequence, additional treatment and

extension of wastewater treatment plants (WWTPs) by tertiary treatment becomes necessary. The 1-

STEP® filter is a compact, cost-effective, efficient and effective total tertiary treatment solution,

developed and successfully implemented in the Netherlands over the last decade. This paper presents

the results of full operation over five years (2012-2017) of the 1-STEP® filter at WWTP Horstermeer of

Waternet, Amsterdam. The 1-STEP® filter combines four distinct processes in one filter unit: (1)

coagulation and flocculation of phosphorous and suspended solids, (2) denitrification of nitrate, (3)

filtration of flocs and particles (4) adsorption of micro pollutants.

Keywords

Advanced Phosphorous removal, Advanced Nitrogen removal, Micro pollutant removal, Priority

chemicals removal, total effluent polishing, 1-STEP® filter, Biological activated carbon filtration

Introduction

Authorities responsible for the quality of surface and groundwater have increasing attention for

improving effluents from wastewater treatment plants (WWTPs). This directly results from the

implementation of the European Union’s Water Framework Directive (WFD) in 2000. The WFD requires

that good ecological and chemical conditions for both surface and groundwater are achieved in 2015.

Part of the WFD is a list of substances that have to be dealt with in high priority with regard to the impact

on the water quality. To deal with these substances, the removal of these substances (partly from

WWTP effluent) should be increased. It is emphasized here that the water quality requirements

mentioned in WFD are only related to surface water and not directly to WWTP effluent.

The WFD and UK legislations focus on a few different groups of substances, of which nutrients (N and

P), suspended solids, organic micro pollutants and priority chemicals are some prominent examples.

For the UK especially a standard of 0.25 mg P/L is expected, at this time considered as an ultra-low

phosphorous standard. Studies have shown that the common activated sludge systems are unable to

remove these substances sufficiently to newly required low levels. At the moment, a few technologies

are available that effectively and consistently produce WFD compliant effluents, but only a limited

number of these technologies cost-effectively deals with all relevant substances in one compact total

tertiary treatment solution. The 1-STEP® filter is an advanced example of these technologies.

Water and Wastewater Companies in England and Wales are at present implementing the capital

expenditures according to the performance commitments over the various AMP6 plans, to meet their

existing commitments and targets for customers, the regulators and stakeholders. Meanwhile, the

methodology for the 2019 price review (PR19) is about to be finalised. Further quality enhancement







schemes might be included as performance commitments in the “new-to-be-developed” AMP7 plans

for 2020-25, requiring a significant improvement on the 2015-20 levels of performance. This due to

more stringent standards, amongst others for Phosphorous, but also to comply with the upcoming Water

Industry National Environment Programme (WINEP) currently being drafted by Environment Agency

and a final version expected by March 2018. Results from the ongoing Chemical Investigation

Programme 2 with demonstration trials and pilot studies will be included in the PR19 submission of

Water and Wastewater Companies to Ofwat. The initial findings of the trial plants show there is no

´Magic Bullet´ that is effective on all problem substances so far. The 1-STEP® filter might be an effective

solution for various problem substances, as demonstrated at WWTP Horstermeer of Waternet in the

Netherlands.

Methods

The “One Step Total Effluent Polishing filter” or 1-STEP® filter development has been initiated in March

2005 by Witteveen+Bos Consulting Engineers, TU Delft, CABOT-Norit and Waternet (the Water and

Wastewater Company of the city of Amsterdam). The goal was to create a cost-effective filter that was

capable of removing nutrients and suspended solids from WWTP effluents to comply with the WFD

requirements. The 1-STEP® filter has been developed by a ‘Research & Development to Construction’

project in the Netherlands and has proven to be reliable, flexible, simple in design, easy to operate, and

very robust by both demonstration scale (up to 10 m3/h) and full scale operation (from 250 up to 1,250

m3/h).

Figure 1. Schematic representation of

four 1-STEP® filter processes in one

filter unit: (1) Coagulation and

Flocculation of phosphorous and

suspended solids, (2) Denitrification of

nitrate, (3) Filtration of flocs and

particles (4) Adsorption of micro

pollutants






The proof-of-principle of combining the four processes, namely filtration, advanced coagulation (P

removal), biological removal (N removal) and activated carbon adsorption (removal of priority and

emerging substances) was proven at pilot scale at several WWTPs in The Netherlands and

demonstrated at the WWTP Horstermeer (East of Amsterdam). With the laboratory scale results a pilot

installation was designed, built and tested. Subsequently, the 1-STEP® filter was built on the WWTP

Horstermeer on full-scale (1,250 m3/h) in 2012, where it has been successfully producing WFD

compliant effluent and the strict legal discharge limits set for the receiving surface water.

The 1-STEP® filter, shown in figure 1, is a fixed bed filter operated with a downward flow. The filter

medium is a robust type of activated carbon, specially developed for this technology solution. Next to

the removal of suspended solids which is typical for filters, this filter also achieves simultaneous

biological denitrification, physical/chemical removal of phosphate and removal of micro pollutants and

priority substances by adsorption to the activated carbon.

At present, all these features require extensions and additional treatment units at existing WWTPs.

Conventional treatment units lead to high investment and operational cost since several successive

treatment processes have to be applied with additional high energy and chemical consumption, and

associated sludge streams. The 1-STEP® filter is an innovative compact fixed bed activated carbon

filter operated at a relatively high rate downward flow combining four processes in one single additional

treatment unit. Besides removal of suspended solids by filtration, it performs excellently on nitrogen

removal by simultaneous biological denitrification (using a selective carbon source), chemical

phosphate and heavy metals removal (by coagulation and flocculation with a low dose of metal salt)

and, if required, removal of organic micro pollutants by adsorption to the activated carbon.

The uniqueness of the 1-STEP® filter lies in the integration of four treatment processes in one filter unit

saving significantly on investment costs and plot size while producing an excellent and stable treated

effluent quality with ultra-low concentrations on phosphorous, nitrogen and suspended solids.

Optimised biological, chemical and hydraulic conditions within the 1-STEP® filter guarantee low

chemical and energy consumption with respect to the demonstrated treatment performances.

The 1-STEP® filter ensures i) advanced nutrient (N and P) removal to minimise eutrophication of surface

waters and algae blooming, ii) effective TSS removal, iii) removal of heavy metals and hazardous micro

pollutants to protect ecological and human health, iv) improving treated effluent quality (by removal of

total suspended solids) for future water reuse purposes, and v) greater stability of treated effluent to

safeguard against consent failure.

Results

Advanced nutrient removal during the Monitoring Programme 2013

The period from January 1st till September 22nd 2013 has been considered as the monitoring (or “full-

scale trial”) period of the 1-STEP® filter at WWTP Horstermeer. From January 1st 2013 performance

control measurements were executed to verify whether the filter removes total nitrogen (N-total), nitrate,

phosphorous (P-total) and orthophosphate as specified in the contract. The conclusion was that the

filter complies to the contract. As other construction works at WWTP Horstermeer continued during this

monitoring period, the monitoring program was troubled by regularly off-spec filter feed water quality

(compared to design parameters) especially with regard to the concentration of phosphorous. Although

the filter performance was robust, the removal of nitrogen and phosphate, averaged over the whole

monitoring period, was less than expected based on the design. The concentrations and removal

efficiencies of N-total, nitrate, P-total and orthophosphate, averaged over the entire monitoring period,

are presented in table 1.






During the monitoring programme, the average suspended solids concentration in the feed was 9.6 mg

DS/L and in the filtrate 6.5 mg DS/L, resulting in an average removal efficiency of 32%. The COD

removal was 19% on average which resulted in an average concentration in the filtrate of 29.9 mg O2/L.

The removal efficiency of nutrients does not decrease over time. The removal efficiency of these

processes is not related to adsorption, so saturation of the activated carbon has no influence.

Table 1: Nutrient removal performances 1-STEP® filter Monitoring Program 2013

Figure 2: The 1-STEP® filter at WWTP Horstermeer (Waternet) on June 20th 2017

The 1-STEP filter at WWTP Horstermeer is in continuous operation from January 1st 2013 onwards.

After the monitoring period, performance control measurements have been executed as part of the

operation of the WWTP and in accordance with the environmental permit.

1-STEP® filter performance 2013-16

In addition to the monitoring period data summarized in Table 1, about four years of operational data is

available to demonstrate the performance of the filter unit. The first three diagrams below show seven

years of operational data of WWTP Horstermeer. The years 2010-12 show the concentrations of the

specified parameter without the 1-STEP® filter and the years 2013-16 with the 1-STEP® filter in

operation.






Figure 3: Phosphorous removal performance 1-STEP® filter: Portho

Figure 4: Phosphorous removal performance 1-STEP® filter: Ptotal

Figure 5: Nitrogen removal performance 1-STEP® filter

0,0

0,1

0,2

0,3

0,4

0,5

2010 2011 2012 2013 2014 2015 2016

PO

4-P

(m

g/l)

PO4-P Removal 1-STEP® filter Horstermeer

PO4-P in STP Effluent PO4-P in 1-STEP Treated Effluent

without 1-STEP 1-STEP in operation

0,0

0,2

0,4

0,6

0,8

1,0

2010 2011 2012 2013 2014 2015 2016

Ph

osp

ho

rou

s to

tal

(m

g/l)

Ptotal Removal 1-STEP® filter Horstermeer

Ptotal in STP Effluent Ptotal in 1-STEP Treated Effluent


0

2

4

6

8

10

12

14

2010 2011 2012 2013 2014 2015 2016

Nit

roge

n t

ota

l (m

g/l)

Nitrogen Removal 1-STEP®Filter Horstermeer

Ntotal in STP Effluent Ntotal in 1-STEP Treated Effluent







Figure 6: Yearly specific energy consumption of the 1-STEP® filter (2013-16)

Micro pollutant removal

Only three of the 45 substances from the WFD priority substances were found above the detection limit

in the effluent from WWTP Horstermeer. These three substances were the pesticide diuron and the

metals lead and nickel. These results are positive regarding the impact of the discharge of WWTP

Horstermeer on the concentration of priority substances in the river De Vecht, the receiving surface

water. But it also means that during the monitoring period in 2013 only limited information was obtained

regarding the removal of priority substances by the 1-STEP® filter. The concentration of diuron in the

effluent of WWTP Horstermeer was far below the quality limit for surface water and just above the

reporting limit. In the filtrate, diuron was no longer detected above the reporting limit. The concentration

of the metals lead and nickel in the feed to the filter were also below the quality limit for surface water.

Lead was removed with 20 to 70% up to 22,500 treated filter bed volumes. Unfortunately, nickel was

not removed.

Figure 7: Specific pharmaceuticals removal performance over time and bed volumes

Only micro pollutants in the group of pharmaceuticals (which are not WFD priority substances, but are

worldwide considered as emerging substances) were often detected above the reporting limit as well in

the feed to the filter as in the filtrate. In total 44 pharmaceuticals were analysed of which 34 were

reported more than once above the reporting limit. Of the 34 pharmaceuticals, 21 were detected in

concentrations that were higher than the concentration in the river De Vecht. Most of these

pharmaceuticals were removed by 30 to 90% at the start of the filter run. After approximately 15,000 -

0

0,02

0,04

0,06

0,08

0,1

2013 2014 2015 2016

Ene

rgy

con

sum

pti

on

kW

h/m

3

Specific energy consumption1-STEP® filter Horstermeer






20,000 treated bed volumes (filter runtime of 4.5 - 6 months at WWTP Horstermeer) the removal

efficiency for most pharmaceuticals is reduced to 0%.

The monitoring programme indicated that the assumption “WWTPs are considered as point sources

of WFD priority substances” is not valid. For WWTP Horstermeer most of the priority substances were

not detected (discharge limits are above detection limits) at all.

Operational aspects and energy efficiency

Coagulant (PAX 214) for removal of orthophosphate was dosed to the full-scale installation with an

average Me:P molar ratio of 3.1. This was lower than the advised ratio of 4 based on the pilot

research.

The average amount of methanol used for denitrification in the full-scale installation, during the entire

monitoring period, was 4.2 g COD/g NOx-N. After optimisation in the beginning of May 2013 the

average ratio was 3.9 g COD/g NOx-N. In both cases the ratio was lower than the advised ratio of 4.5

g COD/g NOx-N based on the pilot research.

Consumption of electricity by the full-scale 1-STEP® filter was on average 0.04 kWh per m3 of treated

water. The expected electricity consumption based on the pilot research was 0.06 kWh per m3 and

the full-scale filter unit proves to be more energy efficient than the demonstration plant.

The average total downtime related to backwashing is about 5%. The overall experience with the full-

scale 1-STEP® filter is that during normal operating conditions there are limited to no process upsets.

Optimizations in the process can be easily implemented. Other important lessons learned are:

- Despite the unstable supply of nitrate and dosing of methanol during start-up, denitrification

started rapidly and performed well.

- Positioning of analysers in a side stream can result in unrepresentative results due to side effects.

Sufficient mixing and flow through of a side stream is crucial for representative analysis.

- Excess dosing of methanol resulted in H2S gas production. H2S gas safety and optimization of

dosing of methanol are important aspects of filter operation.

- A high load of solids on the filter combined with high load of organic material, nitrate and

orthophosphate can quickly result in clogging of the filters (nearly all simultaneously). This results

in a problem with backwashing of the filters. In case of extreme situations manual intervention is

required to prevent further problems.

Concluding Summary

This 1-STEP® filter would efficiently and cost-effectively improve the quality of the WWTP treated

effluent when compared with conventional tertiary treatment processes such as sand and drum

filtration.

- Excellent Treated Effluent quality:

· Ntotal < 2.2 mg/l: 60-70% removal performance;

· NO3-N < 0.75 mg/l: 82-89% removal performance;






· Ptotal < 0.15 mg/l: 60-77% removal performance;

· PO4-P < 0.03 mg/l: 78-90% removal performance;

· TSS < 5 mg/l: 25-55% removal performance;

· removal performance of 20-85% for a broad spectrum of high priority organic micro pollutants,

depending on their logKow;

· removal of the bulk of heavy metals like Cu and Zn.

- Efficient process:

· high filtration rates: up to 15 m3/m2h;

· nitrogen removal load up to 4 kg N/m3 filter bed volume;

· energy efficient (long filter run times, efficient backwash water consumption): < 0.033 kWh/m3;

· superior pre-treatment for membrane filtration (UF/NF/RO) due to low specific ultrafiltration

resistance (SUR decrease: 20 - 70 %).

- Low chemical consumption (efficient metal precipitant and carbon source dosage) due to optimal

dosing methods:

· Me/P-ratio: 1.4 - 1.7 mole Al3+/mole PO43-;

· C/NOX: 3.5 - 4.5 kg COD/kg NOX.

- Cost-effective process:

· 30-50% saving in investment- and operational costs compared to conventional successive

processes (Horstermeer WWTP experience);

· low Totex: 3 - 6 pence per m3 (N, P and TSS) depending on flow rate, excluding reactivation of

activated carbon.

Future Opportunities

The results of the research on the 1-STEP® filter led, in the perspective of the European Union’s Water

Framework Directive, to a unique approach for the advanced treatment of WWTP effluent focused on

the removal of nitrogen and phosphate combined with priority substances. The technology is registered

and the 1-STEP® filter is available as a patented black box solution. The system is unique because the

biofilm on an ordinary granular activated carbon filter, which is typically used for “clean water streams”

instead of WWTP effluent, is much thinner.

In a filter, also priority substances are removed in some extent. A large amount of the present bacteria

are removed by filtration. Due to the low phosphorus and nitrogen concentration re-growth will be

minimal. In the pilot study at the WWTP Horstermeer, it is determined that after filtration of the WWTP

effluent, the standards based on the WFD water quality objectives are met. The development of the 1-

STEP® filter is a technological step in improving the quality of WWTP effluent compared to the currently






available post-treatment methods, such as sand filtration (continuous and fixed bed). The filtrate water

is of such high quality that opportunities are created for further treatment and reuse of the produced

water.

References

Ofwat (2017). Delivering Water 2020: Consulting on our methodology for the 2019 price review. July

2017. Available from: www.ofwat.com.uk.

Graig, M. (2017). Chemicals Investigations Programme 2. Presentation made at the WWT

Wastewater 2017 Conference, Birmingham, UK, January 31st 2017.

Martijn, A.J. (2015). Impact of the water matrix on the effect and the side effect of MP UV/H2O2

treatment for the removal of organic micropollutants in drinking water production, Wageningen

University. Available from: http://edepot.wur.nl/364041.

STOWA (2013). Monitoring 1-STEP® filter Horstermeer. Amersfoort: STOWA report 2013-35.

Available in Dutch only (with summary in English).

Scherrenberg, S.M., Neef, R. de, Menkveld, H.W.H., Nieuwenhuijzen, A.F. van & Graaf, J.H.J.M. van

der (2012). Investigating phosphorus limitation in a fixed bed filter with phosphorus and nitrogen pro-

file measurements. Water Environment Research, 84(1), 25-33.

Scherrenberg, S.M., Kloeze, A.M. te, Janssen, A.N., Nieuwenhuijzen, A.F. van, Menkveld, H.W.H.,

Bechger, M & Graaf, J.H.J.M. van der (2010). Advanced treatment of WWTP effluent with filtration

leading to a pre-treatment technique for membrane filtration. Water Science & Technology, 62(9),

2083-2089.

STOWA (2009). Pilotonderzoek RWZI Horstermeer 1-STEP® filter als effluent polishing techniek.

Utrecht: STOWA report 2009-34; 2009. Available in Dutch only (with summary in English).

Scherrenberg, S.M., Menkveld, H.W.H., Bechger, M. & Graaf, J.H.J.M. van der (2009). Phosphorus

and nitrogen profile measurements to locate phosphorus limitation in a fixed bed filter. Water Science

& Technology, / Volume 60, Issue 10.




65

PRACTICAL APPLICATION OF MODULAR OFF-SITE BUILD: A COMMISSIONING

PERSPECTIVE

Baird, A. Technical Director, WPL, UK

Email [email protected]

Abstract

A range of biological wastewater treatment technologies is available from manufacturers in the UK including

variations on the submerged aerated filter (SAF). One of those companies is WPL which has supported

postgraduate university research at Cranfield University on the internal hydrodynamics of reverse cyclic

SAF reactors. The research was applied in a pilot engineering design project investigating how the rate of

removal of organic contaminants can be improved by reducing the ratio of media to wastewater inside the

tank. The focus of the research and development work has been to develop a WPL Hybrid-SAF™, which

incorporates offsite build and modularisation to aid the efficiency of installation and commissioning as well

as enhanced process and energy efficiency. The hybrid model, which has a patent pending, has been

applied in nearly 10 UK water utility and large commercial projects for carbonaceous and tertiary

nitrification. Substantial efficiency gains have been achieved in terms of process and energy consumption

as well as installation. In one application a 2,000 population equivalent (PE) plant was installed in a single

morning, with commissioning commencing immediately. Furthermore, 3D computational fluid design (CFD)

modelling of the treatment process for an individual site makes it possible to offer an accurate estimation

of project resource requirements, including process requirement, workforce skills, costings, construction

and commissioning.

Keywords

Chemical parameters

BOD5 Biochemical oxygen demand (5-day test)

COD Chemical oxygen demand

DO Dissolved oxygen

MLSS Mixed liquor suspended solids

MLVSS Mixed liquor volatile suspended solids

NH4+ Ammonium ion

NO3- Nitrate nitrogen

NTU Nephelometric turbidity units

O2 Oxygen

TSS Total suspended solids

VSS Volatile suspended solids

Reactor acronyms

BAF Biologically aerated filter

CSTR Continuously stirred tank reactor

IFAS Integrated fixed film activated sludge

PFR Plug flow reactor

SAF

WPL SmartCell™

Submerged aerated filter

WPL’s patent pending design

Introduction

Water scarcity and need to reuse water are increasing the demand for packaged wastewater treatment

plants globally. In addition, regulation is tightening as governments face growing environmental concerns

about wastewater disposal. Market analyst Technavio has projected growth in the global packaged

wastewater treatment market at a compound annual rate of about 9% from 2015-19.

The legacy of the higher operational cost of packaged treatment plants over traditional activated sludge

and trickling filter plants has historically posed a challenge to the growth of the market, but new

technological approaches have made significant efficiency gains in recent years. When whole life

equipment costs are considered, there are multiple areas in the delivery of these systems that can be

honed. Firstly, they require electricity - for air blowers, water pumps and drives - so improved efficiency




66

throughout the process train can reduce energy footprint and cost. The configuration of the treatment

process and hydraulics, including the media, have historically been underexplored and are ripe for

optmisation. Finally, the industry is learning quickly that offsite construction, faster installation and easier

maintenance also provide significant scope for enhanced competitiveness.

The two main segments in the packaged wastewater treatment market are municipal and industrial.

Technavio’s analysis shows the global municipal segment is expected to grow steadily from 2015-19,

reaching a revenue of around US$13 billion by the end of 2019. Uptake of packaged treatment plants by

UK utilities is increasing as they seek to meet tightening Environmental discharge consents on constrained

sites.

Available technologies

UK packaged treatment plant manufacturers offer a complete range of biological treatment technologies

including rotating biological contactors (RBCs), sequential batch reactors (SBRs) membrane bioreactors

(MBRs) and variations of submerged aerated filter (SAF) technology. These larger above ground packages

are generally designed around one section of the treatment process and cover inlet screens, grit removal,

primary settlement, sludge storage, biological treatment, final settlement and clarification.

The package can be supplied as a full turnkey manufactured product and delivered as a single unit for

smaller population numbers or as multiple units connected in series making up a complete treatment works.

Packaged treatment plants are built offsite and delivered on the day they are required. In terms of efficiency,

the smaller the footprint the lower the cost of installation, both in terms of backfill materials (generally

concrete) and transportation. Offsite build offers the manufacturer a number of benefits:

• Quality of manufacture in a controlled environment

• Consistency in quality

• The ability to test the integrity of products produced

• Just-in-time manufacture

• More complex design, allowing improved hydrodynamics

• Reduced health and safety risk when compared to onsite construction

• Higher tolerance build minimises footprint without compromising process

SAF technology

WPL operates globally and has been manufacturing packaged treatment plants for over 25 years at its

facility in the UK town of Waterlooville, Hampshire. The packages have been based on submerged air

filtration (SAF) technology and incorporate primary settlement, biological treatment - in the propriety SAF

- and final clarification. SAFs are well established in the treatment of municipal and industrial wastewaters

and are often used as an alternative to conventional biologically aerated filters (BAFs) (Pedersen et al.,

2015, Priya and Ligy 2015, Moore et al., 2001; Moore et al., 1999). SAF reactors are widely used for

secondary and tertiary treatment applications with configurations including upflow, downflow or a

combination of the two when internal recirculation is used (Khoshfetrat et al., 2011; Gálvez et al., 2003).

SAF reactors are typically arranged in a series of cells, with internal dividing baffles to increase treatment

efficiency and encourage segregation of heterotrophic and nitrifying processes (Hu et al., 2011). Biofilms

are grown on a plastic support media, which is submerged in wastewater and can be either, structured or

random packed. Media specific surface areas range from 150 to 1,200 m2/m-3 with SAFs most commonly

using 150 to 600 m2/m-3 (Khoshfetrat et al., 2011; Hu et al., 2011). Aeration of submerged media is provided

by coarse or fine bubble diffusers. Coarse bubble diffusers are often selected over fine due to the shearing

of air bubbles when in contact with the media, which generates a larger mass transfer surface area, similar

to that of fine bubbles (Rusten, 1984; Hodkinson et al., 1998). In SAFs with static media, contaminant

diffusion into the biofilm occurs from a combination of forward fluid velocities and circulatory mechanisms

influenced by aeration.

Wang et al. (2005) investigated the influence of support media fill ratio in suspended carrier biofilm reactors.

A media fill ratio of 50% was found to be optimal for the removal of chemical oxygen demand (COD) with

efficiencies of 73%, whilst a >70% media fill ratio was crucial to achieve 52% ammonium (NH4+) removal.

Rusten et al. (2006) investigated the effect of media fill ratio on treatment efficiency in moving bed biological

reactors (MBBRs), concluding that <70% was the optimum media fill ratio for NH4+ removal. Similarly

Pedersen et al. (2015), compared the removal of NH4+ in fixed and moving media bed reactors, concluding



67

that static media beds offered significant advantage over moving beds, providing protection of the nitrifying

bacteria in the biofilm. The WPL bioreactor moves very slowly in comparison to MBBR technologies and,

due in part to its shape, offers protection to the nitrifying bacteria, whilst giving the benefits of a moving bed.

Internal fluid velocities within SAF reactors have a direct influence over process performance, with high

velocities increasing biofilm detachment and washout potential (Burrows et al., 1999). In contrast, low

velocities increase solids retention, causing media blinding and reduction in mass transfer efficiency (Priya

and Ligy, 2015).

Airlift reactors use aeration and a fraction of flow is constantly recirculated to maintain a stable biofilm

thickness. Airlift reactors have two zones, an inner aeration column and an outer draft space. The inner

aeration column is raised a nominal distance from the bottom of the reactor. When aerated, a density

difference occurs between inner aeration column and draft space, causing fluid movement through the draft

space and into the aeration column (Kilonzo et al., 2006). Therefore changes in draft space baffle position

can be used to control the recirculation flow rate (Benthum et al., 2000). In fixed-film bioreactor engineering,

directional fluid movement is typically achieved through strategic aeration as described by Rusten et al.,

(2006).

At high volumetric loadings, SAFs suffer from flow shortcutting, therefore it is important to study the

hydrodynamic properties of the reactors in order to optimise flow regimes. Biofilm diffusion and dispersion

characteristics are influenced by SAF hydrodynamics, which is defined by three parameters:

• Mass transport efficiency (dispersion/advection)

• Mass transfer efficiency (aeration)

• Diffusive mass transfer (bulk flow to biofilm)

Many studies have been performed using estimations of Peclet number (Pe), axial dispersion coefficients

(D), dispersion number (d) and Reynolds number (Re) to describe mass transport and transfer dynamics

in biofilm reactors (Stevens et al., 1986; Romero et al., 2011; Lamia et al., 2012). Application of these

methods would offer insight into how fluid moves and mass transport dynamics of SAF reactors (Rexwinkel

et al., 1997).

In many applications, upflow and airlift SAF reactors are preferred due to improved fluid circulation, however

the hydrodynamics are not yet fully understood (Pedersen et al., 2015). Most research on SAFs focuses

on process efficiency and particularly variations in the influent wastewater volumetric and aeration loadings,

without considering the internal hydrodynamics (Gálvez et al., 2003; Albuquerque et al., 2012; Bibo et al.,

2011). The impact of media fill ratio, internal recirculation, internal baffles, aeration rates and biofilm growth

on the media is often excluded from SAF design practices.

The importance of hydrodynamic conditions in SAF reactors is investigated in a limited number of studies

to date. Fluid dynamics and efficiencies impact significantly on design and ultimately site and carbon

footprint. Refining the hydrodynamics can deliver multiple knock-on benefits to a utility or industrial user as

the plant size required to service a given population is smaller and requires less space on site. Reducing

site footprint also has a direct impact on the ability to produce compact SAF units offsite - it makes them

easier to transport, install and commission. This is especially important where access for large vehicles is

restricted. A reduction in the operating power required reduces carbon footprint and energy costs. It also

speeds up commissioning.

Main content

WPL SAF treatment plants, sold under the WPL HiPAF® brand, provide easily-installed below-ground

tanks. They are well established for use in both small commercial and water utility applications across the

UK. During 2011, demand for above-ground SAF technology in easily transportable units increased. This

was driven by the requirement for hire equipment and led WPL to develop modular above-ground

equipment that was easier and significantly lower cost to install than traditional in-ground civil construction.

This led WPL to develop hire equipment supplied in modules treating up to approximately 1,000 PE. These

modules can be used in various applications including treatment provision during temporary shut-down for

maintenance or to ensure a site meets its discharge consent where it is at risk of a breach. The WPL SAF

is manufactured in steel tanks for ease of transportation, offloading and set-up. This was an easy step for

WPL as it required upscaling of the biological zone to treat larger population loadings. The result was a

WPL SAF installed in a steel above-ground box using the existing HiPAF® design.



68

Figure 1: Example of a below ground WPL HiPAF® packaged multi-unit wastewater treatment

plant.

Figure 2: A 10,000 PE steel WPL SAF modular tertiary wastewater treatment plant.

More recently, in a drive to reduce power consumption, installation costs and footprint, WPL has been

improving SAF efficiency by refining the design of the process hydraulics. One aspect of this work has

involved reviewing the size and parameters of the plastic media used and investigating its density and

efficiency at different fill rates. During 2015-16 a substantial amount of work was done to calculate the

efficiency of the SAF bioreactor design used by WPL. It was found that the flow characteristics did not

follow the expected plug flow design. Instead, the reverse was seen, with the baffle walls not just allowing

removal of diffuser legs, but playing an intrinsic part in the reactor cell hydraulics. This understanding led

to a comprehensive review of the components used in WPL SAF technology. The first element under review

was the media used within the biozone and the investigation included its surface area, construction material

and density.

Initial lab tests revealed that the existing product had a high specific gravity, leading to it sinking to the

bottom of the tank. Its tubular shape meant the media soon filled with biomass and sludge, increasing its

weight to the point that a significant amount of air was required to move and therefore scour sludge and

dead biomass. The shape of the product did not lend itself to even flow distribution and dead spots were

seen in the tank, potentially leading to septicity. The surface area of 150m2/m3 was at its most effective



69

when exposed to high loading rates, however the inherent drawback was the limited potential surface area

available for treatment.

A new design was sought to replace this product and a buoyant open-structured media identified, with a

larger surface area, not exceeding 310m2/m3. In laboratory trials this media was found to be very buoyant

and again required significant aeration to get movement of media and subsequent scouring. Due to the

buoyant nature of the media, when used in full-scale installations, care had to be taken to ensure the media

was retained in the tank. Having used this media for a number of years in both full-scale and hire equipment,

further works were undertaken to look at efficiency of treatment, air requirements and optimisation. After

preliminary laboratory trials working on varying densities of media, an optimum specific gravity was

selected. This proved to have all the benefits WPL were looking for to complement the unique hydraulic

design of the HiPAF®.

Having been the focus of many years of research, mainly on the internal flows across the media, media fill

rates and energy requirements, the WPL SAF is now close to its final iteration. Branded as the WPL

SmartCell™, it is now ready for commercialisation, with the caveat that research and development will

continue over the next three years.

The SmartCell™ retains additional traditional primary settlement and final clarification either as an integral

tank or as individual treatment cells, with an enhanced biozone. It is this element that is critical to the

treatment process. It can be used in new build, retrofit and refurbishment applications. The modular design

optimises installation and reduces health and safety risk and maintenance requirements.

A number of sites have now been designed to incorporate downflow pipes rather than the traditional baffle

walls. This offers a number of benefits, but primarily allows flexibility in how flows are dispersed over the

media, optimising treatment and further reducing the size of the required biozone by up to 20%. The

downpipes can also be strategically placed to ensure the tank has no dead spots and, rather than using

scour air to prevent the accumulation of dead biomass, the process air undertakes this task. Whilst during

the day-to-day running of the plant, scour is only used intermittently, it does mean that a smaller plant using

one blower can be sized for process only rather than scour reducing the air requirements.

Case study 1: Hydraulic reconfiguration deployed at two sites

WPL has installed enhanced models of its WPL Hybrid-SAF™ wastewater treatment plants at two utility

sites – Site 1 in Cheshire and Site 2 in Staffordshire. Some 12 Hybrid-SAF™ modules were installed in four

treatment streams at Site 1 wastewater treatment works (WwTW) and six modules are now providing two

additional treatment streams at Site 2 WwTW.

Each of the two site installations was delivered and offloaded over two days and was set up and

commissioned within five days – Site 2 in October and Site 1 in early November 2015 and the plants are

likely to be needed at these sites for approximately five years. However, they can be transported from site-

to-site as required in the future. The utility was very clear about the number of streams required and the

need for the SAF units to be transportable. The utility did not want to be left with stranded assets if the

requirements of the site changed after five years and saw the flexibility of the Hybrid-SAF™ as a key

advantage.

WPL was asked to develop the transportable systems as the existing single-stage rock tricking filter plants

were experiencing ammonia compliance incidents during low flow and colder conditions. The new plants

are required to achieve an effluent quality of 4mg/l ammonia and can be switched off when warmer weather

returns – saving energy.



70

Figure 3: Packaged treatment plants can be easily lifted onsite.

In dry weather and colder weather the levels of ammonia in untreated effluent can rise considerably. The

previous winter the utility had had to hire additional packaged plants from WPL at short notice when the

plant at Site 2 struggled to cope with the higher levels of ammonia. Rather than having to hire equipment it

was decided to upgrade Site 1 and Site 2 by adding SAF units. The SAF systems have been individually

sized to treat the dry weather flow (DWF) for each site: 2,050m3/d for Site 2 and 4,000 m3/d for Site 1. Each

stream treats a nominal 1,000m3 wastewater and the additional treatment is expected to be required for

three-to-five months a year.

Site 1 and Site 2 installations were both designed to optimise flow dispersal in the biozone by incorporating

downflow pipes rather than traditional baffle walls. The pipes were carefully located to ensure the tank has

no dead spots which meant that the size of the biozone in each unit was reduced by up to 20%.

The client specified variable speed drives (VSDs) on the blowers to optimise energy efficiency. The use of

VSDs can achieve an energy saving of up to 30% over standard drives. Design developments focusing on

health and safety issues have removed the requirements for high-level access for routine maintenance.

Ground-level sensors and fixed diffusers Hybrid-SAF™ units at Site 1 and Site 2 have removed the need

for high-level inspection.

Figure 4: A 5,000 PE WPL Hybrid-SAFTM modular wastewater treatment plant installed at

Site 1 in 2017.

Case study 2: Media replacement halves air requirement

WPL has undertaken a number of small refurbishments on plants that were struggling to maintain

performance within the required Environment Agency discharge consent. One of these sites was a utility

100 PE plant consistently close to the consent requirements.



71

The plant was regularly impacted by storm water ingress and recovery was slow. This was due in part to

peak-loading occurring in the mornings as high flow levels arrived from a school. The challenge was to

upgrade the plant within the confines of the existing site and regulate effluent quality with sufficient margin

to keep within consents during periods of slow recovery.

As the SAF had a predetermined volume to house the media, a higher surface area product of 310m2/m3

was used throughout the unit, replacing the existing 220m2/m3 media. However a freeboard of 10% was

provided in the biozone to enable the slow but free movement of the media. The existing aeration had been

installed some 15 years previously and used coarse bubble diffusers. The decision was taken to retain it,

which allowed a straightforward replacement of the media in this instance without upgrading the baffle

configuration. The media replacement meant that air requirement to achieve the same PE was halved,

meaning that the plant could now process 189 PE – nearly double the flow and load.

Other efficiencies were leveraged in terms of project delivery. Removal of the legacy media was undertaken

with a vacuum tanker rather than manually, reducing time on site and the risk associated with entering a

confined space. As the SAF tank was over 15 years old, minor alterations and improvements were

undertaken. All aeration pipes were inspected and those showing signs of wear were removed and

replaced. These works accounted for only a further day onsite.

Media was delivered to site in 2m3 bags that could be moved and positioned next to the tank using a vehicle

with a small hydraulic crane attached. It was then manually loaded into the SAF tank, retaining grid replaced

and flow introduced.

Case study: Graph showing ammonia reduction from sampling positions in five different zones

within the SAF tank, from commissioning of the 189 PE plant. The samples were taken over six

weeks and full nitrification was achieved relatively quickly.

Whilst the project was a straightforward switch in media, different techniques in media removal, site

organisation and project management enabled a significant reduction in time on site. As a consequence,

significant cost savings were made. The alternative to this enhancement of the process would have been

a much bigger capital project, involving an extension or replacement. Refurbishment of the existing

infrastructure reduced capital cost, health and safety issues of major construction works, and has reduced

tanker movements reducing the environmental impact.



72

Figure 5: A 15 year old WPL HiPAF® wastewater treatment plant – the refurbishment nearly

doubled the flow and load from 100 to 189 PE.

Conclusion

This paper deals primarily with the biological phase of treatment, offering small footprint, high-efficiency

above-ground Hybrid-SAF™ modules that can be installed as temporary works or as permanent

installations. The advantages are significant when comparing with traditional onsite build. The package is

a fully pre-assembled unit ready to be placed directly onto a concrete base.

Figure 6: The flexible packaged WPL Hybrid-SAF™ units can be installed in series to form a

multi-unit wastewater treatment plant.

Integral packaged treatment plants manufactured off-site offer easy-to-install units with a reduced site

footprint. These systems are mainly viable when using high-rate filters and where the population equivalent

measure does not exceed 1,000. The advantages are significant when compared with the traditional onsite-

build of trickling filters or concrete tanks used as activated sludge plants.

Constructing a works based on packaged treatment plant reduces onsite installation time and mitigates

issues with prolonged confined entry, reducing health and safety risk. The primary drawback with packaged

plant is the size of tanks that can be produced and shipped to site, limiting the population equivalent that

can be served.

A small number of UK manufacturers offer modular off-site manufactured tanks. The units can be used

temporarily to allow remedial works to be undertaken on existing assets or permanently installed. Due to

the modular nature of the SAF units, they can be expanded or reduced to meet fluctuations in demand.



73

Modular tanks can handle higher population equivalents than single packaged plants and can offer

economical secondary treatment for up to 5,000 PE and economical tertiary treatment for up to 10,000 PE.

A number of sites in the UK have been retrofitted with improved media or media has been added to activated

sludge tanks to provide a fixed-film process. These have had varying levels of success and have generally

used media with a significantly higher surface area than 310m2/m3, with some manufacturers claiming

>1,000m2/m3. Generally speaking they have been converted or adapted to work as an MBBR. However

MBBR technology relies on a lower fill rate, around 50%, and has significantly higher turbulence than the

WPL SAF, causing biomass to be mechanically scoured as the media collides, thus theoretically negating

the outside surfaces of potential for fixed-film growth.

WPL is one of the first companies to closely examine the internal hydrodynamics of SAFs and the research

has proven that the removal of organic contaminants can be improved by reducing the ratio of media to

wastewater inside the tank, eliminating dead zones, reducing the need for scouring and cutting energy use.

By reconfiguring the baffles to fully-segment the biozone, the risk of process shortcuts and dead zones has

been eliminated. Each biozone segment, where both carbonaceous and nitrifying processes take place, is

filled with high-voidage plastic filter media. Air to oxygenate the influent and scour excess biomass from the

media is introduced continuously below each chamber by a series of diffusers. Optimisation of the specific

gravity of the media ensures it now circulates as efficiently as possible. Refining multiple aspects of the

process has succeeded in delivering a variety of benefits and mitigating process risks associated with

variable loads.

Research and development into SAF technology undertaken by WPL has culminated in the WPL

SmartCellTM, which is a compact module with counter-current hydraulics to ensure constant flow over

optimised media, avoiding the requirement for reduced media fill rates. With a 90% fill rate, the SmartCelllTM

has sufficient open-voidage to avoid fouling or becoming sludge-bound when applied in secondary

treatment. The design also takes into account the impact of physical colliding or scouring of the media, and

protects the growth of the biomass due to its unique shape and speed of movement. With the slow

movement of the media, any build-up of sludge is carried to the surface where it washes out with other

solids to be captured in final settlement.

Research into SAF efficiency is on-going and WPL is supporting a PhD at Portsmouth University over the

next three years (2017-2020) to review the efficiencies achieved in more detail and to optimise energy use.

The study will examine the data and technical information already generated and hone it into a design

protocol for all WPL SAF systems. Consideration is given to the following elements, in relation to the design

for the SmartCelllTM.

1. Module size

The size of the modules should give optimum performance, balanced by the need to transport, manufacture

and handle them safely. Size of the individual modules is therefore governed by the following constraints:

• Maximum permissible width without it being classed as an abnormal load notifiable to the highways

authorities. This aspect of sizing is critical as access at most rural wastewater treatment sites is

designed to the maximum width, which is required for desludging tankers.

• Cube-shaped modules were deemed the most efficient for maximising transportable tank volume,

even though some oxygen transfer efficiency can be gained using taller tanks. Height can limit

road transportation options, so using taller tanks delivered to site horizontally was dismissed in the

standard module design, but can be considered for a design-and-build project. The external tank

envelope was sized to maximise media capacity but minimise transport risk.

2. Inlet and outlet pipework

In instances where a pumped flow is to be used, investigation into incoming flow velocities will be carried

out and the system and optimised. The same goes for the exact entry points of flow within the bed – at

present some short circuiting on the first cell occurs. Gravity-fed systems suffer less from the first cell being

hydraulically bypassed, but the study will optimise inflows. The outlet pipework appears to offer no process

benefits, but this will be investigated using CFD modelling in a PhD research study.

3. Gridding

Both top and bottom gridding is used to encapsulate the media and prevent it from discharging into the

environment. This product is expensive and its loss reduces the operational volume available. Due to its

structure, the media also creates back-pressure within the reactor. The nature of this back-pressure is



74

largely understood, but an investigation is required to fully evaluate the process benefits or drawbacks of

using this type of encapsulation.

4. Downpipes

Currently the downpipes are spaced evenly throughout the tank on the assumption that lateral flows will

be even and so will mixing. No consideration has been fully given to the effect of backflow during low

diurnal peaks or forward flow during peak flow. This element will be critical to product development as it

will allow WPL to optimise the cells for enhanced secondary treatment, BOD removal and to predict

where within the process nitrification will stabilise. Due to the significant variances in flow seen in

packaged treatment plants this is seen as an area that may be difficult to design.

5. Diffusers

WPL currently uses coarse bubble diffusers as these were the subject of a significant investigation into

oxygen transfer rates using fine, medium and coarse bubble diffusers. The study formed part of a previous

PhD, An Evaluation Of The Robust Aerobic Digestion System by Nick Sherlock (2009). This study looked

into oxygen transfer rates using fine, medium and coarse bubble diffusers, which found coarse bubble

diffusers gave the best performance with the added benefit of reduced fouling and maintenance. The

ongoing work will look at maintenance and long-term installation when using a fixed diffuser leg versus a

removable one.

6. Media

As previously described, the neutrally buoyant media has been optimised for buoyancy, biomass thickness

and type but has not been fully investigated. Its performance in an anoxic environment will also be reviewed

and considered.

It is hoped that data gathered from the PhD research will enable full optimisation of the existing plant,

allowing greater treatment efficiencies per cubic meter of media and enhanced oxygen transfer through

positioning of the downpipes. It is also hoped that clearly identified and stabilised zones will be created

allowing both nitrification and denitrification to occur in the same SmartCelllTM. If bio-stabilisation can be

achieved, a review of phosphorus (P) removal will be undertaken, subject to time constraints. The SAF

process has come a long way, but there is still further to go in terms of process efficiency and power

consumption.

References

Benthum, W. A. J., Lans, R. G. J. M., Loosedrecht, M. C. M. and Heijnen, J. J. (2000), "The biofilm airlift

suspension extension reactor - II: Three phase hydrodynamics", Chemical Engineering Science, vol. 55,

pp. 699-711.

Bibo, H., Wheatly, A., Ishtchenko, V. and Huddersman, K. (2011), "Effect of shock loads on SAF bioreactors

for sewage treatment works", Chemical Engineering Journal, vol. 166, pp. 73-80.

Burrows, L. J., Stokes, A. J., West, C. F., Forster, C. F. and Martin, A. D. (1999), "Evaluation of different

analytical methods for tracer studies in aeration lanes of activated sludge plants", Water Research, vol. 33,

pp. 367-374.

Hu, B., Wheatley, A., Ishtchenko, V. and Huddersman, K. (2011), "The effect of shock loads on SAF

bioreactors for sewage treatment works", Chemical Engineering Journal, vol. 166, pp. 73-80.

Khoshfetrat, A. B., Nikakhtari, H., Sadeghifar, M. and Khatibi, M. S. (2011), "Influence of organic loading

and aeration rates on performance of a lab-scale upflow aerated submerged fixed-film bioreactor", Process

Safety and Environmental Protection, vol. 89, pp. 193-197.

Kilonzo, P. M., Margaritis, A., Bergougnou, M. A., Jun, T. Y. and Qin, Y. (2006), "Influence of baffle

clearance design on hydrodynamics of a two riser rectangular airlift reactor with inverse internal loop and

expanded gas-liquid separator", Chemical Engineering Journal, vol. 121, pp. 17-26.

Moore, R. E., Quarmby, J. and Stephenson, T. (1999), "BAF media: Ideal properties and their

measurement", Institute of Chemical Engineers, vol. 77, pp. 291-297.

Pedersen, L. F., Oosterveld, R. and Pedersen, P. B. (2015), "Nitrification performance and robustness of

fixed and moving bed biofilters having identical carrier elements", Aquacultural Engineering, vol. 65, pp. 37-

45.



75

Priya, V. S. and Ligy, P. (2015), "Treatment of volatile organic compounds in pharmaceutical wastewater

submerged aerated biological filter", Chemical Engineering Journal, vol. 266, pp. 309-319.

Rexwinkel, G., Heesink, A. B. M. and Swaaji, V. (1997), "Mass transfer in packed beds at low Peclet

numbers - wrong experiments or wrong interpretations", Chemical Engineering Science, vol. 52, pp. 3995-

4003.

Rusten, B. (1984), "Wastewater treatment with submerged biological filters", Water Pollutant Control

Federation, vol. 56, pp. 424-431.

Wang, R. C., Wen, X. H. and Qian, Y. (2005), "Influence of carrier concentration on the performance and

microbial characteristics of a suspended carrier biofilm reactor", Process Biochemistry, vol. 40, pp. 2992-

3001.

Sherlock N. "An Evaluation Of The Robust Aerobic Digestion System" 2009.

Watkins S. "Physico-Chemical And Microbial Factors Affecting The Operation Of a Package Waste Water

Treatment Plant" 2011.

Holloway T.G. "Influence Of Aeration And Media Fill Ration On Submerged Aerated Filter Hydrodynamics

And Process Performance For Municipal Wastewater Treatment" 2016

LIVERPOOL WWTW SBR CARBONACEOUS TRIAL

Akinola, O., Black, J., Sherwood, A. and Hornsby J., United Utilities, UK


Abstract

In January 2017, a Trial commenced on the Sequencing Batch Reactor (SBR) Plant at Liverpool WwTW.

The 16-basin SBR Plant is normally operated in ‘nitrification mode’. Annual aeration savings of

approximately £300k were previously estimated if carbonaceous operation is adopted. The Trial was carried

out on 2 No. basins with a target aerobic sludge age of 4 days. The key objectives of the Trial were to

assess process risk, OPEX benefits, and the likely impact on the sludge stream. Process modelling using

BioWin was carried out in advance to provide guidance on key operating parameters. Liaison with site

operators and managers was necessary throughout the Trial to implement changes and monitor

performance. Key Performance Indicators (KPIs) included: effluent quality (particularly ‘hard’ COD), sludge

age, settleability, microscopy and energy consumption. The Trial results demonstrated effluent quality

consistently below consented limits and 37% aeration energy reduction.

Keywords

Basin, Carbonaceous, Configuration, Energy, MLSS, Operation, SAS, SBR

Introduction

The 16-basin SBR Plant at Liverpool WwTW was designed to operate in ‘nitrification mode’, due to historical

issues with ‘hard COD’ received at the Works and the need to ensure adequate treatability. There is

currently no ammonia consent at Liverpool WwTW. It has been previously estimated that by adopting

carbonaceous operation (i.e. not removing ammonia), annual aeration savings of approximately £300k

could be achieved.

The process of transitioning from ‘nitrification mode’ to ‘carbonaceous mode’ requires some reconfiguration

of the process and could entail some process compliance risk. Therefore, it was necessary to undertake a

Trial to further understand the risks, limitations and reconfiguration needed.




76

Background

Trial Objectives

The objectives of the Trial were as follows:

• To quantify the process risk of carbonaceous operation (in particular, BOD & COD compliance,

sludge settleability, microbiology, odour)

• To understand how best to manage the transition from ‘nitrification mode’ to ‘carbonaceous mode’

• To understand any reconfiguration required on the SBR and sludge treatment processes to

enable full carbonaceous operation of the SBR

• To understand the optimum operating conditions for carbonaceous operation

• To quantify the OPEX savings and costs of operating a carbonaceous SBR

Site Details and Permit

Treatment Process

Treatment at the Works consists of: preliminary treatment followed by primary sedimentation, and

secondary treatment in the SBR Plant. Furthermore, there are storm tanks present which provide storage

during periods of high flows. Primary and secondary sludge thickening are carried out separately, followed

by combined digestion on site.

Permit

The Permit for Liverpool WwTW includes consents for the following:

• Flow to Full Treatment (FTFT) ≈ 346 Ml/d

• BOD: 25mg/l (95%ile) / 50mg/l (UTL) / 70% Removal [Urban Wastewater Treatment Directive

(UWWTD)]

• COD:125mg/l (95%ile) / 250mg/l (UTL) / 75% Removal (UWWTD)

• Suspended Solids: 250mg/l UTL

• Several metals including iron and aluminium; and organic compounds such as chloroform and

trichloroethene

SBR Operation

The 16No. SBR basins are continuously filled and are operated according to a 4 hour cycle as shown in

Table 6. This cycle is repeated throughout the day for each pair of basins.

Table 6: Typical SBR Cycle

Time (Hours) 0 – 1 1 – 2 2 – 3 3 – 4

Phase Fill-Aerate Fill-Aerate Fill-Settle Fill-Decant

During each cycle, Surplus Activated Sludge (SAS) can be removed from the basins during the ‘aerate’

(hour 1) or ‘decant’ (hour 4) phase.

Methodology

Prior to start of the Trial, process calculations and modelling1 using BioWin 5.0 software, were carried out.

The Carbonaceous Trial was then carried out on 2No. basins (namely Basins 5 and 6), operating as a

hydraulically-linked pair, with basin 1 as the Control. The target aerobic sludge age was 4 days, and this

equated to a Mixed Liquor Suspended Solids (MLSS) target range of 1,500 to 2,000mg/l.

Throughout the Trial, monitoring of the Trial and Control basins was undertaken.

_________________________________________________________________________________ 1 Discussed in a separate section

Configuration of Trial and Control Basins

In order to achieve (and subsequently maintain) the required aerated sludge age of 4 days and MLSS target

of 1,500 to 2,000mg/l, it was necessary to reduce the MLSS in the Trial basins by increasing the amount of

SAS solids being wasted during each cycle.

Initially, the Trial and Control basins were set according to the configuration in Table 7 below:



77

Table 7: Initial Basin Configuration

Basin Configuration

5 SAS during ‘aerate’ phase

6 SAS during ‘decant’ phase

1 SAS during ‘aerate’ phase

However, following a number of mechanical issues with the SAS Drum Thickeners approximately seven

weeks into the Trial, surplussing for all the remaining basins was changed to occur during the ‘decant’

phase. The following basin configuration (Table 8) was thereby adopted for the remainder of the Trial. It is

worthy of note that these mechanical issues were unrelated to the Trial.

Table 8: Final Basin Configuration

Basin Configuration




A key advantage of surplussing solids from the basins during the ‘decant’ phase is significant reduction in

SAS volumes. Therefore, this would result in overall less pressure on the Secondary Drum Thickeners’

capacity and contribute to more resilient operation of the Sludge Thickening Plant.

Process Set-points

Throughout the Trial, the following process set-points were adjusted and recorded for the Trial basins:

• volume of SAS removed (m3 per cycle)

• duration of SAS event (minutes)

Monitoring

As part of the Trial, qualitative and quantitative monitoring was undertaken.

Qualitative monitoring comprised of daily visual inspections of the Trial and Control basins, weather and

weekly odour readings. Quantitative monitoring comprised of daily/weekly collection and analyses of

effluent samples, MLSS and SAS samples, and weekly downloads of relevant site data. Determinands of

particular interest included Ammonia, Biochemical Oxygen Demand (BOD), Filtered BOD (BODF),

Chemical Oxygen Demand (COD), Filtered COD (CODF), Nitrate (NITR), pH, Total Suspended Solids

(TSS), and % Dry Solids (DS). Furthermore, weekly microscopy was carried out for the MLSS and/or SAS

samples.

Preliminary BioWin modelling and Outputs

In order to predict potential impacts of operating in carbonaceous mode, BioWin 5.0 modelling software

was used. One of the 16 No. basins was modelled to assess typical performance of a single SBR basin.

The models were run in nitrifying, transition and carbonaceous modes at 11°C (i.e. worst-case

temperature). Furthermore, several iterations were performed by varying a number of factors including

duration of model run, Dissolved Oxygen (DO) concentration, alpha factor, SAS pump flow rate and SAS

volume removed per cycle.

Figure 1 is a layout of one of the model iterations.



78

Figure 1: BioWin Model Layout

The input data for the models was SBR feed data from Performance Tests undertaken in June 2016 as this

provided a robust data set. The duration of the model runs was initially 2 weeks, then this was increased to

4 weeks to observe the MLSS trend over a longer period of time. More detailed analysis of BioWin showed

that it would typically take approximately 40 days for the MLSS in an SBR to stabilize. The duration of the

model runs was therefore increased to 8 weeks in order to assess the MLSS, SAS concentration and

effluent quality trends over this ‘stabilization’ period.

The following sub-sections summarise the outputs of the modelling exercise.

Transition from Nitrification to Carbonation Mode

MLSS and SAS

The transition models suggested that it would take up to two weeks to reach the target MLSS range of

1,500 to 2,000mg/l. The differences in ‘SAS aerate’ and ‘SAS decant’ configurations were also highlighted.

The output SAS concentration from the modelled ‘decant’ configuration was much higher (about 4 times

greater) than the ‘aerate’ configuration; thereby, enabling a significantly reduced SAS volume to be

removed from the basin (approximately one-quarter of that in ‘aerate’).

SBR Effluent Quality

Overall, in transitioning from nitrifying to carbonaceous mode, the following trends were observed from

the 95%-ile effluent quality results. These can also be seen in Figure 2.

.

• Ammonia increased

• BOD change was negligible (slight reduction)

• COD reduced

• Solids reduced



79

Figure 2: Effluent Quality – Nitrifying vs. Carbonaceous mode (BioWin)

Blower Energy Trends

The BioWin models showed that switching the SBR basins to carbonaceous mode (without making any

changes to current DO set-points2) would result in 40 to 50% reduction in blower energy consumption.

Impact of Dissolved Oxygen Control in Carbonaceous Mode

Blower Energy Trends

By halving the DO set-points that the basins currently operate to, the models also suggested potential

additional3 aeration energy savings of approximately 25%. There is an opportunity for this to be explored

on site during the imminent blower optimisation work.

Effect of Reduced Aeration on Microbiology

Results from the BioWin models showed that nitrification was still occurring even at significantly reduced

basin MLSS concentrations in carbonaceous mode.

While modelling in carbonaceous mode, the impact of reducing the DO set-points on the nitrifying

microbiology was observed. The nitrifying microorganisms modelled are classified into three groups,

namely, Ammonia Oxidising Biomass (AOBs), Nitrite Oxidising Biomass (NOBs) and Anaerobic Ammonia

Oxidising Biomass (AAOs). AOBs and NOBs are aerobic organisms which oxidise ammonia to nitrite, and

nitrite to nitrate respectively; while AAOs are anaerobic organisms which oxidise ammonia using nitrite, to

nitrogen gas and nitrate.

Figures 3, 4 and 5 highlight the impacts of halving the DO set-points on these organisms.

From these graphs, it can be observed that in carbonaceous mode, steady populations of AOBs and AAOs

were maintained, while a steady reduction in NOBs is observed. Once the DO set-points were halved

however, the AOBs reduced steadily, the NOBs reduced more rapidly and the AAOs increased to almost

twice their previous concentration. Thus, the AAOs show a competitive advantage over the aerobic nitrifiers

(AOBs and NOBs) when DO levels are reduced. ______________________________________________________________________________________________________________________________________

2 See Appendix for current blower DO profile applied across all basins; 3 Additional savings estimate is a proportion of

carbonaceous energy consumption

Figure 3: Impact of Reduced Aeration on AOBs (Carbonaceous Mode)



80

Figure 4: Impact of Reduced Aeration on NOBs (Carbonaceous Mode)



81

Figure 5: Impact of Reduced Aeration on AAOs (Carbonaceous Mode)

Trial Results and Discussion

Flows

Figure 6: Average Flows into SBR Basins

Figure 6 shows the average flows which passed through the entire SBR Plant during the Trial. There were

some periods of heavy rainfall (average flow > 300Ml/d) and drier periods (average flow < 200Ml/d). During

the Trial, the 20%-ile and 99%-ile flows were 187Ml/d and 344Ml/d respectively. In 2016, the 20%-ile and

99%-ile flows recorded on site were 167Ml/d and 343Ml/d respectively. Therefore, the flows experienced

during the Trial were representative of the typical flow range on site.

MLSS

Figure 7 shows the trends in Mixed Liquor Suspended Solids (MLSS) throughout the duration of the Trial.

From this graph, it can be observed that there was a significant reduction of MLSS in both Trial basins as

the Trial progressed. Furthermore, within 1 month of the Trial, basin 6 reached the target MLSS range of

1,500 to 2,000mg/l (average – 1,750mg/l); however, throughout the Trial, basin 5 did not reach this target.

This may have been hindered by wider site issues.

For example, the MLSS concentrations at the beginning of the Trial were elevated (above 4,000mg/l) for

all three basins. These concentrations were above the operational target of 3,800mg/l. In addition, in early

March, an increase in MLSS concentrations was observed for both Trial basins. Both these instances of

solids increase were attributed to increased SBR feed loads, which ultimately exerted a strain on (and

contributed to mechanical issues with) the SAS Drum Thickeners, due to the need for higher SAS removal

rates across the entire SBR Plant. The mechanical issues with the SAS Thickeners resulted in a reduction

in throughput; and consequently, a backlog of sludge to be processed by the Thickeners. This eventually

led to a build-up of solids within the basins.

Furthermore, from 24th to 28th March, basin 6 was out of service due to a mechanical issue. This issue was

considered unrelated to the Trial. During this time, flow in and out of the basin was stopped, and it was

placed in continuous aeration. There was therefore no effluent or MLSS sample collected on those days.



82

Figure 7: MLSS Trends

Sludge Settlement

Throughout the Trial, both Trial basins showed very good settlement. For example, the Stirred Sludge

Volume Index (SSVI) was significantly less than 100ml/g, compared to the UU Asset Standard of 120ml/g.

In addition, the Control basin generally showed good settlement; however, there were some occasions

during which the SSVI was greater than 100ml/g. Figure 8 shows the SSVI samples at the end of a test.

Figure 8: SSVI Test

As settleability was not an issue during the Trial, a detailed standardised SSVI (SSVI 3.5) test was not

considered necessary for each sample. From the entire Trial dataset, approximate SSVI 3.5 figures for all

three basins were as follows: Basin 5 – 79ml/g, Basin 6 – 81ml/g and Basin 1 – 79ml/g. Therefore, SSVI

3.5 was consistent across the Trial and Control basins.

SAS

For the Liverpool WwTW SBR basins, SAS can be removed during the ‘aerate’ or ‘decant’ phases of the

cycle. Due to increased SBR feed loads and consequent mechanical issues with the SAS Drum Thickeners

in March, there was need for a reduction in the sludge volume to be processed by the Secondary Thickening

Plant. Therefore, Operations adjusted all the remaining SBR basins to surplus sludge during the ‘decant’

phase (including Trial basin 5).

Following the repair and recovery of the SAS Drum Thickeners, Operations decided to maintain the sludge

surplussing regime of the basins in ‘decant’ phase for the foreseeable future, as this puts less pressure on

the Drum Thickeners’ capacity; thereby contributing to more resilient operation of the Thickening Plant.



83

Table 4 highlights advantages and disadvantages of either method of SAS removal.

Table 4: ‘SAS Aerate’ vs. ‘SAS Decant’ Configuration – Advantages & Disadvantages

SAS Removal Phase Advantages Disadvantages

Aerate

Relatively automatic sludge age

control Greater sludge volumes (3 times ‘decant’ phase volume)

Less intensive operator intervention Secondary sludge processing capacity

exceeded – major additional capital

expenditure (CAPEX) likely

Higher operational expenditure

(OPEX) – i.e. SAS pumping and

sludge processing

Decant

Much lower sludge volumes

(up to one-third of ‘aerate’ phase

volume)

More intensive operator intervention

(MLSS control)

Within volumetric capacity of

secondary sludge assets – no/minimal

additional CAPEX

Potential adverse impacts on process

due to more unknowns e.g. mass of

basin solids

(no evidence of this during Trial)

Potential OPEX savings (i.e. reduced

SAS pumping and sludge processing)

Settleability risk

(no evidence of this during Trial)

The SAS volumes removed from the basin(s) with the ‘SAS decant’ configuration varied throughout the

Trial. This was mainly due to a number of factors; for example, difficulties in estimating mass of sludge

being withdrawn from the system at any given time (as this depends on sludge settleability), and changing

sludge concentrations at the base of the SBR during ‘settle’ and ‘decant’ phases. Therefore, close

monitoring of basin MLSS was essential for this configuration. The ‘optimum’ range for basin 6 SAS volume

in order to maintain the MLSS target, was found to be between 65 and 85m3 per cycle (in ‘SAS decant’

configuration). As a comparison, the maximum SAS volume withdrawn from basin 5 whilst it was in ‘SAS

aerate’ configuration, was 185m3 per cycle.

The SAS pump flow rate typically used on site is 70l/s. Triplicate sampling of SAS was carried out for both

Trial basins in order to observe the change in SAS concentration over each withdrawal event. At this rate

of 70l/s during the ‘decant’ phase, it was observed that the SAS concentration reduced significantly over

the event – on average from approximately 0.8% Dry Solids (DS) at the start through to 0.4% DS at the

end.

Therefore, in the last two weeks of the Trial, the SAS flow rate was reduced from 70 to 50l/s for both Trial

basins and the resulting impact on SAS concentrations noted. It is worthy of note that both basins 5 and 6

were surplussing sludge during the ‘decant’ phase at this point. Following this reduction in SAS flow rate,

the SAS concentrations were approximately 1% DS at the start, and 0.5 to 0.6% DS by the end of the event

for both basins. This therefore suggested that a reduction in SAS pump flow rate resulted in more consistent



84

SAS concentration throughout the SAS event, as there was a lower possibility of ‘rat-holing’ (i.e. drawing

off more dilute sludge).

Effluent Quality

Daily spot and 24-hour composite samples were generally collected for the Trial and Control basins.

However, in the last month of the Trial, the frequency of sampling was reduced for a number of

determinands, as sufficient information had already been gathered.

The following sub-sections discuss the composite effluent results only, as these are representative over a

24-hour period. Furthermore, the spot effluent results generally showed similar trends across all

determinands.

Ammonia

From the second week of the Trial, the ammonia concentrations in effluent from the Trial basins increased

significantly more than the Control. This was expected as increased amount of solids were being surplussed

from the Trial basins in order to reduce the MLSS concentrations. Consequently, the sludge age and

populations of nitrifying organisms in these basins were being reduced accordingly. Furthermore, in March,

there were two periods of increased effluent ammonia concentrations in the Trial and Control basins. These

corresponded with increased SBR feed loads.

As the Trial progressed, a reduction in ammonia load removed by basin 5 was observed, and by February,

an average load removal of approximately 300kg/d was estimated. In early March (i.e. during the period of

the sludge backlog on site and consequently, increased basin solids), a slight increase in ammonia load

removed (to approx. 420kg/d) was observed for both Trial basins. However, this reduced towards the end

of the Trial. There is therefore potential for further optimisation i.e. achieving further reductions in OPEX, if

ammonia load removal is reduced further.

Through the Trial, nitrification (i.e. conversion of ammonia to nitrate) reduced in basins 5 and 6. A

corresponding reduction in denitrification (i.e. conversion of nitrate to nitrogen gas) was also estimated for

these basins. This was expected because if lower levels of nitrate were being produced through nitrification,

there would be less nitrate available for conversion to nitrogen gas, through the process of denitrification.

On the converse, denitrification for basin 1 remained relatively stable, despite occasional fluctuations.

BOD

The effluent BOD concentrations for basin 5 were relatively steady for the entire duration of the Trial.

However, on 29th March, there was an atypical result of 25.8mg/l. This sample showed a 77% BOD removal

(from corresponding crude BOD concentration of 113mg/l); therefore, was still within the Permit conditions.

On this day, it rained and visible solids were noted in the effluent; hence, the exceedance may have been

linked to rain which would have contributed to increased flow through the basin, resulting in solids carryover.

The corresponding soluble BOD (BODF) concentration was significantly lower (8.6mg/l); and this suggests

that this total BOD exceedance may have been linked to solids carryover. However, given that the

corresponding COD concentration for that sample was 72mg/l (within typical range experienced on site), it

is possible that this high BOD concentration may have been a spurious result. Also, by the next day, the

effluent BOD concentration for basin 5 had reduced significantly to 15.8mg/l.

Furthermore, several BOD spikes in effluent from basin 1 (Control) were observed. These spikes also

corresponded with spikes in COD and Total Suspended Solids (TSS) concentrations. Overall in March, a

steady rise in effluent BOD concentration was observed for both basins 5 and 1. This rise corresponded

with increasing SBR feed loads.

Typically, for Trial basin 5, effluent soluble BOD (BODF) concentrations were significantly less than total

BOD concentrations (average of 3mg/l BODF vs. 10mg/l for total BOD). Similarly for basin 1 (Control),

typically, BODF concentrations are significantly less than total BOD concentrations (average of 3mg/l BODF

vs. 17mg/l for total BOD). Both basins 5 and 1 generally showed a similar BODF concentration range of 1

to 10mg/l; thereby, confirming BOD treatability at lower basin MLSS concentrations.

No distinct difference in BOD load removal was observed between basins 5 and 1; however, in some

instances and on average, results from basin 5 showed slightly higher removal.



85

COD

Initially, a slight reduction in COD concentrations was observed for Trial basin 5, followed by a period of

relatively steady results. However in March, there were two periods of increased effluent COD

concentrations for the Trial and Control basins. These corresponded with increased SBR feed loads

(potentially linked to several factors including increased Primary and Secondary Thickener filtrate loads and

crude loads coming into the Works). In addition, for majority of the Trial, one of the Primary Settlement

tanks was out of service, and this could have contributed to greater loads in the SBR feed. However, the

effluent COD concentrations for basin 5 were still within the UWWTD Consent of 125mg/l.

Overall, effluent COD results for the Trial basins were typically lower than the Control; and the Trial basins

did not show spikes in COD concentrations compared to the Control. This has contributed to alleviating

major concerns over treatability of (‘hard’) COD in carbonaceous mode.

Furthermore, both basins 5 and 1 showed a similar effluent soluble COD (CODF) concentration range of

approximately 20 to 80mg/l; thereby, confirming (‘hard’) COD treatability at lower basin MLSS

concentrations.

Basins 5 and 1 (Control) showed similar COD removal throughout the Trial. COD load removal increased

from early- till about mid-March, and this corresponded with increased basin MLSS following the sludge

backlog on site, as well as increased feed loads. This increased removal was expected as an increase in

MLSS would result in more microbiological organisms being available to provide more treatment, and

increased feed COD load meant more COD was available for removal. COD load removal then reduced,

before increasing again in the last week of the Trial. This latter increase was also attributed to increased

basin MLSS and feed loads around the same time.

Total Suspended Solids

Effluent TSS concentration from basin 5 was generally steadier than the Control (basin 1). This was

attributed to there being less solids in basin 5, as well as good settlement of solids during the ‘settle’ and

‘decant’ phases.

For the Control basin, spikes in BOD, COD and TSS concentrations sometimes occurred when the basin

MLSS was greater than 4,000mg/l, during high SBR feed loads, rainfall and when SSVI was slightly above

100ml/g. However, the Trial basins did not show such spikes. This has contributed to alleviating concerns

over ‘carbonaceous mode’ operation.

Summary: Basin 5 vs. 1

Table 9 summarises the effluent quality for basins 5 and 1, from the point at which basin 5 MLSS had

reduced to a relatively stable level (20th February) up until the end of the Trial.

Table 9: Effluent Quality Summary – Basin 5 vs. 1

Determinand

(mg/l)

95%-ile

Consent

(mg/l)

Basin 5 Basin 1

Average 95%-ile Risk ratio 4 Average 95%-ile Risk ratio 4

Ammonia - 32.4 48.8 N/A 16.4 32.8 N/A

BOD 25 11 23 2.2 17 59 1.4

COD 125 65 91 1.9 70 159 1.8

TSS 250 (UTL 5) 21 27 12.0 44 145 5.7

The risk ratios for basin 5 were significantly higher than those for basin 1 (Control), particularly for BOD and

TSS. Furthermore, the 95%-ile effluent results for basin 5 were all within the Permit limits; whereas, the



86

Control showed 95%-ile exceedance for the determinands listed above. Therefore, these results confirm

that process risk is significantly reduced at lower basin MLSS (i.e. in/approaching ‘carbonaceous mode’).

4 This is the Consent/Average ratio. Typical operational target for BOD and COD > 2 5 Upper Tier Limit (UTL) value i.e. absolute maximum

Blower Energy

After about three weeks into the Trial (from 3rd February 2017), a steady decline in blower energy for the

Trial basins was observed (Figure 9). From that point to 24th March, the reduction in total estimated blower

energy consumption was approximately 37%. Furthermore, from 3rd February, the difference between the

total energy6 for the Trial basins and that for basins 1 and 2 was approximately 1,000kWh on average.

Applying this across all 16No. basins would correspond to a potential annual aeration energy saving of

approximately £292,000.

From the first week of March, the blower energy consumption for the Trial and Control basins was seen to

increase. This was attributed to several factors including increased basin MLSS concentrations following

issues with the SAS Drum Thickeners (and resulting sludge backlog) and increased SBR feed loads. An

increase in temperature also occurred, and this resulted in increased nitrification (and therefore, increased

oxygen demand) as well as a reduction in oxygen transfer efficiency. The combined effect of all these

factors was increased blower power output in March.

Furthermore, during the last few weeks of the Trial, there were a few instances where discrepancies

between the DO concentrations of basins 5 and 6 were observed. Following temporary removal and

cleaning of the DO probes, these discrepancies seemed to be removed and the DO concentrations for both

basins returned to similar levels. These occurrences may likely have contributed to increased air demand

from the blowers during these times; and consequently, increased blower energy consumption.

Lastly, there is currently no air flow measurement in the pipes supplying air to the basins. This information

would have been useful to better understand the factors contributing to blower energy consumption. Going

forward as time progresses, by monitoring the air flow to the basins, it would enable site to attribute the

blower energy consumption to either blower or diffuser efficiency.

6 Between 28th and 30th March, a data gap in blower power data was observed. This was identified as a site-wide issue. Due

to several issues on site from 24th March, the blower energy consumption after this date has been omitted from calculations

and Figure 9.

Figure 9: Total Daily Blower Energy



87

Microscopy

Figure 10: Trial Basin - Example Image from Microscope (x40 Magnification)

Figure 10 is an example microscope image for basin 6. This shows good floc structure and representative

life forms in the basin.

As carbonaceous mode was achieved in both Trial basins, overall observations from microscopy were as

follows:

• Good floc formation and generally low filament density in both Trial and Control basins

• Trial basins: a greater proportion of lower life forms i.e. free-swimming flagellates, crawling

ciliates and some higher life forms i.e. stalked ciliates; thereby, indicative of a lower sludge age

• Control basin: Much lower proportions of lower life forms i.e. very few flagellates and much

greater proportions of higher life forms; thereby, indicating a higher sludge age

Odour

Throughout the Trial, no impacts on odour were observed.

Comparison of BioWin and Trial Results

Both the preliminary BioWin modelling and the Trial showed the following results for carbonaceous mode

operation:

• Effluent quality well within Permit limits (BOD, COD and TSS)

• Blower Energy reduction – 37% vs. 50% prediction by BioWin (Black, 2016)

• Potential additional aeration savings if DO set-points are halved (BioWin results)

• Continued nitrification even in ‘carbonaceous mode’

The BioWin modelling exercise suggested that it would take up to two weeks to reach the target MLSS

range. However, it took four weeks to achieve this target during the Trial, and only basin 6 reached the

target. This delay may have been contributed to by elevated starting basin MLSS concentrations (i.e.

>4,000mg/l versus the preferred site operating range of 3,000 to 3,600mg/l).

Furthermore, BioWin generally suggested that more treatment would be provided in carbonaceous mode

i.e. lower average effluent concentrations for Ammonia and BOD. However, this may be linked to

differences in sewage characteristics between North America and UK e.g. COD/BOD ratios; as well as

under-/over-exaggerated processes or assumptions within the BioWin software e.g. settling, mixing and

nitrification.

Conclusions

From the Trial, the following conclusions about carbonaceous mode operation were drawn:

1. Operation of the Liverpool WwTW SBR basins in carbonaceous mode is possible and is capable

of delivering consistently compliant effluent quality. Operational risk ratios (i.e. Consent/Average)

from the point of relatively stable MLSS in the Trial basins, were 2.2 and 12 for BOD and TSS

respectively for basin 5; whereas those for basin 1 (Control) were 1.4 and 5.7. This confirmed that

process risk is significantly reduced at lower basin MLSS (i.e. in/approaching ‘carbonaceous

mode’)



88

2. Removing surplus solids from the basins during the ‘decant’ phase is preferable due to several

benefits such as: significantly reduced sludge volumes (up to one-third volume of ‘aerate’ phase)

and consequent reduced operation of secondary sludge processing assets, and OPEX savings

3. For the ‘SAS decant’ basin configuration, the required SAS volume per cycle varied throughout

the Trial, in order to achieve (and subsequently maintain) the MLSS target of 1,500 to 2,000mg/l.

The ‘optimum’ range of SAS volume for the Trial basin that achieved this target (basin 6) was

found to be 65 to 85m3 per cycle

4. Close monitoring of basin MLSS is required, particularly if the ‘SAS decant’ configuration is

adopted

5. A lower SAS pump flow rate of 50l/s resulted in more consistent sludge quality over a typical SAS

event, as this prevented “rat-holing”

6. Significant savings in blower energy consumption was achieved – average of 1,000kWh/d

savings estimated. This may have been higher if a steady basin MLSS concentration was

maintained (although this was affected by wider issues on site e.g. existing sludge treatment

capacity and resulting sludge backlog issues)

7. The Trial objectives were achieved; thereby, rendering the Trial successful

8. Following the success of this Trial, the roll-out of carbonaceous mode operation across all the

16No. SBR basins at Liverpool WwTW, has begun.

Acknowledgements

I would like to thank Jeremy B., Andrew S., and Jon H. from United Utilities, for their review and

constructive criticism while writing this paper.

References

EnviroSim. (2016) BioWin 5.0 Software

Appendix

Current Blower Control DO profile

Time

(cycle min) 0 15 30 45 60 75 90 105 120

DO Target

(mg/l) 0 0.37 0.75 1.12 1.50 1.67 1.85 2.05 2.4



89

MODULAR UPGRADE OF AN ASP TO MEET RAPID POPULATION UPSURGE,

NEW WASTE STREAMS AND TOUGHER CONSENTS IN THE UK

Bassey, B.O.1, Ogarekpe, N.M.2, Njunbemere, N.1, Ajare, T.O.1, 1Coventry University, UK, 2Cross River University of Technology, Nigeria

Corresponding Author Email: [email protected]

Abstract

BIOSAS wastewater treatment plant (BWWTP) is activated sludge plant (ASP) located in Warwickshire

County, Southern England. In response to anticipated population increase due to new infrastructure and

industrialization of the area, and stiffer discharge consents by industry regulators, the BWWTP needs a

major upgrade. This paper reports the proposed plant upgrade process aimed at enabling greater influent

wastewater treatment in the face of higher BOD loading and COD reduction requirements. A techno-

economic proposal to guide management in retrofitting the existing facilities in the secondary and tertiary

treatment processes in order to comply with these conditions and remain in business, forms the nucleus of

this work. The most appropriate design solution to retrofit the biological process units of the current plant,

justification for design decisions, design drawings and dimensions, operational and economic efficiency

considerations, and the project delivery plan are presented. Attempts are made to address possible

hydraulic and maintenance problems that could impact the modified plant’s performance during start-up

and continuous operation. The results and recommendations are hoped to be useful for operators,

contractors and consultants involved in similar projects around the globe.

Keywords

Additional tanks; BOD reduction; Capital Expenditure; Gantt chart; Performance monitoring; Pocket ASP;

Retrofitting

Introduction

There is no gain saying the fact that the Activated Sludge Process (ASP) has consistently demonstrated its

capability to meet treated effluent discharge limits. The low footprint requirement makes it a preferred

wastewater technology to waste stabilisation ponds, trickling filters, aerated lagoons and other options for

densely populated areas. Numerous research efforts have been devoted to enhance the applicability of

ASP and thereby improve its treatment efficiency. These include determining the effects of hydraulic and

solid retention times on the fate of tetracycline in the process (Kim et al., 2005); biodegradation and

adsorption of antibiotics (Li & Zhang, 2010); chemical reduction of excess sludge produced (Liu, 2001); fate

of water-soluble azo dyes (Shaul et al., 1991); strategy for minimization of excess sludge production (Liu &

Tay, 2001); comparison of sludge characteristics and performance of a submerged membrane bioreactor

and an ASP at high solids retention time (Massé, Spérandio, & Cabassud, 2006), etc.

The latter study had based its comparison on the entire range of sludge retention time (SRT) of 10–110

days, observing that the Submerged Membrane Reactor (SMBR) achieved very good organic removal

efficiencies as compared to the ASP (Massé, Spérandio & Cabassud, 2006). The immersed membrane

activated sludge process, which entails the coupling of a bioreactor with effluent separation by

microfiltration hollow fibres immersed directly in the bioreactor, provided high degree of treatment in terms

of suspended solids (100%) and organic matter (>96% for COD) as compared to the conventional ASP.

When operated in nitrification-denitrification mode, 99% ammonia and 80% total nitrogen removal were

obtained (Côté, Buisson, Pound & Arakaki, 1997). However, there is paucity of information with regards to

retrofitting an existing plant having the pocket arrangement.

Thesing (2015) had, nevertheless, observed that by retrofitting Integrated Fixed-Film Activated Sludge

(IFAS) systems, communities could upgrade and expand wastewater treatment without the expense and

complications associated with new builds. It, however, stressed that municipal and regional authorities that

make critical decisions regarding expansions and treatment upgrades must understand the capabilities and

limitations of retrofitting IFAS systems, as well as the specialized approach to an IFAS retrofit project. As

the IFAS process has been reported to be similar to ASP and can actually be retrofitted into existing ones

(Di Trapani et al. 2010, Sriwiriyarat et al. 2008), the imports of that work and Kaindl (2010) are instructive

to the wider objectives of the present study.


http://www.sciencedirect.com/science/article/pii/004565359190269J#!

http://www.sciencedirect.com/science/article/pii/S0011916497001288#!






90

Figure 1: Picture of BIOSAS Wastewater Treatment Plant (Bassey et al. 2016:3)

BIOSAS ASP, shown in Figure 1 above, treats combined municipal wastewater for 10,500 heads; with

effluent consent of 25/45/5 (BOD/SS/NH3). In November 2017, phosphorus limits will be added to the

current consent. A proposed petrochemical plant and university sewer are to channel their effluents to the

ASP. These developments thus throw up two choices of either replacing the plant with a new one or

upgrading the existing facilities. The targets set to evaluate the attainment of the plant’s new status were

as follows:

• “1mg/l of total phosphorus in wastewater discharged from the ASP into surface waters

• Reduction of oil and grease content in petrochemical plant effluent wastewater (4ml/day, COD

860mg/l) from 300 mg/l to 5 mg/l in the preliminary process

• Reduced BOD loading (2.5Ml/day, BOD 250mg/l) of the additional municipal wastewater from the

newly established university’s sewer system to the existing BOD/SS/NH3 limits (25/45/5 on a 95

percentile basis)

• Possible onsite sludge processing for resource recovery, energy efficiency and carbon footprint

reduction” (Bassey & Odigie 2016:2).

Bassey & Odigie (2016) and Bassey et al. (2016), previous publications in a series that reports the research

carried out, had addressed the first two and the fourth targets, respectively. This paper thus complements

the earlier publications and is aimed at meeting the third target for retrofitting the plant. Building a new plant

would disrupt municipal sewage processing in the area, being a market community, and pose severe

environmental health risks. Moreover, acquisition of land and obtaining permits for a new build would

require too much time to deliver upon the project objectives and deadlines. The pocket configuration, as

identified in two earlier papers, makes it possible to retrofit parts of the plant without shutting down

production.


This study involves the development of a techno-economic proposal for retrofitting existing secondary and

tertiary treatment facilities at a pocket municipal ASP to meet anticipated population increase due to new

infrastructure and industrialization of the area, as well as existing and new discharge consents. Operational

data used were extracted from laboratory and performance monitoring reports obtained from the plant

operator; environmental data from websites of concerned government agencies; costs averaged from

figures quoted by a handful of equipment and service providers across the UK; while the missing

parameters were obtained from the plant design manuals, industry standard practice guidelines, regulatory

specifications and relevant literature.

Established and improvised design calculations were then carried out to enable specification of sizing for

the required additional process units and modified operational parameters of the retrofitted plant, based on

the foregoing. Costing approximations and task scheduling performed to predict CAPEX and project



91

timeline. The project delivery plan was prepared on a Gantt chart using Microsoft Office Project. The results

so obtained were generally analysed and recommendations made for improved performance.

Plant Retrofit Design

Aeration Zone

This comprises both the aerobic and anoxic basins for ammonia, BOD and suspended solids removal from

the wastewater stream after preliminary and primary treatment (Downing & Nerenberg 2008). Hydraulic

retention time was selected as 12 hours for the entire aeration zone and 2 hours for the anoxic zone from

the ranges provided by standard practice guidelines for wastewater treatment plant design. From the sizing

calculations performed for the retrofitted ASP, one more anoxic tank of 500m3 capacity needs to be

constructed in the secondary process. Considering a maximum volume of 500 m3 for each aerobic tank in

a pocket ASP from the design manuals; four extra aerobic tanks, each of 500m3, are to be constructed to

handle the increased sewage load in a pocket configuration with the additional anoxic tank; in conformance

to the original design of the facility and to prevent hydraulic issues. These and other designed units are

represented in Figure 2, which simplifies the design drawings for the entire upgrade project. Tank sizes are

maximised in order to reduce the number of additional tanks required to handle the upsurge in wastewater

intake, reduce project footprint and optimise the usage of available land at the property. Additional land

acquisition in the UK is a lengthy and rigorous process, which would significantly escalate project costs and

extend the time scheduled for completion.

The food to mass ratio was then calculated as follows, maintaining the same mixed liquor suspended solids

(MLSS) for the existing plant.

F/M Ratio = Qi × BOD ÷ MLSS × Va ---------------------------------- (1)

The MLSS could, however, be altered after plant restart if condition monitoring shows it to be incompatible

with the new set of process parameters, in order to arrest early operational problems (Kaindl 2010).

Final Sedimentation Tank (FST)

The FST section, where the tertiary treatment process occurs, was designed for total capacity that would

handle the additional as well as existing wastewater that exits the secondary process to ensure discharge

quality is met. The surface overflow rate was determined using the relationships below.

SOR = Qi / Ac ------------------------------------------------------------- (2)

Qi = ½(Qt) ----------------------------------------------------------------- (3)

for each FST, as there are two FSTs in the retrofitted plant design and laboratory results show that half of

the sewage load are usually removed in the primary settlement tank (PST) in the existing process. The total

daily flow, Qt, excluded the return activated sludge (RAS) flow, which was assumed to have been

completely removed from the PST once that cycle had been completed and had been taken care of in the

sludge mass balance calculations reported by Bassey et al. (2016).

The clarifier surface area for each FST was computed from the equation below.

Ac = πdc2 / 4 -------------------------------------------------------------- (4)



92

Figure 2: Simplified process flow diagram for the retrofitted BIOSAS ASP (Bassey et al.

2016:16)

A diameter of 12 m was specified for each clarifier from the existing plant’s design manual. Following the

sizing calculations, an additional FST was designed for the retrofitted facility in order to ensure proper

polishing of the effluent from the secondary process in a similar pocket arrangement that would allow

interconnection between both tanks and suspension of either of them for maintenance, whenever the need

arises, without having to shut down the entire process, as illustrated in Figure 2.

Table 1: Designed Dimensions for Proposed BWWTP Upgrade

Operating Parameter Designed Value (Units)

Hydraulic retention time, RTa, for aeration zone 12 hours

Total volume of aeration zone, Va 4.65Ml or 4,650 m3

Total daily flow, Qt, excluding RAS 9.3 Ml/d

Hydraulic retention time, RTaz, for anoxic zone 2 hrs

Volume of anoxic zone, Vaz 0.775Ml or 775m3

Additional sewage volume to anoxic zone 500m3

Volume of aerobic zone = Va – Vaz 3,875m3

Total capacity of 4 existing aerobic tanks, each 445m3 1,780m3

Remaining volume of sewage for aeration 2,095m3

Required number of additional aerobic tanks 4.19 ~ 4

Food to mass ratio, F/M Ratio 0.47kg.BOD/kg.MLSS/day

Diameter of each clarifier, dc 12m

Clarifier surface area, Ac for each FST 113.112m2

Influent flow rate, Qi = 1/2(Qt) for each FST 4650 m3/d

Surface Overflow Rate, SOR 41.11m3/m2.d



93

Table 1 details the various parameters designed in order to retrofit both the secondary and tertiary

processes of the BIOSAS ASP. This study also assumed that since 50% of the sewage load and practically

all the sludge, except little remnants of RAS in the mixed liquor recycle, is removed at the PST, no additional

PST is needed for the plant. The workability of this is, however, dependent upon plant performance

monitoring results after startup following completion of the upgrade project.

Considering the new consent and the expected additional pollution load from the petrochemical company,

there might be need to by-pass some of the storm water in order to avoid hydraulic overloading, which

could lead to the depletion of a large percentage of the active biomass and possibly result in failure of the

system. This is a subject for future research after the upgraded plant has been commissioned.

Project Economics

From September 2016, when the major parts of this study were initially completed to the time of publication

of this paper, the average costs obtained from wastewater contractors and service companies would

definitely have risen. It thus seemed rational to factor this into the overall project costing in order to reflect

the prevailing economic realities.

As the actual current costs could not easily be obtained immediately due to obvious commercial reasons

and bureaucratic bottlenecks, an escalation factor of 1.01 was applied to the overall project costs which

were estimated and presented in Bassey et al. (2016). This was determined by averaging the monthly

inflation rates in the UK from November 2016 to September 2017, which gave approximately 1%. This was

then multiplied by each item of expenditure to obtain the figures presented in Table 2; an increase of

$523,775 over the previous estimates. This could, however, be significantly reduced through proper and

thorough contracting processes, without sacrificing valuable project time and quality of job delivered.

Operating costs were not covered by this paper as these would be largely based on speculations which

may not be realistic considering the uncertainties of startup, with the attendant operational losses, extra

maintenance costs and legal liabilities. Whilst every care had been taken during design to drastically reduce

the probabilities of these issues surfacing, experience in practice would show that they are not entirely

impossible. While normal operating costs except labour could be similar across like facilities elsewhere in

the UK, making such projections would be better informed after three to six months of continuous operations

after commissioning the retrofitted facilities.

Table 2: Capital Expenditure for BWWTP Modification (adapted from Bassey et al. 2016)

ITEM DESCRIPTION COST ($)

Site excavation 521,463

Dissolved air flotation (DAF) for O&G removal in PST 4,550,050

Anaerobic Tank for Phosphorus Removal 654,985

Anoxic tank 2 494,900

Aerobic tank 5 366,630




Anaerobic Digester 1 1,012,020






Sludge Pre-treatment Processes 4,221,800

Biogas storage tank 505,000

Head works – Pumps, fine screens, primary sludge tanks, etc. 28,281,010

Contractors’ Fees 4,040,000

Approvals and permits 73,427

Miscellaneous/Overhead 2,020,000

Total Capital Expenditure 52,901,275



94

Project Delivery Plan

The timeline showing the project delivery schedule for various activities that make up the plant upgrade

programme is illustrated in Appendix 1. From the Gantt chart, it can be seen that the project was originally

billed to commence on 15 October 2016 and scheduled to be completed on 30 April 2017; a total delivery

time of 61/2 months. However, due to delays experienced in the pre-design planning, environmental review

and permitting, as well as contract bidding and award stages, it is still in the concrete works stage.

Nevertheless, since the petrochemical plant is set to take off in late 2018 and the university presently has

a test sewage treatment plant that works effectively during holidays, the BWWTP upgrade can continue

without disruptions to municipal wastewater management in the community. In seasons of peak load, the

reserve capacity of BIOSAS conveniently contains the upsurge from both the community and the excess

sewage load from the university. This, notwithstanding, efforts are presently geared towards

debottlenecking the project execution process. Details of this are beyond the scope of this paper.

Conclusion

• This paper has summarized the technical and economic imperatives of a modification project on

the BIOSAS WWTP aimed at expanding its treatment capacity, particularly in the secondary and

tertiary processes. The goals were to ensure the plant remains relevant for the next 15 years to the

market community and environs, while ultimately boosting returns on investment to the operator.

Proposed improvements were therefore carefully designed to accommodate the anticipated

population upsurge and ensure the discharge effluent meets stringent discharge requirements. The

larger quantity of recycled water delivered is hoped to be useful for agricultural practices at the

university, domestic nondrinking uses around the area and industrial needs at the petrochemical

plant.

• Microbial performance monitoring, especially of the bacteria used at various stages of treatment,

is necessary to prevent sludge bulking, scum formation and other operational problems. F/M ratio

should also be adjusted with wide variations in ambient and operating temperature. These would

enhance the availability of the retrofitted plant. Objective sampling for effluent quality assurance is

also recommended for consistent consent compliance.

• Rigorous condition monitoring, data mining, performance measurement, necessary changes in

process parameters and life cycle cost analysis are required from start-up through operations and

maintenance to address any hydraulic or other ‘teething’ problems that may arise. To this end,

computational fluid dynamics (CFD), economic and project risk management models are currently

being developed for activation with live data beginning from the testing and commissioning phases

of the project to decommissioning. This will help in developing lessons learnt for continuous

improvements and future similar projects.

References

Azimi, A. A., Hooshyari, B., Mehrdadi, N. & Nabi Bidhendi, GH. (2007) Enhanced COD and Nutrient

Removal Efficiency in a hybrid integrated fixed film Activated Sludge Process, Iranian Journal of Science &

Technology, Transaction B, Engineering, 31(B5): 523-533.

Bassey, B.O. & Odigie, O.P. (2016) " Build New or Retrofit: Project Evaluation for an ASP Faced with Oil,

Grease and Phosphorus Removal Consents in England". 10th European Waste Water Management

Conference, held 11 – 12 October 2016, Manchester, UK.

Bassey, B.O., Odigie, O.P., Ajare, T.O. & Oloruntoba, F.M. (2016) “Modification of UK Municipal Activated

Sludge Plant to incorporate Onsite Sludge Digestion”, European Biosolids and Organic Resources

Conference, 15-16 November, Edinburgh, Scotland.

Côté, P., Buisson, H., Pound, C. & Arakaki, G (1997). Immersed membrane activated sludge for the reuse

of municipal wastewater. Desalination, 113 (2-3), 189 – 196.

Di Trapani, D., Mannina, G., Torregrossa, M. & Viviani, G. (2010) Comparison between hybrid moving bed

biofilm reactor and activated sludge system: a pilot plant experiment, Water Science & Technology, 61 (4):

891-902.







95

Downing, L.S. & Nerenberg, R. (2008) Total nitrogen removal in a hybrid, membrane-aerated activated

sludge process, Water Research 42: 3697–3708.

Kaindl, N. (2010) Upgrading of an activated sludge wastewater treatment plant by adding a moving bed

biofilm reactor as pre-treatment and ozonation followed by biofiltration for enhanced COD reduction: design

and operation experience, Water Sci Technology, 62(11): 2710-2719.

Kim, S., Eichhorn, P., Jensen, J. N., Weber, A. S., & Aga, D.S. (2005). Removal of Antibiotics in

Wastewater: Effect of Hydraulic and Solid Retention Times on the Fate of Tetracycline in the Activated

Sludge Process. Environ. Sci. Technol., 39 (15), 5816–5823.

Li, B. & Zhang, T. (2010). Biodegradation and adsorption of antibiotics in the activated sludge process.

Environ. Sci. Technol., 44 (9), 3468–3473.

Liu, Y (2002). Chemically reduced excess sludge production in the activated sludge process.

Chemosphere, 50, 1-7.

Liu, Y. & Tay, J. (2001). Strategy for minimization of excess sludge production from the activated sludge

process. Biotechnology Advances, 19 (2), 97 – 107.

Massé, A., Spérandio, M. & Cabassud, C. (2006). Comparison of sludge characteristics and performance

of a submerged membrane bioreactor and an activated sludge process at high solids retention time. Water

Research, 40 (12), 2405 – 2415.

Shaul, G.M., Holdsworth, T.J., Dempsey, C.R, & Dostal, K.A. (1991). Fate of water soluble azo dyes in the

activated sludge process. Chemosphere, 22(1-2), 107 – 119.

Sriwiriyarat, T., Pittayakool, K., Fongsatitkul, P. & Chinwetkitvanich, S. (2008) Stability and capacity

enhancements of activated sludge process by IFAS technology, Journal of Environmental Science and

Health, Part A, 43:11, 1318-1324.

Thesing, Glenn. (2015) Retrofitting IFAS Systems in existing Activated Sludge Plants, a Kruger Case Study

[online] Available from

<http://www.etec-sales.com/pdf/Kruger%20-%20Case%20Study%20-

%20Retrofitting%20IFAS%20in%20existing%20Activated%20Sludge.pdf







96



97





KINGSPAN CALLS FOR FOOD SERVICE INDUSTRY TO CONSIDER

FOG AS FUEL

Curran, J.,

Kingspan Environmental Northern Ireland

Email [email protected]

Abstract

Sewer blockages are epidemic in the UK, with estimates of 366,999 per year, 70% of them caused by

fat, oil and grease (FOG). Owners of catering businesses must manage their FOG waste, capturing

before it enters the drains, or run the risk of prosecution under Section III of the Water Industry Act

1991. Water companies are increasingly clamping down on infringements, prosecuting offenders who

face heavy fines.

The problem is getting worse, despite the fact that food outlets now have access to smart ways of

disposing of FOG safely with technology and monitoring providing better ways to manage the process.

Kingspan is seeking legislative changes to make monitoring of grease traps a mandatory requirement,

as it is with oil/water separators (PPG 3).

The company also wants to advance the debate around changing societal perception of FOG, from that

of a waste product, to fuel and encourage greater use of it as an ingredient in bio-diesel.

Introduction

Our love of coffee shops and eating out shows no sign of abating, but while this enthusiasm has been

very good both for the hospitality industry and the wider UK economy, it has come at a cost in terms

of food waste.

The rise in the number of food outlets across the country is one of the reasons for a huge increase in

sewer blockages, which are now almost an epidemic.

Water UK, which represents the UK’s water and wastewater utilities, estimates there are now around

366,000 sewer blockages per annum across the country and 70% are caused by fat, oil and grease

(FOG).

The reasons are largely two-fold. There is widespread acceptance and acknowledgement that our

drainage infrastructure was never designed to cater for the demand that we now place upon it – and

that’s true of both wastewater and (increasingly) floodwater run-off. However, it’s also fair to say that

owners of these outlets are not always observing their legal responsibilities about disposing of food

waste – particularly when it comes to FOG.


[email protected]





The reason why FOG is such a problem lies in its physical properties. As hot fat and grease cool,

they congeal, binding together other solids in the sewer and adhering to any object in its path,

including the inside of drainage pipes creating ‘fat bergs’. Thousands of tonnes of FOG are now

believed to be in the sewer at any one time.

Sewer blockages cause huge disruption and a public health nuisance. They can also cause problems

for individual food outlets as a local blockage will cause bad smells, blocked WCs, and will land the

owner with a large bill for the cost of the clean-up.

It is, not surprisingly, a criminal offence under Section 111 of the Water Industry Act 1991 to discharge

into the public sewers any matter which may interfere with the free flow of wastewater. For all

businesses which handle or prepare food, this means they must adhere to a number of legally binding

grease management procedures. Uncontrolled discharge of FOG from a food service

establishment could contravene Section 111 of the Water Industry Act 1991 and result in the

water and sewerage company bringing a prosecution against that establishment, leading to a

fine and recovery of the water company’s costs (See: The Legislative Imperative).

Table 1: The UK Catering Industry by Sector

The largest sector in the catering industry is the hospitality sector, with restaurants, cafes and bars

accounting for around 54% of the market. Health and Education account for a further 20%, whilst






hotels account for a 12% share and the remaining 14% is from other sectors such as offices, leisure

and retail etc.

A breakdown of the estimated number of catering outlets in each sector can be found in Appendix 1.

1. The Legislative Imperative

As indicated in the Introduction, there are a number of legal grease management duties that all

businesses which handle or prepare food must adhere to, regardless of the size of the business.

Building Regulations state that kitchens in commercial hot food premises should be fitted with a BS

EN1825 compliant grease separator or other effective means of grease removal.

Despite this, grease management is increasingly becoming an issue for both the Environment Agency

and the water and sewerage companies. As a result, regulations are tightening and stiffer financial

penalties are being issued for infringement.

Severn Trent is one water company which is taking high-profile action to clean up its sewer network.

In October 2016, in a landmark case, Café Saffron, a restaurant in Codsall, near Wolverhampton, was

reportedly fined more than £5,000 including costs by a court in a prosecution brought by Severn Trent

for blocking nearby sewers with FOG. The restaurant had been pouring fat used in cooking down the

drain. The problem came to light after neighbouring businesses complained they could not flush their

toilets.

In the report, a company spokesperson was quoted as saying that Severn Trent is now clearing a

mammoth "45,000 blockages a year”, and that “fat contributes to the majority of those, as it binds

together all the other things that end up in the sewer rather than the bin and creates huge lumps which

block the sewers.” It also reiterated that this type of legal action was “a last resort for us” but given the

current trend, we can probably expect to see more prosecutions in future.

2. How can FOG be Disposed of Safely?

There are several ways that commercial kitchens can dispose of FOG safely. These include new and

innovative procedures along with the latest technology, which together can better manage the process.

All that is required is a determination by food outlets to make grease management a key focus for their

business. For example:

a) Maintaining grease traps. These separate the FOG from the rest of the wastewater in a

drainage pipe. The collected waste is then collected at regular intervals by a licenced waste oil

collector and a written record of maintenance must be kept

b) Investing in bacterial dosing – the latest generation of specialist grease traps use multiple

strains of environmentally-friendly bacteria which are highly effective at breaking down FOG

c) Selecting remote monitoring – some bacterial dosing systems use sensing technology to

provide remote measuring and monitoring. This technology automatically alerts servicing

teams to the need for emptying trap, eliminating the need for manual checking by kitchen staff.






Kingspan’s new innovative SmartServ Grease System incorporates:

• The Trapzilla Grease Trap which captures up to 96% of grease waste – a major improvement

over existing technology which typically only traps 25% of the grease. The trap holds the fat, oil

and grease with very little water, making it highly efficient and reducing the frequency of pump

outs.

• The IFOG Bacteria Dosing Unit an outlet-pipe bacterial dosing system which is highly effective

at digesting fats, oil and grease, eliminating smells and reducing pipe blockage.

• The Kingspan measuring and monitoring system uses an innovative sonic probe that

measures the levels of grease remotely within the trap, identifies and alerts you to when the trap

is full via email and text and can also map grease traps in multiple locations onto the remote

hosting platform.

• Scheduled Servicing & Maintenance Plans are developed with Kingspan’s engineers. Our

teams use the online platform to identify, remotely, when a trap needs emptying, as well as any

alerts about operational issues, ensuring a faster response.

• Tankering services, through Kingspan’s preferred partner scheme, which provides national

coverage and the option to introduce users to a range of payment options for installation of the

system – including leasing, to help reduce capital outlay

Implementing better management practices through innovative and intelligent FOG management

solutions offers food outlets a number of key benefits. These include:

a) Reduced costs through fewer pump-outs, as a consequence of installing remote measuring

and monitoring systems

b) Clear pipes and drains using bacterial bio-remediation

An M&S store in Dublin reduced pump-out frequency from every month to once every three months,

lowering operations costs by an estimated £2,600 per annum (See: Case Study - M&S In-store

Catering Team Solves Grease Problem).

An effective grease management culture within the kitchen is also important. This could include staff

training around the problems caused by irresponsible FOG disposal and better kitchen practices -

meaning plates should be scraped and wiped down with a paper towel before being placed in the sink

or dishwasher. Pouring boiling water down the sink will not unblock it!






Case Study: M&S In-store Catering Team Solves Grease Problem

Kitchen staff based at Marks & Spencer’s store in Bloomfield, Bangor, Northern Ireland had been

experiencing significant issues with the disposal of fats, oils and grease resulting from food preparation

and waste.

The 500 litre grease trap that the team had been using needed pumping out at least once a month. The

frequency of these regular pump-outs was not only extremely expensive, it caused a lot of unnecessary

disruption and interruption within an already busy working environment.

What’s more, the unpleasant and pervasive smells that emanated from the trapped grease were

blighting nearby food and clothing stores, while the process of opening the trap to check the levels of

grease is never a popular task in any commercial kitchen for the same reason.

Kingspan’s expert team with partners McLaughlin Harvey surveyed the problem on-site before

recommending the complete removal of the grease trap and replacing it with the innovative new

SmartServ Grease System

Since it was installed at the store, the system has proved a huge success. Pump-out frequency has

been reduced to once every three months, lowering operations costs by an estimated £2,600 per

annum.

The remote-monitoring capability provided by the Sensor Probe now means the service company can

respond in good time, before the trap becomes too full.

Furthermore, the outlet pipes are clear and the quality of the effluent discharged into the drains has

significantly improved; a key benefit for Marks & Spencer, supporting its 'green philosophy'.

Gary McClernon, Account Manager for Mc Loughlin Harvey advises, “The introduction of Kingspan’s

SmartServ Grease solution has allowed us to further demonstrate our innovation to M&S; one of our

most important accounts.“

“Kingspan’s SmartServ system clearly demonstrates that disposing of fats, oils and grease can be done

simply, easily, and with the minimum of interruption. Operational costs have been more than halved

due to a reduced pump-out frequency from every month to once every three months and that’s not

accounting for the cost of the disruption of emptying a trap every month. So, overall, it has been a

solution which more than fulfils the brief.”

3. More Legislation Needed

Kingspan is currently working with partner companies to gather cross-industry support for further

changes to the legislation which covers FOG.

We’re keen to see a shift in emphasis towards new best practice methodology for managing grease,

rather than simply tankering (which itself adds to greenhouse gasses) or jetting (flushing it down the

sewer network).

The legislative change that we seek to make monitoring of grease traps a mandatory requirement, as

it is now with oil/water separators (PPG 3).






It’s our belief that only collective action can drive this change through. Furthermore, legislation would

not only act as a lever to change behaviour; it should also provide an incentive to encourage more

companies in the sector to develop more solutions.

An UK industry standard for FOG management products has recently been proposed by the Catering

Equipment Suppliers Association (CESA), the representative body for manufacturers of commercial

kitchen equipment. We’re delighted to have been invited to take part in an exploratory meeting to

discuss this opportunity and potentially help with outlining the form that any such standard should take

before a proposal is put to the BSI.

4. Kingspan Bio-Diesel Partnerships

We also want to use the process of tightening legislation to advance the debate around society’s

perception of FOG, changing the narrative so that it is no longer considered a waste product, but a

valuable commodity.

The latest, and to our minds most sustainable, way of thinking is that FOG should not simply be

discarded but should be considered for recycling and reuse – notably as a source of energy or fuel,

given its high calorific value.

Reusing the recovered FOG means that, as a society, we can optimise source control and ensure future

(bio-) energy security. It also offers attractive payback for food outlets, in partnership with strategic bio-

fuel manufacturers.

Kingspan is working with Argent Energy, a company which pioneered the large scale commercial

production of bio-diesel in the UK. It is now the UK’s foremost sustainable biodiesel producer and fuel

supplier.

The company specialises in the supply of high grade, sustainable diesel for fleet operators. The bio-

¬component of the diesel is made from UK waste FOG.

Argent has two manufacturing facilities for waste-based biodiesel and following a recent £75million

investment, its production capacity will be 145 million litres per year.

Supplying biofuel companies like Argent with grease at more than 90% concentration (from the Trapzilla

Grease Trap) vs that from existing technology, which typically only traps 25% of the grease, offers them

a much more attractive raw material.

Trapzilla FOG Yield

A rough calculation suggests that 600lbs / 270kg of FOG from a single

Trapzilla Grease Trap would yield approximately 40kg (15% yield). That’s

about 40 litres of fuel per trap. So the payback time would be quick.

In the USA, a similar initiative, dubbed “Brown Grease” recovery offers a yield

of just 5%.

Our System will deliver 300% more.






The yield from the FOG is certainly attractive!

We’re also working with tanker companies who cannot (currently) monitor FOG levels in grease traps

remotely. Remote monitoring of FOG levels creates a huge opportunity. The SmartServ system

provides this capability, meaning tanker companies (wherever they are in the country) can gather

information from the in-built Sensor technology and then judge for themselves when to remove the

grease, making the logistics and route planning far simpler. Added to which it takes the burden of

responsibility for monitoring the traps away from the restaurant.

We envisage a new business model that allows tanker companies to get paid by bio-fuel companies to

collect and deliver the FOG.

Conclusion

1. UK sewer blockages are on the rise, as a consequence of the rise in the number of food

outlets, and the growing popularity of eating out. The vast majority of blockages are caused

by FOG

2. Food outlets that do not put in place measures to manage their grease waste face

prosecution, and there is clear evidence that water companies are increasingly seeking to

take this route to recover the cost of removing fat blockages

3. Technology, along with changes in culture and behaviour, can drive the change the sector

needs.

4. To get real change however probably requires tighter legislation, making monitoring of grease

traps mandatory

5. Attitudes towards FOG should also change, shifting from our current perception to it as waste,

to that of fuel – a change that, if widely adopted, will help lower greenhouse gas emissions

and give a commercial value to FOG disposal. Products such as the Trapzilla Grease

Trap that dewater FOG, support this aim by increasing its potential yield

www.kingspanenviro.com/foodwaste

References

Water UK: sewer blockage statistics: http://www.water.org.uk/policy/environment/waste-and-

wastewater/fats-oils-and-grease

Section 111 of the Water Industry Act 1991: http://www.legislation.gov.uk/ukpga/1991/56/section/111

AMA Research: NAME OF REPORT (Non-Domestic Catering Equipment Market Report - UK 2017-

2021 Analysis)

Building Regulations 2015 Part H: Drainage and Waste Disposal (2015 edition)

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/442889/BR_PDF_AD_

H_2015.pdf

Café Saffron prosecution, October 2016: (www.bighospitality.co.uk/Business/Restaurant-fined-5k-for-

blocking-sewers-with-fat)






Office of National Statistics (ONS) food outlets 2014 – link to online

Appendix 1 – An Overview of Food Outlets

• The restaurant and mobile food services sector includes both fast food and licensed

restaurant operators but also covers take-away food and sandwich shops and stands.

• This sector as a whole has been showing a rising trend out of the recession with March 2014

ONS figures showing the number of enterprises having risen by 5.6% between 2013-2014 to

68,500 enterprises.

• Allowing for large chain restaurant enterprises the actual number of restaurant and mobile

food service establishments rose from 76,265 units in 2013 to 81,085 in 2014, a rise of 6.3%.

• The majority of restaurants (c.55,000) are small establishments with less than 10 employees.

• The number of beverage serving establishments declined from 49,960 units in 2013 to 49,180

units in 2014, a decline of 1.6%. This sector represents the licensed alcohol establishments

and is dominated by pubs but also includes night clubs, social clubs, bars and taverns

covering independents, managed and tenanted units.

• In 2013, more than 80% of pubs were serving food.

• There are more than 16,500 coffee shops in the UK

• Health and education represent the largest users of catering equipment in the public sector.

More than 8.5m pupils attend over 24,000 schools in England including nursery schools, state

funded primary schools, state-funded secondary schools, special schools, pupil referral units

and independent schools. Meals are served to around 4m pupils daily and in England over

70% of school meals are provided in-house.

• Smaller schools without kitchen facilities, are currently estimated at 1,700 in England, and

these rely on catering providers to cook and deliver meals. They are being encouraged to

consider where possible having meals cooked on site rather than relying on a transported or

brought in services.

• According to Hotel Industry Magazine there are c 45,000 hotels in the UK.

• The contract catering market is believed to have doubled in size over the last 20 years and

continues to expand and show positive growth. Facilities Management companies have over

14,400 outlets and employ over 132,000 staff in the UK.

• In retail, many grocery outlets and department stores offer some form of catering – from

coffee shops through to full self-serve style restaurants. Supermarkets often possess

rotisserie ovens, delicatessen counters, meat and cheese counters, staff canteens and

customer cafes.

• Garden centres have also seen catering provision increasing, with food becoming one of their

primary sources of revenue.




106

TECHNICAL AND COMMERCIAL CONSIDERATIONS IN THE REMOVAL OF

PRIORITY SUBSTANCES AS SPECIFIED WITHIN THE EU WATER FRAMEWORK

DIRECTIVE FROM TREATED DOMESTIC SEWAGE: CASE STUDY FROM A

16,000 PE NON-RURAL WWTP

Nkrumah-Amoako, K., Khan, M., Parocki, D., Nabeerasool A.M.

Arvia Technology Ltd, UK


Abstract

Arvia Technology, a specialist water treatment company, has developed a unique approach for the

treatment of hard to treat (recalcitrant) and toxic compounds. The treatment process is based on a

proprietary graphite adsorbent Nyex™ which self regenerates in-situ without the use of chemicals.

Over a 12-month period ending March 2017, Arvia Technology in conjunction with Anglian Water have

operated a pilot unit at a 16,000 PE WwTW to investigate the removal of a range of micropollutants

including emerging contaminants (Priority Substances). The project was part of the EA sponsored

Chemicals Investigation Programme (CIP2).

An extensive list of contaminants was selected for study, including metals, pharmaceutical and chemical

residuals. Removal efficiencies of up to 99 % were routinely demonstrated at an operating cost of up to

0.15 kWhm-3. Results of the trial were used to cost up a range of full scale systems from 2,000 - 500,000

PE for companies such as Anglian Water and Thames Water.

Keywords

Electrochemical oxidation, Adsorption, Advanced Oxidation Process, Self-regenerating adsorbent,

Nyex, Organic Destruction Cell

Introduction

Arvia Technology commissioned a 12-month Pilot demonstration of its Organic Destruction Cell

technology at St. Ives Wastewater Treatment Works (WwTW) in Cambridgeshire for tertiary removal of a

range of micropollutants from the effluent of the wastewater plant. This was part of a UK Water Industry

Research’s (UKWIR’s) Chemicals Investigation Programme (CIP), a £35 million investigation into the

source and removal of trace substances in the WwTWs’ effluents with the view to establish a basis of a

national-scale assessment of the risks posed by chemicals in WwTW discharges (European Council

Directive 2000/60/EC).

In preparation of the new Asset Management Period 6 (AMP6, 2015-2020) under the Water Framework

Directive, this CIP was updated into a new £140 million CIP2 programme which focused on more site-

specific issues and removal of priority substances, substances pending review to be classified as priority

substances and other micropollutant compounds of specific concern (Bolong et al., 2009; Luo et al.,

2014). Arvia’s ODC technology was selected as a tertiary treatment option for Phase C2b – “Pilot plants

– Technologies not looked at in AMP5 CIP”.

The fates of a total of 55 determinands, listed below, were investigated over the course of the 12-month

study:




107

• Metals: nickel, lead, copper, zinc, cadmium, mercury, iron, aluminium, chromium, both

dissolved and total fraction;

• Priority substances: “penta” cogeners 28, 47, 99, 100, 153 and 154 (BDEs), di (2-ethylhexyl)

phthalate (DEHP), nonylphenol, nonylphenol with 1, 2 and 3 ethoxylate units, tributyltin

compunds (TBT), octylphenols (4-(1,1',3,3'- tetramethylbutyl)-phenol),

hexabromocyclododecane, benzo(a)pyrene;

• Special pollutants: Perfluorooctane sulfonic acid (PFOS), perfluorooctanoic acid (PFOA),

fluoranthene, triclosan, cyphermetrin;

• Steroids: estrone (E1), 17α-estradiol (E2), 17β-ethinyl estradiol (EE2);

• Pharmaceuticals: diclofenac, ibuprofen, atorvastatin, para and ortho hydroxy atorvastatin,

propranolol, atenolol, erythromycin, norerythromycin, azithromycin, clarithromycin, ciprofloxacin,

metformin, ranitidine, carbamazepine, 10,11-epoxy carbamazepine, sertraline, norsetraline,

fluoxetine, tamoxifen, trixylenyl phosphate, 1,2,3 benzotriazole (tautomers 1H and 2H),

tolytriazole (4, 5 methyl benzotriazole).

Arvia’s ODC technology is primarily an organics treatment technology, and thus no treatment of the

metals was anticipated. For the purpose of this work, a selected list of priority substances and

pharmaceutical compounds are discussed.

The ODC Technology

Arvia’s ODC technology is an innovative modular technology that combines adsorption and

electrochemical oxidation in a single process. The process works by adsorbing organic compounds from

solution and achieving complete mineralisation of the adsorbed species off the surface of the proprietary

adsorbent (Asghar et al., 2009; Nkrumah-Amoako et al., 2014). Complete mineralisation is achieved

electrochemically by passing a small current density across the bed of adsorbent, allowing subsequent

adsorption onto the same adsorbent. The self-regenerating nature of the adsorbent eliminates the need

for adsorbent replacement, as is the case in most adsorption systems. The low current densities applied

in the ODC technology makes it economically more viable than a range of advanced oxidation processes

such as ozone and UV, as well as reducing the kinetics of by-product formation.

Some of our main benefits include:

• No toxic by-products or sludge produced

• Low operational costs as energy is used in proportion to the organics being destroyed (trace

level of organics = trace level of energy) • Chemical free and environmentally sound

• Modular and scalable design to suit requirements

• Safe to operate, low maintenance system

A schematic of the Arvia ODC technology is shown in Figure 1, and shows the fate of organic

compounds as they get undergo simultaneous adsorption and electrochemical oxidation to full

mineralization.



108

Figure 1: Schematic of the Arvia ODC Technology cell

Arvia’s proprietary adsorbent, Nyex, is central to this innovative technology, and acts as a three-

dimensional electrode to allow electrochemical oxidation at low voltages, owing to its high electrical

conductivity and non-porous nature. A schematic of Nyex is shown in Figure 2.

Figure 2: A schematic of Nyex, the three-dimensional electrode adsorbents used in the Arvia

Process

Pilot Case Study – Micropollutant tertiary treatment of WwTP effluent

A full turn-key containerised Arvia ODC system, shown in Figure , was installed on Anglian Water’s St.

Ives site to treat the effluent from their wastewater works. The system was commissioned to remove the

organic fraction of the 55 determinands.

Standard Arvia modules contain 22 cells, and the system used in this study was equivalent to one

standard Arvia cell. The containerised system contained telemetry and online data capture to allow

Arvia to monitor system performance remotely.



109

Table 1: Standard Arvia module parameters

Parameter Value

Flowrate (up to) 13 m3/hr

Typical Operating Current 50 Amps

Typical Operating Voltage

Module Footprint

20 – 50 Volts

1.2 m x 0.7 m

Figure 3: Schematic of Arvia 22-cell module

Figure 4: Schematic of the WwTP showing where the Arvia Technology is best located to offer

effective tertiary treatment for micropollutants



110

Figure 5: Schematic of the turn-key containerized ODC treatment system commissioned at

the St. Ives site for the CIP2 study

This plant was run over the course of 12 months, and produced excellent removal rates of the priority

substances in the effluent of the treated wastewater.

Laboratory Treatability Study

Besides the Pilot system, a treatability study on a laboratory scale system of the technology was carried

out to evaluate and optimise the system. Raw water was spiked with ppb levels of these compounds,

and treated through a lab version of the Arvia ODC in a single pass mode. Results from this study were

used to optimise the pilot system.

Results and Discussions

The results generated from laboratory treatability trials and the Pilot plant are shown in Figure 6.

Average removal efficiencies of 96% was achieved for the priority substances in the lab trials, while the

pilot trial results were comparably effective.



111

Figure 6: Comparison between the lab-scale trial results and the pilot plant results

showing comparable effective removal of the micropollutant compounds

Case Study for 16,000 PE WwTW

Results from the CIP2 study have been extrapolated to a 16,000 PE case study. This section aims to

provide details of an Arvia ODC system that will be required to provide tertiary treatment for the effluent

from such a WwTW.

The following assumptions have been made in the calculation:

• Plant sized for peak flow

• Peak flow is assumed to be 3 X DWF for PE >= 10,000

• Per capita consumption estimated as 228 litres/person/day

A 16,000 PE plant capable of providing excellent tertiary removal of trace metals was sized as follows:

• Estimated Arvia ODC cost = circa £525,000

• Estimated out turn cost = circa £1,050,000

• Power consumption = 0.09 kWh/m3

• OpEx per annum = circa £43,000 including maintenance

• % of current UK bills = circa 6%

• Total plant cell surface area (excluding ancillaries) = circa 500 m2



112

Figure 7: A model of 2 standard 22-cell modules connected in series as part of a suite of

modules that will serve a 16,000 PE WwTP. Each module is capable of circa 13 m3/h

Comparison against other technologies

A comparison of the performance of the Arvia ODC technology against other more conventional tertiary

treatment processes in the form of a multi-criteria analysis (MCA) indicated that the Arvia Treatment

solution comparted favourably. Conventional wastewater technologies including GAC, Ozone, UV-

H2O2, Membrane Bioreactor and Membrane Filtration were compared for micropollutant removal. For

removal of micropollutant organics compounds from WwTP effluent, the MCA concluded the following

across the nine criteria measured, Arvia’s ODC technology scored highest in 6 out of the 9 pairwise

technology comparisons, as shown in Table 2 (Pizzagalli, 2017). The technologies were compared and

scored relative to each other across the following criteria:

• Range of technology (RT)

• Micropollutant removal ability (MPR)

• Total cost of installing technology (TC)

• Energy Consumption (EC)

• Chemicals use dependency (CU)

• Production of waste by-products (PWB)

• Flexibility (F)

• Reliability (R)

• Ease of maintenance and operation (EMO)



113

Table 2: Pairwise technology comparison of Arvia ODC against 5 more conventional

wastewater treatment technologies. The results showed that the ODC scored

highest in 6 out of the 9 comparisons

RT MPR TC EC CU PWB F R EMO

ODC 0.48 0.14 0.50 0.49 0.27 0.44 0.17 0.13 0.54

GAC 0.23 0.24 0.09 0.28 0.27 0.11 0.35 0.13 0.32

Ozone 0.09 0.08 0.28 0.03 0.06 0.04 0.03 0.04 0.03

UV-H2O2 0.09 0.08 0.15 0.03 0.27 0.04 0.03 0.04 0.10

MBR 0.07 0.24 0.03 0.16 0.04 0.10 0.35 0.30 0.06

MF 0.03 0.04 0.04 0.09 0.04 0.23 0.09 0.30 0.06

Pizzagalli 2017 concluded that “according to this multi-criteria analysis, the Organic Destruction Cell

could, when scaled up for future application, lead on other technologies for tertiary treatment, presenting

itself as a valid alternative for micropollutants removal from wastewater on every account”

Conclusions

The Arvia ODC technology has proven to be successful at efficiently treating the effluent from WwTP to

remove a range of micro-pollutants. This has been demonstrated through a variety of laboratory

treatability trials and a 12-month pilot study at a 16,000 PE WwTP. This report has discussed the

treatment achieved, and a scaled-up plant to treat the whole effluent from this 16,000 PE WwTP and

similar plants.

The modularity of the technology allows for linear scalability, and thus, the baseline costs of the system

can be extrapolated from the presented costs for this case study 16,000 PE tertiary plant.

On the back of this study, a range of full scale systems for 2,000 – 500,000 PE WwTPs have been costed

up for Anglian Water and Thames Water for micropollutant removal. Arvia remain confident of being able

to supply containerised turnkey pilot plants to treat micropollutants from WwTP effluents, and welcomes

interests from water bodies seeking solution to get in touch.

Acknowledgements

Arvia Technology would like to thank Anglian Water for their contribution to this Pilot study, as well as

UKWIR for their continued support to helping tackle some of the persistent water issues facing the UK.

References

European Commission, “DIRECTIVE 2000/60/EC OF THE EUROPEAN PARLIAMENT AND OF THE

COUNCIL of 23 October 2000 establishing a framework for Community action in the field of water

policy,” Off. J. Eur. Communities, vol. L 269, no. September 2000, pp. 1–15, 2000.

N. Bolong, A. F. Ismail, M. R. Salim, and T. Matsuura, “A review of the effects of emerging

contaminants in wastewater and options for their removal,” DES, vol. 239, no. 239, pp. 229–246, 2009.

Y. Luo, W. Guo, H. H. Ngo, L. D. Nghiem, F. I. Hai, J. Zhang, S. Liang, and X. C. Wang, “A review on

the occurrence of micropollutants in the aquatic environment and their fate and removal during

wastewater treatment,” Sci. Total Environ., vol. 473–474, pp. 619–641, 2014.



114

H. M. A. Asghar, E. P. L. Roberts, S. N. Hussain, A. K. Campen, and N. W. Brown, “Wastewater

treatment by adsorption with electrochemical regeneration using graphite-based adsorbents,” J. Appl.

Electrochem., vol. 42, no. 9, pp. 797–807, Sep. 2012

K. Nkrumah-Amoako, E. P. L. Roberts, N. W. Brown, and S. M. Holmes, “The effects of anodic

treatment on the surface chemistry of a Graphite Intercalation Compound,” Electrochim. Acta, vol. 135,

pp. 568–577, 2014.

Pizzagalli, G. (2017) MSc_Res Dissertation: Removal of Micropollutants From Wastewater Combining

Adsorption And Electrochemical Regeneration, Cranfield Water Institute.



115

ADVANCED OXIDATION PROCESSES AND NON-THERMAL PLASMA FOR THE

REMOVAL OF EMERGING CONTAMINANTS IN WATER

Tizaoui, C. and Ni Y.

College of Engineering, Bay Campus, Swansea University, Swansea SA1 8EN, UK

Email: [email protected]

Abstract

The occurrence of emerging contaminants (ECs) commonly known as micropollutants or trace

contaminants into the aquatic environment has become an important issue due to the negative effects

that these substances cause to aquatic species and humans. Examples of ECs include steroid hormones,

pharmaceutical and personal care products, pesticides and surfactants. Conventional wastewater

treatment plants were not designed to specifically remove ECs, which makes them a major source of

these substances in the environment. New treatment technologies are therefore required to address this

emerging problem. This paper discusses the application of two novel oxidation techniques for the removal

of estrogenic contaminants in water. An ozone-based process termed as liquid/liquid-ozone (LLO) used

a solvent that was initially charged with ozone before being reacted with the wastewater. After reaction,

the solvent was separated from the wastewater and recycled back for further ozone loading. Non-thermal

plasma (NTP) used high electrical voltage to generate a cocktail of reactive species that efficiently

degraded the contaminants. Both techniques have been found effective to degrade potent estrogenic

hormones (Estrone (E1), 17β-estradiol (E2), and 17α-ethinylestradiol (EE2)). Particularly the LLO process

has shown significant cost savings as compared to conventional gas/liquid ozone processes.

Keywords

Emerging contaminants, estrogens, liquid-liquid extraction, non-thermal plasma, oxidation, ozone.

Introduction

The widespread occurrence in water bodies of micropollutants, also termed as emerging contaminants

(ECs), has triggered over the past years increased concerns regarding exposure of human and aquatic

species to these substances (Luo et al. 2014; Klatte et al. 2017; Tijani et al. 2013). This is because

exposure to ECs, despite being found at low concentrations, resulted in effects such as disruption of the

endocrine system, reduction in sperm count in males (Swan et al. 1997; Uzumcu et al. 2004), intersex in

fish (Harries et al. 1997; Jobling et al. 2009), cancers (Bergman et al. 2013) and proliferation of

antimicrobial resistance genes (Malik et al. 2015; Guardabassi et al. 1998). ECs exist in water bodies at

very low concentrations ranging from sub-ng/L to few g/L and cover a vast and expanding list of

anthropogenic and natural substances (e.g. steroid hormones, pharmaceuticals and personal care

products, pesticides, plasticizers, and many other substances). Because of their extremely potent effects,

the natural and synthetic steroidal estrogens, are of major concerns. They originate principally from

human excretion, are ubiquitously spread in almost all waters, and they can cause upon exposure to them

effects such as disruption of the endocrine system. Although the topic of micropollutants occurrence in

the environment has been raised as an issue as early as the 1950s (Fisher et al. 1952), only recently that

it has gained significant interest from the scientific community, water industry, regulatory bodies and the

media. This is largely due to the recent advancement in analytical instrumentation which provided the

tools necessary to accurately measure and identify these substances, commonly found at extremely low

concentrations. The main source of ECs in the aquatic environment is effluent discharge from wastewater

treatment plants (WWTPs). This is because conventional WWTPs were designed to reduce mainly the

parameters BOD and suspended solids rather than to remove specific substances such as ECs.

[email protected]



116

Development of new technologies to efficiently and economically remove MPs from wastewater is hence

very important to the wastewater industry since regulations aimed at controlling the discharge of selected

ECs have started to emerge (Audenaert et al. 2014; Carvalho et al. 2015). This paper will discuss two

novel technologies used to remove potent estrogenic hormones including Estrone (E1), 17β-estradiol

(E2), and 17α-ethinylestradiol (EE2) in water. The first of the two technologies was liquid/liquid-ozone

(LLO) process which used a solvent that is immiscible in water and has high capacity for ozone absorption

to extract and degrade the estrogens. The solvent was recycled in the process so it was recharged with

ozone and used again. A second process used non-thermal plasma (NTP) to generate a cocktail of

reactive species that efficiently degraded the contaminants. NTP was produced by applying high voltage

between two electrodes.


Materials

Estrone (E1), 17β-estradiol (E2), 17α-ethinylestradiol (EE2) and deuterated estrone (internal standard))

at high purity (>98%) were obtained from Sigma Aldrich (Dorset, UK). All solvents were HPLC grade

purchased from Sigma Aldrich UK or VWR UK. Ultrapure deionised water (Resistivity 18M.cm, Millipore

Q system) was used for preparing the aqueous solutions. Stock solutions at 1g/L of each steroid were

prepared in methanol from which working solutions were prepared by water dilution to a desired

concentration. The stock solutions were stored at -18oC. Real effluent was sampled from Marley sewage

treatment works, Yorkshire, UK. Decamethylcyclopentasiloxane (D5) solvent was purchased from Dow

Corning, UK. Ozone gas was generated from pure oxygen with a Lab2B ozone generator (Ozonia Triogen

Ltd., UK) and an ultraviolet ozone analyser (BMT 963, BMT Messtechnik, Germany) was used to measure

the ozone gas concentration.

Liquid/liquid-ozone treatment process

Ozone was first dissolved in D5 using a pre-set inlet ozone gas concentration of 40 g/m3 NTP until

saturation of the solvent. The ozone-rich solvent was then delivered to the wastewater reactor. A mixer

was used to thoroughly mix the two phases. At the end of the reaction time, the two phases were set to

separate by gravity and the solvent was recovered and used for further ozone absorption. A sample from

the treated wastewater was collected for estrogens analysis. A representation of the LLO process is

shown in Figure 1.

Figure 1: LLO process

Non-Thermal Plasma treatment process

Non-thermal plasma was generated using a high voltage (up to 30 kV) pulse generator (MPC3010S-

50SP, Suematsu Electronics, Japan). A four-channel oscilloscope (Tektronix MDO3024) with sampling

rate of 2.5 GS/s was used to measure the pulse voltage and current waveforms using appropriate voltage

Extraction

and oxidation

of

contaminants



117

and current probes. The high voltage (HV) electrode was placed above the surface of the solution

containing 1mg/L of EE2 and the ground electrode was immersed inside the solution. The reactive

species generated in the plasma were transferred to the liquid phase via the interfacial area between the

liquid and gas phases. The volume of the liquid was set to 80 mL and the gap between the HV electrode

and the water surface was kept at 1cm.

Analytical methods

Ozone concentrations in the aqueous and solvent phases were measured by the indigo method (Nobbs

and Tizaoui 2014). The estrogens concentrations were measured by MRM assay using LC-MS-MS

(Waters Alliance 2695, triple quadrupole Quattro Ultima mass spectrometer, ESI mode by Waters,

Elstree, Hertfordshire, UK). The chromatographic conditions were: C18 column (Gemini NX, 3m,

150×2mm, Phenomenex, Macclesfield, UK), 55%ACN:45%water mobile phase, 0.2 mL/min flow rate,

and 10 L injection volume. Solid phase extraction (Supelco Manifold, C8 cartridges) was used to pre-

concentrate and purify the effluent samples before LC-MS-MS analysis. Calibration curves were

determined for each estrogen using the ratio of the signal area to the signal area of the internal standard

as function of concentration.


The MRM transitions and retention times are summarised in Table 1. Solid phase extraction recoveries were around 80%.

Table 1: MRM transitions and retention times

E1 E2 EE2 DE1

MRM (m/z) 269143 271143 295147 273147 RT (min) 3.85 3.17 3.51 3.82

Ozone absorption in the solvent and estrogens distribution

The concentration of ozone absorbed into the solvent was measured using either the indigo method or

integration of the output ozone concentration curve (Equation 1). A typical curve is shown in Figure 2.

Both methods gave similar results. The ozone solubility into the solvent was found to be around ten times

higher than that in water. The distribution coefficients of the estrogens were around 2 indicating that the

solvent does not only exhibit high ozone solubility but is also efficient at extracting the estrogens.

Figure 2: Output O3 concentration

dtCCV

QC

t

AGAGAL 0

0 (Eq. 1)

where: CAL solvent O3 concentration, Q gas flow rate, V volume of solution, CAG0 and CAG gas O3 concentrations at the inlet and outlet respectively, and t time.

Removal of estrogens with LLO

0

10

20

30

40

-2 0 2 4 6 8

CA

G0-C

AG

(g/m

3)

Time (min)



118

The degradation of the estrogens was studied using individual solutions and mixtures in deionised water.

The LLO technique was used and its performance was compared to gas phase ozonation (termed here

liquid/gas-ozone LGO). The times to reach almost full degradation of the three estrogens using LGO were

2.5 min, 0. 8 min, and 0.9 min for E1, E2 and EE2 respectively whilst the degradation with LLO was

extremely fast to the point that all samples exhibited concentrations below the detection limit within times

less than 0.3 min. Figure 3 shows the change of E1 concentration as function of time for both techniques

LGO and LLO; similar curves were also obtained for E2 and EE2.

Figure 3: Removal of E1 by LGO and LLO

The sharp drop of E1 concentration when LLO was used indicates that LLO is indeed very effective to

remove the estrogens. This is possibly the result of synergetic effect of extraction and oxidation reactions

by the ozone-rich solvent in contact with the wastewater. The estrogens are hydrophobic organic

molecules, so they have tendency to distribute to the solvent phase where the ozone reactions take place.

The LLO reaction time was then set to 0.5 min to treat a secondary treated effluent collected from the

final point of the treatment process at Marley STW and the ozone dose was changed between 0.5 to 1.8

mg/L by changing the ratio of solvent volume to the wastewater volume. The results showed that the

three estrogens were fully removed by at least 95%. An ozone dose of 1mg/L and 1% solvent ratio were

hence chosen to simulate the LLO process to treat 400L/s wastewater at concentrations of each estrogen

set to 100 ng/L. The solvent was recycled in the process at a flow rate of 4 L/s and the contact time

between the ozone-rich solvent and wastewater was 1 minute. The simulation results indicated that the

operating cost of the LLO process was around £0.11/ML (ML: million litre of wastewater), which is

significantly lower than a cost figure reported for conventional ozone gas process (£4/ML) (Churchley et

al. 2011).

Removal of estrogens with NTP

Non-thermal plasma generates an array of highly reactive species including ozone, hydrogen peroxide,

hydroxyl radicals, peroxynitric acid, UV, shock waves, ultra sound etc. Exposure to this cocktail of plasma

species leads to significant degradation of the estrogens. In this study the degradation of EE2 was

evaluated using HV discharge above the surface of the water. Figure 4 shows the effect of pulse

frequency on the degradation of EE2 as function of time. As the frequency increased, the degradation

rates also increased. The half-life times (i.e. C/C0=0.5) were 38, 21 and 12 min for the frequencies 100,

300, and 500 Hz respectively. The increased rates of degradation as function of the pulse frequency is

the result of increased energy deposited in the system. This has also been confirmed by increased rates

when the voltage was increased.

0

0.2

0.4

0.6

0.8

1

0 1 2 3

C/C

0

time (min)

LGO

LLO



119

Figure 4: Degradation of EE2 with NTP at different pulse frequencies

Conclusion

Substances such as pharmaceuticals and personal care products, estrogenic hormones, and pesticides

(collectively known as micropollutants or emerging contaminants) are causing “in silence” significant

damage to the environment and human health and only recently that the problem has been recognised.

New emerging regulatory measures to control the discharge of these substances into the environment

are likely to come which will require upgrades with new treatment technologies since conventional

wastewater treatment processes were not designed to deal with such substances. A novel liquid/liquid-

ozone (LLO) process was tested to remove the estrogenic hormones E1, E2 and EE2 and was found

highly effective since it provided synergistic effect of both extraction of the estrogens and their degradation

with ozone. Simulation has shown that the LLO process is largely less expensive than conventional ozone

processes but further pilot trials are required to ascertain the true cost of the technology. Non-thermal

plasma was also found affective to degrade 17α-ethinylestradiol and further study on the energy demand,

by-products formed and better NTP reactor design are under way.

Acknowledgments

Funding was received from the Engineering and Physical Sciences Research Council (Grant no.

EP/M017141/1) for the NTP study.

References

Audenaert, W. T. M., Chys, M., Auvinen, H., Dumoulin, A., Rousseau, D. and Hulle, S. W. H. V. (2014)

(Future) Regulation of Trace Organic Compounds in WWTP Effluents as a Driver of Advanced

Wastewater Treatment. Ozone News: The Newsletter of the International Ozone Association, 42, 17-23.

Bergman, Å., Heindel, J. J., Jobling, S., Kidd, K. A. and Zoeller, R. T. (2013) State of the Science of

Endocrine Disrupting Chemicals 2012 Summary for Decision-Makers. Available at

http://www.who.int/ceh/publications/endocrine/en/.

Carvalho, R. N., Ceriani, L., Ippolito, A. and Lettieri, T. (2015) Development of the first Watch List under

the Environmental Quality Standards Directive. Available at https://ec.europa.eu/jrc/en/news/first-watch-

list-emerging-water-pollutants.

Churchley, J., Drage, B., Cope, E., Narroway, Y., Ried, A., Swierk, T., Alexander, K. and Kanda, R. (2011)

Performance of ozone for EDC removal from sewage effluent. 20th IOA World Congress - 6th IUVA World

Congress. Ozone and UV Leading-edge science and Technologies. Paris, France: International Ozone

Association.

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25 30

C/C

0

time (min)

100 Hz

300 Hz

500 Hz

http://www.who.int/ceh/publications/endocrine/en/



120

Fisher, A. L., Keasling, H. H. and Schueler, F. W. (1952) Estrogenic action of some DDT analogues. Proc

Soc Exp Biol Med, 81, 439-41.

Guardabassi, L., Petersen, A., Olsen, J. E. and Dalsgaard, A. (1998) Antibiotic Resistance in

Acinetobacter spp. Isolated from Sewers Receiving Waste Effluent from a Hospital and a Pharmaceutical

Plant. Applied and Environmental Microbiology, 64, 3499-3502.

Harries, J. E., Sheahan, D. A., Jobling, S., Matthiessen, P., Neall, M., Sumpter, J. P., Taylor, T. and

Zaman, N. (1997) Estrogenic activity in five United Kingdom rivers detected by measurement of

vitellogenesis in caged male trout. Environmental Toxicology And Chemistry, 16, 534-542.

Jobling, S., Burn, R. W., Thorpe, K., Williams, R. and Tyler, C. (2009) Statistical modeling suggests that

antiandrogens in effluents from wastewater treatment works contribute to widespread sexual disruption

in fish living in English rivers. Environ Health Perspect, 117, 797-802.

Klatte, S., Schaefer, H. C. and Hempel, M. (2017) Pharmaceuticals in the environment - A short review

on options to minimize the exposure of humans, animals and ecosystems. Sustainable Chemistry and

Pharmacy, 5, 61-66.

Luo, Y., Guo, W., Ngo, H. H., Long Duc, N., Hai, F. I., Zhang, J., Liang, S. and Wang, X. C. (2014) A

review on the occurrence of micropollutants in the aquatic environment and their fate and removal during

wastewater treatment. Science of the Total Environment, 473, 619-641.

Malik, A., Khan, F., Rizvi, M., Shukla, I., Afaq, S. and Sultan, A. (2015) Trends in Antimicrobial Resistance

of Bacterial Isolates Circulating in Sewage Waters of Aligarh Region over a Period of 14 Years. Asian

Journal of Water Environment and Pollution, 12, 69-74.

Nobbs, J. and Tizaoui, C. (2014) A Modified Indigo Method for the Determination of Ozone in Nonaqueous

Solvents. Ozone: Science & Engineering, 36, 110-120.

Swan, S. H., Elkin, E. P. and Fenster, L. (1997) Have sperm densities declined? A reanalysis of global

trend data. Environmental Health Perspectives, 105, 1228-1232.

Tijani, J. O., Fatoba, O. O. and Petrik, L. F. (2013) A Review of Pharmaceuticals and Endocrine-Disrupting

Compounds: Sources, Effects, Removal, and Detections. Water Air and Soil Pollution, 224.

Uzumcu, M., Suzuki, H. and Skinner, M. K. (2004) Effect of the anti-androgenic endocrine disruptor

vinclozolin on embryonic testis cord formation and postnatal testis development and function.

Reproductive Toxicology, 18, 765-774.



121

EARLY LIFE PERFORMANCE OF STONEYFORD INTEGRATED CONSTRUCTED

WETLAND

Hall, L.1; Woodward, D. 1; McDermott, R. 1; McCurdy, D. 2; Griffin, J. 2; Crabbe, D.2 1Ulster University, 2 Northern Ireland Water, Northern Ireland


Abstract

A lack of understanding of the design and appraisal of Integrated Constructed Wetlands (ICW) is limiting

their development for domestic wastewater treatment in Northern Ireland. This study develops the

understanding of ICWs, and their potential as sustainable alternatives to traditional wastewater treatment

works. This information will aid in the decision making process for future provisions of ICWs.

A full scale, 5 pond system and small scale ‘Test Rig’ were pioneered by Northern Ireland Water and

Ulster University to assess ICW appropriateness as wastewater treatment works and test key variables

that impact performance, including ICW design, weather conditions and time.

Weekly samples were taken from each pond to monitor water quality for BOD, Ammonia, Suspended

Solids and COD. Results identified and confirmed trends in ICW performance over time and area. Data

from the small-scale test rig highlighted connections between design variables and treatment

performance. The results found no significant correlation between changes in weather and ICW

performance.

Keywords

Design, Integrated Constructed Wetlands, Performance, Sustainable, Treatment, Wastewater

Introduction

This paper considers the early life performance of Stoneyford Integrated Constructed Wetland in treating

domestic wastewater in Northern Ireland. It details issues regarding the design, construction, operation

and maintenance of an ICW from a full scale pilot system commissioned by Northern Ireland Water (NIW).

The ICW is located at Stoneyford, County Antrim.

An ICW is an engineered system specifically designed to simulate the bio-filtration processes of a natural

system to remediate contaminated wastewater. This is done using strategically chosen aquatic plants

which filter and remove contaminants from the water as it flows through the ponds of the constructed

wetland. ICWs are designed to work as an integrated ecosystem, combining the functions of the natural

environment with human activities (Moshiri, G. A, 1993). In other words, they are a natural means of

treating wastewater in a controlled and manageable method (Vymazal, J., 2011).

The use of an ICW is now regarded by Northern Ireland Water as a sustainable alternative to traditional

wastewater treatment works and prompted the full-scale development at Stoneyford. There is currently a

lack of understanding of the design principles and performance monitoring relating to the use of ICW’s

for the treatment of domestic wastewater. This research uses data from the full-scale pilot scheme and

small-scale test rig to further knowledge and understanding of design principals and wastewater treatment

performance monitoring. This paper aims to improve the understanding of ICW performance for the

treatment of domestic wastewater in Northern Ireland.




122

Methods

Stoneyford ICW Planning and Design

Stoneyford ICW was proposed by Northern Ireland Water (NIW) to replace an existing Waste Water

Treatment works (WwTW) that had been servicing the village of Stoneyford since 1980. The system was

designed by VESI Environmental and built by BSG Civil Engineering. The development process from

initial planning stages to final completion occurred between August 2013 and July 2016.

The treatment system was designed to treat effluent comprising of both foul effluent and storm water for

a capacity in excess of 950 Population Equivalent (pe) and consists of 2 settlement ponds, 5 treatment

ponds, a control building, site access road, boundary planting and landscaping, and monitoring

equipment. The ICW pond surface area was determined using the DEHLG design guide recommendation

that the area should be between 20m2 – 40m2 per population equivalent (Carty, A., et al., 2008). To

ensure the sustainability of the ICW in terms of future development in Stoneyford Village, the wetland

area designed for the site will provide treatment for a population of 950, with an overall area of

approximately 38,000m2. Figure 1 illustrates the layout drawing of Stoneyford ICW.

Figure 1: Stoneyford ICW Layout Drawing

The operational water depth within each treatment pond is between 150mm - 200mm, with a maximum

depth of 300mm. The pond embankments are sloped with a gradient of 1:1.5 - 1:2 and the upper

embankments are a minimum of 3m wide to allow for access and maintenance. The ponds are connected

using 150mm diameter pipes which are placed on the wetland floor. Water levels can be managed within

each pond using adjustable bends which are placed on the outlet pipe of each pond. The design includes

purpose built concrete weirs to control the water level within each pond.

Planting involved emergent species within each wetland pond and tree species along the embankments

of the site. The ponds were planted with approximately 60,000 emergent wetland species similar to those

used at Glaslough (Dong, Y. et al, (2011; 2013); Kayranli, M. et al, (2010)).



123

Test Rig Planning and Design

The small-scale Test Rig (TR) consisted of 8 test beds. Seven test beds were based on the design

principles of an ICW and one bed was designed on the principles of a Horizontal Sub-Surface Flow

constructed wetland (HSSF). The 7 ICW test beds allowed study of the effects of varying influent volume

similar to the research by Harrington, C., and Scholz, M., (2010). The HSSF allowed comparison with the

ICW.

Consultation with NIW concluded that the best position for the TR would be within the boundaries of Pond

1. This would allow for adequate area for construction of the TR and gravitation inlet flow from the sludge

ponds. This would not disrupt treatment performance of the rest of the ICW as the outlet from the TR test

beds would flow into Pond 1 to continue with treatment as normal. This also ensured that the original ICW

boundaries remained and no additional land take was required.

The design for the TR is based on principles of being able to test 3 different surface areas per person

equivalent (SA/pe) of 20m2, 30m2 and 40m2 against various water depths (Dw) between 50mm and

250mm. Previous studies had suggested that the plant beds should be constructed at a Width: Length

ratio (W: L) of 1:2 in order to gain optimum hydraulic retention and influent mixing (Carty, A., et al, (2008);

Scholz, M., et al (2007)). In order to test the effect of surface area (SA) and Dw on retention time the W:

L ratio for each of the test beds should remain constant and sized accordingly.

In order to test the recommended design SA of 20-40m2 /pe it was important that there was a test bed to

represent each of the sizes 20m2/pe, 30m2/pe and 40m2/pe. The scaling of each test bed was increased

to a SA of 2pe giving the test beds SA’s of 40m2, 60m2 and 80m2. The W: L ratio was maintained at 1:2.

The dimensions of each test bed are shown in Table 1.

Table 1: Individual Test Bed Dimensions

Test Bed Test Bed Number Surface Area Test Bed Width Test Bed Length

ICW 1, 7 40m2 4.475m 8.95m

2, 6 60m2 5.47m 10.95m

3, 4, 5 80m2 6.32m 12.65m

HSSF 8 10m2 2.24m 4.47m

The design for the HSSF Test Bed is based on a simple 1 pond system at the recommended surface area

of 5m2/pe (Vymazal, J., 2005). In order to have the same scaling factor as the ICW Test Ponds of 2pe to

allow for appropriate comparison to be made, the total surface area of the HSSF is 10m2. The W: L ratio

is also recommended at 1:2 giving the HSSF dimensions of 2.24m x 4.47m.

The design drawing of the TR layout is shown in Figure 2. The influent from the sludge ponds is divided

equally across all test beds using a 10 way splitter chamber (Figure 3). Chamber 3 to 10 allows each of

the 8 test beds to gain equal flow and volumes of wastewater. Chambers 1 and 2 were bypassed directly

into Pond 1.



124

Figure 2: Test Rig Layout within Pond 1

Figure 3: The 10 Way Splitter Chamber

The rate of influent flowing into the constructed wetland is dependent upon the varying rate of usage

within the local community and cannot be altered within a live treatment system. The volume of the test

beds would be varied by way of a combination of surface area (SA) and water depth (Dw). The proposed

depth of the Test Rig 7 ICW beds are a total of 600mm which allowed for 200mm soil depth for planting

and up to 300mm water depth for testing, with a 100mm surplus to prevent overflow.

The water depth of T8 was varied by altering the level of the outlet pipe. The different levels of Dw

investigated in the HSSF bed can be seen in Table 2. The depth of T8 was 500mm. The soil: crushed

rock ratio was 300mm: 200mm which allows adequate depths of soil for plant rooting as well as enough

crushed rock depth to prevent clogging. All 8 test beds were planted with Glyceria maxima to allow for

faster establishment of the test beds so testing could begin sooner (Harrington, C. et al, 2011). Figure 4

shows the completed test rig.



125

Table 2: Test Rig HSSF Water Levels

Water Depth (mm) Soil/ Crushed Rock

Depth (mm)

Difference of water level

from soil surface (mm)

300 500 -200

400 500 -100

500 500 0

Figure 4: Test Rig Completed and ready for Testing

Data Collection

Grab samples from the inlet and outlet of each of the 5 full-scale ICW ponds and 8 small-scale test beds

were taken on a weekly basis by NIW and tested for BOD, COD, Total Suspended Solids, Total Nitrogen,

Total Phosphorus, Ammonium, pH and E-Coli. Analysis of the grab samples were made available by NIW

for a 19 month period from January 2016 to July 2017.The Stoneyford ICW weather station continuously

monitored air temperature, precipitation, wind speed, wind direction and humidity. This was supplemented

with weather data recorded at the nearby Aldergrove Airport from the MetOffice archives.



126

Results from Stoneyford Full-Scale ICW

Figure 6: Comparison of ICW inlet contaminants over time

Figure 7: Comparison of ICW outlet contaminants over time.

0

500

BO

D

(mg/

l) Inlet

WOC

0

500

SS

(mg/

l) Inlet

WOC

0

50

NH

3-N

(m

g/l) Inlet

WOC

Jan 16

Feb 16

Mar 16

Apr 16

May 16

Jun 16

Jul 16

Aug 16

Sep 16

Oct 16

Nov 16

Dec 16

Jan 17

Feb 17

Mar 17

Apr 17

May 17

Jun 17

0

1000

Sample DateCO

D (

mg/

l)

Inlet

0

50

BO

D

(mg/

l) Outlet

WOC

0

150

SS

(mg/

l)

Outlet

WOC

0

15

NH

3-

N

(mg/

l) Outlet

WOC

Jan 16

Feb 16

Mar 16

Apr 16

May 16

Jun 16

Jul 16

Aug 16

Sep 16

Oct 16

Nov 16

Dec 16

Jan 17

Feb 17

Mar 17

Apr 17

May 17

Jun 17

0

300

Sample Date

CO

D(m

g/l)

Outlet



127

Comparison of ICW inlet and outlet analysis

Comparison of the inlet and outlet water quality data shows how the 5 ponds of the ICW are capable of

treating of wastewater. Figure 6 compares the Inlet contaminant levels over the sample period. Figure 7

compares the Outlet contaminant levels over the sample period. Figure 6 shows that Inlet levels of all

contaminants rapidly vary over a wide range and are all above their Waste Order Consent (WOC) levels.

COD did not have a WOC.

Figure 7 illustrates that levels of all contaminants are much lower than Inlet levels with a narrower range

of fluctuation. BOD levels are consistently below WOC of 15mg/l with only 1 sample of 40mg/l in

November. Suspended Solids are also typically below WOC of 25mg/l. High levels of suspended solids

above 30mg/l were recorded in June, August and December 2016 and July 2017. Outlet levels of COD

are more stable than Inlet levels across the sample period.

Ammonia levels are below the WOC of 3mg/l for the majority of the sample period. However, consent

was not met between the months of August 2016 and March 2017, and again in July 2017. Overall,

evidence illustrates that Stoneyford ICW can successfully treat wastewater for BOD, Suspended Solids,

and COD during the first 19 months of its life. More research is needed to improve the early performance

of the ICW in treating Ammonia.

Water Quality between Ponds/Area Analysis

Water quality data for each of the 5 ICW ponds over the sample period of 19 months were compared to

determine if there are differences between the performances of each pond. Plotting water quality data

against time is an inappropriate way to illustrate how each pond performs over time. In order to determine

the performance of each of the ponds over time, hydraulic retention time must be considered. To

demonstrate the effects of retention time and performance of each pond, the treatment performance of

ammonia will be discussed in more detail.

Figure 8 illustrates the various ammonia levels (measured as Ammoniacal nitrogen (NH3-N)) per pond

across the testing period of 19 months. Inlet levels fluctuate throughout the year, particularly between the

months of May to November reaching highs of over 47mg/l. Pond 1 shows more stable levels, although

still follows a similar pattern of gradually increasing levels until November. Pond 2 begins stable, before

dropping rapidly to 0.06mg/l by May. However, levels rapidly increase to 14.94mg/l before the end of

June, and remain high for the remainder of the study.

Ponds 3, 4 and 5 show similar trends with NH3-N levels remaining below the WOC for the first 6 months

of testing; however, in June Pond 3 begins to steadily increase and Ponds 4 and 5 follow a similar pattern

by the end of June/ beginning of July. The 3 ponds continue to increase steadily until January 2017 where

they decrease below the WOC by the end of March 2017. Pond 3 then records another spike of 4.27mg/l

at the beginning of April and continues to rise in a similar trend to the previous year with Ponds 4 and 5

continuing to follow by June.

Ponds 3, 4 and 5 follow an ‘n’ shaped curve, seemingly beginning around spring each year. It could be

suggested that their performance is cyclical or seasonal. It may be suggested that it could take a period

of time longer than 2 years for the ponds to reach a state of equilibrium, or it may be that winter is a period

of low performance in ammonia removal.

It is apparent from Figure 8 that there is a period of retention between the ammonia levels reaching above

WOC levels in Pond 3 in July, Pond 4 in August and Pond 5 in October. It is also apparent that the period

of time taken to reduce ammonia levels back below WOC reduces over time. Pond 3 takes 9 months,

Pond 4 takes 8 months and Pond 5 only 5 months. This evidence would suggest that having more than



128

5 ponds could reduce the levels of ammonia at the discharge point further and mitigate the effects of high

fluctuations caused in earlier ponds.

Figure 8: NH3-N Levels per Pond over Time

The results of the total contaminant levels within each pond were plotted against the number of ponds.

Using this data an ‘area under the line’ graph was plotted for each of the contaminants as shown in Figure

9. This illustrates the correlation between the number of ponds and the amount of contaminant within the

pond. From this it can be seen that the treatment of wastewater within the system improves with the

number of ponds. The results have indicated that there is evidence to suggest that each of the ponds

performed differently in treating ammonia with Ponds 1 and 2 typically less stable than Ponds 3-5. Results

also indicate that retention time and wastewater concentration within ponds must be considered to fully

analyse the treatment performance of each individual pond.

0

50

NH

3-N

(m

g/l) Inlet

WOC

0

50

NH

3-N

(m

g/l) Pond 1

WOC

0

50

NH

3-N

(mg/

l) Pond 2

WOC

0

50

NH

3-N

(m

g/l) Pond 3

WOC

0

50

NH

3-N

(m

g/l) Pond 4

WOC

0

50

Sample Date

NH

3-N

(mg/

l)

Pond 5

WOC



129

Figure 9: Area under line graphs for each contaminant plotted against number of ponds.

Water Quality against Seasonal Variations

Water quality performance results from the ICW were plotted against seasonal variations of precipitation,

air temperature, wind speed and humidity over the sample period of one year. This comparison was

completed to determine if external weather conditions have a direct influence on wetland performance.

Hourly weather data was obtained for the yearly testing period and was then combined into daily and

weekly averages for the week prior to each of the sample dates. This gave the representative weather

conditions for the weeks that each of the samples taken. The trend line equation and R2 values for each

of the contaminants against the weather conditions are displayed in Table 3 to illustrate the significance

in correlation between contaminants and weather.

0

20000B

OD

Are

a u

nd

er L

ine

0

35000

SS A

rea

un

der

Li

ne

0

2000

NH

3-N

A

rea

un

der

Li

ne

0

30000

0 1 2 3 4 5 6

CO

D A

rea

un

der

Lin

e

Number of Ponds



130

Table 3: Significance of Weather Variables against Discharge Water Quality

Contaminant Weather Variable Equation R2 Value

BOD Total Precipitation y = -0.0492x + 5.346 0.0198

Air Temperature y = -0.0535x + 4.8292 0.0015

Wind speed y = -1.3294x + 9.7886 0.0564

Humidity y = 0.1023x - 4.2565 0.0069

Suspended Solids Total Precipitation y = -0.1551x + 17.176 0.0141

Air Temperature y = -0.0265x + 14.191 2E-05

Wind speed y = -6.943x + 42.433 0.1108

Humidity y = 0.0963x + 5.8788 0.0004

Ammonia (NH3-N) Total Precipitation y = -0.0525x + 3.6051 0.0514

Air Temperature y = -0.2354x + 4.6588 0.0653

Wind speed y = -0.401x + 4.1296 0.0117

Humidity y = 0.4224x - 32.934 0.2787

COD Total Precipitation y = -0.3959x + 75.035 0.0197

Air Temperature y = 5.2603x + 16.365 0.2123

Wind speed y = -15.849x + 131.81 0.1238

Humidity y = -0.7092x + 126.11 0.0049

Results demonstrate that the relationships between weather conditions and each of the contaminants

show little to no statistical significance. This suggests that weather and seasonal variations have little to

no impact on the ICW ability to treat wastewater. This is contradictory to the previous results which

suggest seasonal trends relating to treatment performance. It should be noted however that these results

do not account for time, maturity or other external factors such as plant growth, influent concentration,

site disturbances or flow changes which may be impacted by climatic conditions.

Results from Stoneyford Small-Scale Test Rig

Water Quality over Surface Area

Results in relation to the water quality performance of each of the different surface areas of 40m2, 60m2

and 80m2 are plotted in Figures 10 - 13 in order to determine if the design rule of thumb of 20m2 – 40m2pe

is appropriate. The test beds are split into 2 groups; 3 of which are set at a water depth of 50mm and 3

are set at a water depth of 250mm. This allowed for a comparison of each surface area within different

water depths to allow for a better understanding of the impacts of volume on performance.

Figures 10 – 13 shows the water quality performance comparisons between the different surface areas

at different water depths. Results show that surface area has a small impact on the wetlands ability to

treat wastewater, with the larger areas being slightly more effective; however this change between 60m2

and 80m2 is very slight when at higher water depths.



131

Figure 10: BOD surface area comparison at 50mm and 250mm water depth

Figure 11: SS surface area comparison at 50mm and 250mm water depth

0

500B

OD

Lev

el (

mg/

l)50mm water depth

Inlet

T1 40m2

T2 60m2

T3 80m2

0

500

BO

D L

evel

(m

g/l)

Sample Date

250mm Water Depth

Inlet

T4 80m2

T6 60m2

T7 40m2

0

400

SS L

evel

(m

g/l)

50mm Water Depth

Inlet

T1 40m2

T2 60m2

T3 80m2

0

400

SS L

evel

(m

g/l)

Sample Date

250mm Water Depth

Inlet

T4 80m2

T6 60m2

T7 40m2



132

Figure 12: NH3-N surface area comparison at 50mm and 250mm water depth

Figure 13: COD surface area comparison at 50mm water depth

0

60

NH

3-N

Lev

el (

mg/

l)50mm Water Depth

Inlet

T1 40m2

T2 60m2

T3 80m2

0

60

NH

3-N

Lev

el (

mg/

l)

Sample Date

250mm Water Depth

Inlet

T4 80m2

T6 60m2

T7 40m2

0

3000

CO

D L

evel

(m

g/l)

50mm Water Depth

Inlet

T1 40m2

T2 60m2

T3 80m2

0

3000

CO

D L

evel

(m

g/l)

Sample Date

250mm Water Depth

Inlet

T4 80m2

T6 60m2

T7 40m2



133

Water Quality over Water Depth Comparisons

Results in relation to the water quality performance of each of the different water depths of 50mm, 150mm

and 250mm are demonstrated in Figure 14. This data helps determine if the design principles of having

no less than 50mm of water and no more than 250mm of water depth were appropriate. The performance

for this section was based on the performance of each of the different water depths of the three 80m2 test

beds in treating the four main contaminants.

Figure 14: Water depth comparisons at 80m2 surface area

For BOD performance the test beds show similar patterns of fluctuations across the 9 month test period.

T3 typically showed the lowest figures across the testing, whereas T4 and T5 showed similarly higher

0

50

100

150

200

250

300

BO

D (

mg/

l) Inlet

T3 50mm

T4 250mm

T5 150mm

0

300

600

900

1200

1500

SS (

mg/

l)

Inlet

T3 50mm

T4 250mm

T5 150mm

0

15

30

45

60

75

NH

3-N

(m

g/l) Inlet

T3 50mm

T4 250mm

T5 150mm

0

200

400

600

800

1000

CO

D (

mg/

l)

Sample Date

Inlet

T3 50mm

T4 250mm

T5 150mm



134

levels. This would suggest that at a surface area of 40m2 pe, the shallower water depth of 50mm was

more effective in removing BOD. SS performance results appear more stable across the 9 month test

period with the exception of T3, which had notable high levels at the beginning of testing. T3 again tends

to show typically lower figures across the testing period, whereas T4 and T5 showed similar patterns that

were slightly higher in levels. This would suggest that at a surface area of 40m2 pe, the shallower water

depth of 50mm was more effective in removing SS.

In terms of ammonia performance, all test beds again show similar patterns of fluctuations across the 9

month test period; however, there is no water depth that performs consistently lower than the others,

which would suggest that water depth has little to no impact on NH3-N removal at this scale. T3 and T4

showed similar performances, with T4 showing higher levels at the beginning of the study and T3 showing

higher levels at the end. T5, despite showing similar patterns, was more consistent in results suggesting

that a water depth of 150mm could be the most effective in treating NH3-N at an area of 40m2 pe in the

long term. Results however were not significantly different than the other water depths.

HSSF Test Bed Performance against Water Depth Changes

Water quality data from the HSSF bed under the different water levels were plotted to determine how it

performs under each of the different ratios of soil to water. This was achieved by changing the water

depth between surface level 0mm and 200mm beneath the surface across the period of 9 months. The

results from each of the contaminants are displayed on one graph with water depth displayed on a

second axis as shown in Figure 15.

Figure 15: Water depth against water quality within a HSSF test bed

Figure 15 illustrates the impacts of changing water depth on the four main contaminants of BOD,

suspended solids (SS), Ammonia (NH3-N) and COD over the test period of 9 months. BOD levels

fluctuated consistently over the 9 month period, but showed no real relationship to changes in water depth

at any stage. Suspended solids on the other hand showed consistent changes in relation to water depth;

-250

-200

-150

-100

-50

0

50

0

500

1000

1500

2000

2500W

ate

r D

epth

(m

m)

Sample Date

BO

D L

evel

(m

g/l)

BOD

SS

NH3-N

COD

Water Depth



135

as the water level decreased below the surface of the HSSF, suspended solids increased and when water

levels rose again, suspended solids levels decreased.

This would suggest that having a higher water level within the HSSF is more effective in treating

suspended solids than having a lower water level. Ammonia levels within the HSSF remained

consistent throughout the study and showed no reaction to changing water depths within the HSSF.

Conversely, COD reacted much like suspended solids and increased greatly when water levels were

decreased. And decreased when water levels rose again. Thus, this would again suggest that having a

higher water level within the HSSF is more effective in treating COD than having a lower water level.

Discussion

The results from the full-scale ICW indicate that Ammonia and COD performances are impacted over

time, while BOD and SS are not. Figure 8 highlights that it is important to consider retention time when

comparing the treatment of each pond and that the theoretical 90 day retention time is not always

practical. Figure 9 represents how water quality improves with increasing number of ponds. Results from

one year of sampling illustrate that air temperature, precipitation, wind speed and humidity have no

significant correlation with ICW water treatment.

The results for the small-scale test rig illustrate the performance of constructed wetlands in relation to

water quality against the key design variables of surface area, water depth, and wetland type. Results

show that a difference in surface area between the rule of thumb of 20m2 – 40m2 /pe has little impact on

the wetlands ability to treat wastewater. The larger surface areas were more effective than the smaller

surface areas, however the impact on water quality performance was not significant at this scale.

With regards to interrogating the water depth rule of thumb, evidence illustrated that a change in water

depth between 50mm – 250mm had little impact on the wetlands ability to treat wastewater. However, it

was emphasised that the shallower water depth of 50mm was more effective than the deeper ponds in

terms of reduction of levels and consistency over time.

The test rig allowed for the investigation into the impacts of decreasing the water level of a HSSF from

surface level 0mm to 200mm beneath the surface. Evidence demonstrates that changing the water level

from 0mm to 200mm beneath the surface had little effect on BOD or ammonia treatment but there was a

significant correlation between water levels within the HSSF and the levels of suspended solids and COD.

This result allows for the recommendation that keeping the water level as close to the surface as possible

would provide the most effective treatment of domestic wastewater in HSSF systems in Northern Ireland.

Results from both the full scale and small scale systems emphasise that current design parameters

could be made narrower, with the importance of having a larger surface area and shallower water depth

being more appropriately explained.

Conclusion

This research is considered the early life performance of an ICW in domestic wastewater treatment. An

ICW by nature is a holistic concept, reliant on many different factors ranging from design, through to daily

operation. The Stoneyford ICW has identified how all of these factors need to be considered in this holistic

concept. It is concluded that all variables need to be considered to ensure successful operation of the

ICW.

There is limited experience with the use of ICWs for the treatment of domestic wastewater with most of

the design principles based on agricultural wastes. Stoneyford is the first full-scale ICW for the treatment



136

of domestic wastewater in Northern Ireland and allowed for observations to be made which could be used

to optimise the design and decision making process for future installations.

ICW’s are a viable alternative to traditional wastewater treatment works in treating domestic sewage in

Northern Ireland. Stoneyford ICW has shown successful wastewater treatment performances in its early

years and evidence suggests that performance will improve as the wetland matures. Using a full-scale

trial system has brought substantial benefits to the wastewater treatment industry in terms of knowledge

and understanding, providing an excellent example from which to learn and improve future developments.

Acknowledgements

This research project has been supported through the Department of Employment and Learning (DEL)

Co-operative Awards in Science and Technology (CAST) scheme with direct involvement with Northern

Ireland Water.

References

Carty, A., Scholz, M., Heal, K., Gouriveau, F. and Mustafa, A. (2008) The universal design, operation and maintenance guidelines for farm constructed wetlands (FCW) in temperate climates. Bioresource technology, 99 (15), 6780-6792.

Dong, Y., Kayranli, B., Scholz, M. and Harrington, R. (2013) Nutrient release from integrated constructed wetlands sediment receiving farmyard run-off and domestic wastewater. Water and Environment Journal, 27 (4), 439-452.

Dong, Y., Scholz, M. and Harrington, R. (2012) Statistical modeling of contaminants removal in mature integrated constructed wetland sediments. Journal of Environmental Engineering (United States), 138 (10), 1009-1017.

Harrington, C. and Scholz, M. (2010) Assessment of pre-digested piggery wastewater treatment operations with surface flow integrated constructed wetland systems. Bioresource technology, 101 (20), 7713-7723.

Harrington, C., Scholz, M., Culleton, N. and Lawlor, P.G., (2011) Meso-scale systems used for the examination of different integrated constructed wetland operations. Journal of Environmental Science and Health - Part A Toxic/Hazardous Substances and Environmental Engineering, 46(7), pp. 783-788.

Kayranli, B., Scholz, M., Mustafa, A., Hofmann, O. and Harrington, R. (2010) Performance evaluation of integrated constructed wetlands treating domestic wastewater. Water, air, and soil pollution, 210 (1-4), 435-451.

Moshiri, G. A. (1993) Constructed Wetlands for Water Quality Improvement. CRC Press.

Scholz, M., Harrington, R., Carroll, P. and Mustafa, A. (2007) The integrated constructed wetlands (ICW) concept. Wetlands, 27 (2), 337-354.

Vymazal, J. (2005) Horizontal sub-surface flow and hybrid constructed wetlands systems for wastewater treatment. Ecological Engineering, 25 (5), 478-490.

Vymazal, J. (2011) Constructed wetlands for wastewater treatment: Five decades of experience. Environmental Science and Technology, 45 (1), 61-69.



137

ON SITE LANDFILL LEACHATE TREATMENT: INVESTIGATIONS INTO

ECONOMICAL AND ENVIRONMENTAL SUSTAINABLE SYSTEMS FOR

NORTHERN IRELAND

Devlin, Y.1, Nicholl, G.1, McRoberts, C.1, Johnston, C.1, Rosinqvist D.2, Svensson B.M.3 and

Mårtensson, L.3 1Agri-Food and Biosciences Institute, UK, 2Laqua Treatment AB, Sweden,

3Kristianstad University, Sweden


Abstract

This paper presents the potential for the Swedish Laqua system to be used as a sustainable method

for on-site landfill leachate management in Northern Ireland, specifically the potential to use locally

sourced filter materials from Northern Ireland as part of the filter system. Four carbon containing ashes

and four types of peat were tested over a 24 hours period by a shaking test with untreated landfill

leachate. Considering the results of this screening test, and the economical and sustainable supply of

filter materials, one combination of ash and peat was selected to be column tested. Column testing with

artificial leachate containing 7 organic pollutants (3 PAHs and 4 PCBs) and 9 inorganic pollutants

showed that locally sourced filter materials effectively removed both organic and inorganic pollutants.

A subsequent column test with landfill leachate for 13 weeks demonstrated it was feasible to apply the

Laqua system with economical locally sourced filter materials.

Keywords

Artificial leachate, Carbon-Containing Ash, Filter treatment, Landfill leachate, Peat, Wastewater

treatment

Introduction

In the 21st century we are becoming an increasingly ‘throw-away society’; consuming produce and

generating significant volumes of waste materials for disposal immediately after use. For example,

many mothers in the UK no longer wash baby nappies, instead using disposable products which are

ultimately destined for landfill. These products are not just made of simple materials such as paper, but

also contain complex materials that include polymers, adhesives, dyes and perfumes. While there is

some public concern about these waste materials entering landfill and polluting the soil, there may be

less concern or knowledge about the dissolved pollutants that may leave the landfill via leachate. Over

time, waste material decomposes producing leachate which has to be managed and treated to prevent

it from entering ground water sources. Without proper treatment of the leachate, it can pollute not only

around landfills, but also rivers, lakes and the ocean.. However, even after a landfill site has been closed

to further waste, the treatment of leachate is often required for decades.In 2015, there were

approximately 30 authorised landfill sites in Northern Ireland which were estimated to handle 1.4 million

tonnes of municipal waste a year (NIEA, 2015). According to Met Office data (Met Office, 2017) rainfall

in Northern Ireland ranges between 800 – 2000 mm per annum, and subsequent ground water

replenishment can result in large volumes of leachate being generated on landfill sites. On-site, the

landfill leachate receives primary treatment to reduce the levels of Biological Oxygen Demand and

Ammonia and adjust the pH/acidity, before being transferred to waste water treatment facilities for

further treatment before final safe discharge to the environment. One Northern Irish landfill site recorded

their leachate volume to be up to 40,000m3 per year, equivalent to transportation of three water tankers




138

per day (110m3 /day). So, it is easy to estimate the high economic costs solely for leachate

transportation, let alone the additional costs of the final leachate treatment. In Sweden, some

commercial landfill sites have introduced an “on-site landfill leachate management” system named

Laqua. The Laqua system and methodology is a simple filter-based system where leachate is filtered

through a mixed layer of specific peats and ashes to remove potential contaminants (Figure 1).

Figure 1: Laqua system: Leachate is filtered through a mixture of layered ashes and

peats

This study investigated the potential for the Swedish Laqua system to be used as a sustainable

method for on-site landfill leachate management in Northern Ireland, specifically the potential to use

locally sourced filter materials from Northern Ireland as part of the filter system.

Materials

A range of filter media resources were obtained from several sources across Northern Ireland and the

Republic of Ireland. A range of peat materials were sourced along with a selection of combustion ashes

from a number of coal-fired power stations; a waste material or by-product. Approximately 200L landfill

leachate, were collected from an authorised landfill site before primary treatment. The leachate was

stored at 4C until its use.

Analysis of leachate

Municipal Solid Waste (MSW) landfill leachate has a very complex composition and usually contains a

large number of organic and inorganic pollutants. In addition, the leachate’s character is influenced by

factors such as rainfall volume, landfill composition and speed of decomposition. A batch of leachate

collected on 5 December 2016 was analysed for various parameters, including organic pollutants and

inorganic pollutants. A range of general properties, including pH, conductivity, ammonium-nitrogen,

nitrite and nitrate nitrogen and a biological toxicity test were also measured.

Table1: General properties of leachate collected on 5 December 2016

A biological toxicity test was carried out using Artemia salina (Svensson, et al, 2005). Artemia salina

has a high tolerance for chloride ions and therefore it is suitable to measure the toxicity of leachate.

Artemia salina was hatched in artificial seawater and 48-52 hours after hatching, was added to five

concentrations of leachate (48%, 71%, 83%, 89% and 95%) leachate. After 24 hours, the immobility of

pHconductivitiy

(S/cm)

NO2 +

NO3-N

(mg/l )

NH4-N

(mg/l )

7.18 1825 3.77 81



139

the Artemia salina was measured and compared with that in artificial sea water. “Immobility” was defined

as no movement for more than 10 seconds. Reliability of this assessment as a quality control was

carried out with 100ppm potassium dichromate as a reference material.

The results are in Table 2. Artemia salina had less influenced on this leachate until 71% leachate

concentration (20l of artificial sea water and atresia, 300l of leachate and 100l of artificial sea water)

but more than 83% volume ratio Artemia salina showed dramatically high immobility. According to

microscopic observation, Artemia salina were stuck into sludges and killed. Therefore leachate for

toxicity test were filtered with 0.7m glass syringe filter to remove sludge.

Table 2: Biological toxicity test of leachate collected on 5 December 2016

Organic pollutants

The leachate consisted of solid sludge and liquid. Extraction with dichloromethane was carried out

filtered and unfiltered leachate to examine sludge’s influence.

They were analysed by a mass spectrometry for a range of organic contaminants, 62 pesticides, 24

polycyclic aromatic hydrocarbons (PAHs) and 7 polychlorinated biphenyls (PCBs). Two pesticides and

14 PAHs were detected in the unfiltered samples whilst lower levels of one of the pesticides and two of

the PAHs were detected in the filtered leachates. (Table 3) This suggests that the majority of the

compounds are associated with the particulate matter (solids / sludge) in the leachate samples.

Inorganic pollutants

Filtered and unfiltered leachate samples were acid digested and measured by ICP-OES (Agilent

Technologies 5100) in order to determine and compare concentrations of inorganic pollutants.

Filtration of the leachate reduced the concentration of many of the elements, including cadmium,

chromium, nickel, lead, copper, zinc and iron; suggesting that these pollutants are associated with the

particulate material similar to the organic pollutants (Table 4).

Leachate volume

ratio in a test

well (%)

Artemia

immobility (%)

0 3

48 1

71 5

83 53

89 87

95 95



140

Table 3: Screening test for pesticides, PCBs and PAHs - positive identifications indicated by a tick. Where a compound was detected in the both filtered and unfiltered a double tick indicated much higher levels were found in that sample

Table 4. ICP analysis on unfiltered leachate and filtered leachate

Filter materials screening (24 hour shaking test)

The initial screening of the potential filter materials involved a 24 hour extraction test with leachate to

assess the potential to reduce contaminants.

Since high carbon-containing ashes (Svensson, B.M. et al, 2005) showed good potential as filter

materials, carbon concentration was measured by combustion analysis (LECO). Fly ash from a coal

power station (Ash 2) was the highest content. Ash 1 and 3 were similar content but sourced from

different combustion power stations. Biochar was added to the screening test to examine the effect of

a high carbon containing product produced via pyrolysis.

Pesticide Unfiltered filtered

Chlorpyriphos PP P

Pendimethalin P

PCBs

PCB 101

PCB 118

PCB 138

PCB 153

PCB 180

PCB 28

PCB 52

PAHs

Naphthalene PP P

Phenanthrene P

Anthracene P P

Fluoranthene P

Pyrene P

Benzo (a) Anthracene P

Chrysene P

Benzo (b) Fluoranthene P

Benzo (k) Fluoranthene P

Benzo (j) Fluoranthene P

Benzo (a) Pyrene P

Indeno (123,cd) Pyrene P

Benzo (ghi) Perylene P

Dibenzo (a,e) Pyrene P

metal Cd Ni Pb Cu Mn Zn Fe K Na Ca

unit µg/l µg/l µg/l µg/l µg/l µg/l mg/l mg/l mg/l mg/l

Unfiltered

leachate 9.46 46.9 16.4 31.6 491 525 73.7 45.8 82.1 130

Filtered

leachateND 13.9 1.16 8.62 444 33.7 0.069 48.7 85.4 129



141

Peats were selected from 4 different mining areas and production processes. All the peats contained

high moisture except for Peat 3 which was a more fibrous peat. The materials are summarised in Table

5.

Peats were selected from 4 different mining areas and production. All the peats contained high moisture

except for Peat 3 which was a fibrous peat. The materials are summarised in Table 5.

Table 5: Ashes and Peats for filter materials screening

Screening test method

The screening test method was a modified version of British standard BS EN 12457-2:2002. Four

individual peats and 4 individual ashes, and 16 peat/ash mixtures were shaken for 24 hours with

untreated leachate at a ratio of 10L/kg. Various mixtures of peat and ash were made up using a peat to

ash ratio of 3:1 (v/v). The individual materials and the mixes (10g) were weighed into 250ml plastic

bottles and shaken with 100mls of leachate using an orbital shaker at 200rpm for 24 hours. The

processed leachates were centrifuged at 3600rpm for 10 minutes then filtered through Whatman No.40

filter paper. A standard mix supplied by Laqua Treatment AB, Sweden, was also included in the

screening test for comparison.

The filtered treated leachate was measured for pH and conductivity. The concentrations of nitrite,nitrate

and ammonium nitrogen were determined by a colorimetric method.

Biological toxicity tests, simplified to 71% and 95% of leachate composition, were also carried out for

all the filtrates.



142

Table 6: 24 hours shaking test results on general properties

Shaking leachate with ash and the biochar material alone effectively removed colour from the leachate

indicating the possible absorbance of certain organics. The Biochar had some effect on reducing the

level of NH4-N while other ashes had little effect on the NH4-N concentration in the leachate. The ashes

also increased the conductivity, possibly due to the high conductive salts or carbons which might have

eluted from ash. The ashes also showed their potential to reduce the toxicity of the leachate.

The biological toxicity results also showed that the peats were less effective than the ashes, despite the

low NH4-N concentrations. However, they were effective in reducing the conductivity and NH4-N

concentration in the leachate (Table 6).

The peat and ash mixtures varied in effectiveness depending on each characteristic. Mixes using peat

4 showed relatively good performance for all the general properties. Ash 3 showed similar

characteristics to the standard Laqua mix. These results show that there is potential for locally sourced

Filter material pHconductivitiy

(S/cm)

NO2 +

NO3-N

(mg/l )

NH4-N

(mg/l )

71%

leachate

immobility

%

95%

leachate

immobility

%

Leachate 7.18 1825 3.77 81.0 69 100

Leachate (method

blank) 7.65 1811 2.30 81.5 21 93

Standard mix 7.55 2930 4.35 61.5 19 96

Ash 1 7.92 1929 1.97 79.5 15 96

Ash 2 7.69 2260 2.38 73.1 35 98

Ash 3 11.65 2780 3.44 67.0 17 100

Biochar 10.07 17760 3.82 30.0 100 100

Peat 1 7.22 875 2.07 52.4 48 100

Peat 2 6.38 550 2.80 25.2 63 99

Peat 3 7.46 730 3.39 40.8 60 100

Peat 4 7.54 602 3.29 31.3 62 100

Ash1+ Peat 1 7.30 1417 1.92 69.3 31 100

Ash1+ Peat 2 7.84 1484 2.02 64.6 25 100

Ash1+ Peat 3 7.62 1438 3.07 66.5 22 95

Ash1+ Peat 4 7.87 1589 2.55 68.9 8 92

Ash 2 + Peat 1 7.23 1346 2.06 66.2 2 97

Ash 2 + Peat 2 7.16 1493 2.33 61.3 16 99

Ash 2 + Peat 3 7.17 1414 2.86 64.2 52 100

Ash 2 + Peat 4 7.61 1641 2.80 67.1 36 97

Ash 3 + Peat 1 7.60 2280 4.14 71.5 10 100

Ash 3 + Peat 2 7.95 2710 4.09 61.3 14 100

Ash 3 + Peat 3 7.84 2410 4.94 66.2 22 100

Ash 3 + Peat 4 8.12 2470 4.22 64.3 15 100

Biochar + Peat 1 8.32 5550 4.00 43.1 100 100

Biochar + Peat 2 8.25 8420 3.71 7.1 100 100

Biochar + Peat 3 8.25 6540 4.51 17.3 100 100

Biochar + Peat 4 7.56 1365 4.82 64.9 12 100

Potential Materials

Potential Mixes



143

materials to act as a filter to reduce/remove contaminants. The peat/ash mixture is crucial for the

filtration system as essentially the peat will remove metal contaminants and the ash will remove the

organic contaminants.

Table 7: 24 hours shaking test results on metal analysis

As shown in Table 7, the concentrations of copper, zinc and iron decreased after filtration.

Concentration of manganese and alkali metals were not changed by paper filtration.

Although the ashes are effective in reducing the concentration of heavy metals such as copper,

manganese, zinc and iron, it is evident that they also elute arsenic and calcium. However, according to

the experience of Laqua treatment AB, the arsenic concentration produced from the ash in the initial

stage of treatment will reduce after a number of weeks as the filter material becomes conditioned.

Because this is only 24 hours shaking test, it was not unexpected to see an increase of arsenic from

young ash material.

The biochar also reduced heavy metal concentrations and although it did not elute arsenic it produced

high concentrations of potassium and sodium. Carbons in ashes and carbon in biochar seemed to have

different effects on leachate treatment.

Peats were effective in reducing alkali and alkali-earth metals, as well as heavy metals.

The mixtures of ashes and peats reduced both heavy metals and alkali, alkali-earth metals.

Filter materialAs

(µg/l)

Cd

(µg/l)

Cr

(µg/l)

Pb

(µg/l)

Cu

(µg/l)

Mn

(µg/l)

Zn

(µg/l)

Fe

(mg/l)

K

(mg/l)

Na

(mg/l)

Ca

(mg/l)

Leachate <5.0 1.50 <50.0 18.4 18.6 473 142 21.6 46.6 80.4 127

Leachate (method

blank) <5.0 <0.5 <50.0 <3.0 8.62 444 33.7 0.069 48.7 85.4 129

Standard mix <5.0 <0.5 <50.0 <3.0 6.48 18.0 13.9 0.010 66.0 85.2 466

Ash 1 5.57 <0.5 <50.0 <3.0 <3.0 77.0 10.0 0.008 56.2 88.0 114

Ash 2 30.4 1.15 <50.0 <3.0 <3.0 116 39.8 ND 64.2 102 301

Ash 3 6.89 <0.5 164 <3.0 <3.0 <15.0 <10.0 0.009 38.9 72.1 434

Biochar <5.0 <0.5 <50.0 <3.0 6.33 76.0 10.9 ND 3008 578 11.0

Peat 1 <5.0 <0.5 <50.0 <3.0 7.13 55.0 <10.0 1.14 27.8 62.7 13.3

Peat 2 <5.0 <0.5 <50.0 <3.0 18.2 <15.0 12.0 1.54 14.7 43.6 1.83

Peat 3 <5.0 <0.5 <50.0 <3.0 8.52 40.0 <10.0 0.644 22.2 55.6 3.69

Peat 4 <5.0 <0.5 92.7 <3.0 16.0 22.0 18.1 1.32 18.0 45.4 1.27

Ash1+ Peat 1 <5.0 <0.5 <50.0 <3.0 6.18 163 14.6 0.676 40.8 77.2 61.9

Ash1+ Peat 2 <5.0 <0.5 <50.0 <3.0 9.32 56.0 11.0 0.468 40.5 82.8 46.0

Ash1+ Peat 3 <5.0 <0.5 <50.0 <3.0 6.24 271 13.3 0.194 42.1 81.6 72.6

Ash1+ Peat 4 5.90 <0.5 <50.0 <3.0 7.13 107 11.4 0.087 44.4 83.4 87.5

Ash 2 + Peat 1 17.2 <0.5 <50.0 <3.0 10.8 197 19.7 0.303 43.2 83.5 74.1

Ash 2 + Peat 2 211 <0.5 <50.0 4.30 16.0 74.0 43.4 0.735 44.2 91.7 35.6

Ash 2 + Peat 3 75.5 <0.5 <50.0 <3.0 8.77 269 15.0 0.285 44.5 86.2 59.1

Ash 2 + Peat 4 134 <0.5 <50.0 <3.0 6.50 155 32.7 0.099 50.6 94.4 91.0

Ash 3 + Peat 1 8.95 <0.5 <50.0 <3.0 <3.0 124 <10.0 0.054 41.3 77.8 197

Ash 3 + Peat 2 40.0 <0.5 <50.0 <3.0 <3.0 99.0 <10.0 0.054 43.1 84.2 199

Ash 3 + Peat 3 21.2 <0.5 <50.0 <3.0 <3.0 <15.0 <10.0 0.030 42.5 79.3 138

Ash 3 + Peat 4 45.3 <0.5 <50.0 <3.0 <3.0 54.0 <10.0 0.051 43.5 82.5 172

Biochar + Peat 1 39.6 <0.5 <50.0 <3.0 20.1 137 41.7 0.420 1388 224 22.7

Biochar + Peat 2 <5.0 <0.5 67.5 <3.0 16.9 113 33.0 0.318 2049 335 14.7

Biochar + Peat 3 12.1 <0.5 7.46 <3.0 36.2 119 29.4 0.075 1580 253 13.2

Biochar + Peat 4 <5.0 <0.5 <50.0 <3.0 15.9 78.0 20.2 0.066 1978 323 12.4

Potential Mixes

Potential Materials



144

As organic pollutants only the PAHs and PCBs were determined for 24 hours shaking test. Briefly, PAHs

and PCBs were extracted from the leachate samples as follows: 100ml aliquots of each treatment were

transferred in duplicate to separating funnels, 50ml aliquots of dichloromethane (DCM) were added to

each separating funnel and the mixture shaken for 1min and allowed to separate before the lower DCM

layer was run off into a conical flask. This was repeated with two further aliquots of DCM which were

combined with the first extraction and dried over anhydrous sodium sulphate. The dried DCM fraction

was filtered through Whatman No 1 filter paper into turbovap tubes and this was concentrated to 0.5ml

under nitrogen at 35˚c.This was transferred with hexane washings to 10ml glass tubes and dried under

nitrogen at room temperature before re-suspension in 1ml hexane for analysis by Agilent 7000 series

GC-MS/MS (PCBs) and Agilent 5973 inert GC-MS (PAHs).

Organic pollutants were all lower than detection limits for all 24 hours shaking samples.

Filter material screening (column test)

The 24 hour screening test, was used to select the combinations of ash and peat for the column tests.

The column test was designed to simulate filtration conditions at a landfill site. Untreated leachate was

filtered through these column, at a flow rate of about 1 litre / day for thirteen weeks, with samples being

collected weekly for analysis.

Test equipment

The columns were 1.2 metre in length with a diameter of 100mm and connected to timer controlled

peristaltic pumps. (Figure 2) The columns were filled to a depth of 100cm with the different filter

materials (homogenous mixes of ash and peat) before being compressed to a depth of approximately

80cm. Leachate was stored in a 90 litre container and mixed well once a day to help keep particulate

matter in suspension. Every seven days, 300 ml filtrate was collected over an 8 hour period for analysis.

Figure 2: Column test equipment designed by Laqua Treatment AB, Sweden

Selection of filter materials

Filter materials for the column test were selected using the results of the 24 hour shaking test and also

the material’s availability for up-scaling in the future.

Peat 4 was selected for a potential filter mix, mainly due to the 24 hour shaking test results (Table 6)

and also because this peat is a readily available commercial grade horticultural peat.



145

The biochar was ruled out as it produced high concentrations of potassium and sodium, and high

biological toxicity.

All the ashes tested produced similar results from the 24 hour shaking test.

Finally Ash 3 and Peat 4 were selected as a filter mixture and prepared in a ratio of 1 to 4 (v/v).

Column 1 was prepared using the standard mix supplied by Laqua Treatment. AB.

Column 2 and Column 3 were prepared using the locally sourced ash and peat mixture (Ash 3 and Peat

4). Columns 1 and 2 were then treated with landfill leachate and Column 3 was treated with an artificial

leachate.

Artificial Leachate

In order to investigate the performance of the filter material quantitatively, an artificial leachate was

designed and prepared to include a mixture of standard organic and inorganic pollutants at typical levels

in landfill leachate. The artificial leachate was prepared (Table 8) to include nine inorganic and seven

organic pollutants (PAHs and PCBs) and was then filtered through column 3 instead of actual leachate

from a landfill site for a period of 4 weeks.

Table 8: Artificial leachate composition

Results of the treatment of artificial leachate with local filter materials

As the artificial leachate was prepared using 0.5M nitric acid standard solution, the pH was much lower

than the landfill leachate. The Artemia salina used for the biological toxicity test is very vulnerable to

acidic conditions so it was not possible to perform this test. (JOHN MONASH science school, 2014)

However, after treatment through the column, the pH was almost neutral at around pH7 with no impact

on the Artemia salina. General properties are summarised in Table 9.



146

Table 9: General properties of artificial leachate treatment

Figure 3 shows the effective reduction of the 6 heavy metal pollutants by the column filtration.

Concentration of five elements; Cadmium, lead, chromium, nickel and iron were all reduced within the

first week of treatment to below the detection limits of the instrumentation. Copper, zinc and manganese

decreased over time (Fig 3 (a & b)). It was suggested that the filter materials would take some weeks

to condition and perform well. However, the results clearly indicate that local filter materials have real

potential to remove heavy metals. After an initial flush of arsenic from the filter material the concentration

levels seem to be gradually decreasing after the second week.

Figure 3. Inorganic pollutants concentration variation with column treatment time

(a) Arsenic, Cadmium, Lead and Copper

(b) Zinc and Manganese

In all period the amount in the column effluent was <1% that of the artificial leachate itself showing

that the column 3 is capable of removing all seven contaminants in solution.



147

These results clearly demonstrate that the columns produced from locally sourced material offer

potential for the efficient removal of organic contaminants from leachate.

Results of treatment of landfill leachate with local filter materials

Locally sourced filter materials used in column 2 were tested with landfill leachate in comparison with

the standard mix used in column 1. Initially the column test was set up to run for 4 weeks but

performance at week 7 showed a dramatic change in the concentration of ammonium-nitrogen and that

of nitrite/nitrate-nitrogen and improved biological toxicity test results, therefore the running time of the

columns was extended in order to examine this effect further

Table 10 shows the pH, conductivity and ammonium nitrogen and nitrite/nitrate nitrogen concentrations

from week 2 to week 13. Between week 4 and week 7, treated leachate from column 2 filled with the

local ash and peat mixture showed decreasing ammonium-nitrogen concentration while the

nitrite/nitrate nitrogen concentration was increasing. This same phenomena was continuing at week 13.

This suggests that the filter mixture is now behaving similarly to the column with the standard mixture,

effectively nitrifying ammonium nitrogen to nitrite/nitrate nitrogen.

Table 10: Properties of treated leachate: pH, conductivity and concentration of

ammonium nitrogen and nitrate / nitrite nitrogen

Results from the biological toxicity test are summarised in Table 11 The mobility of the Artemia salina

was dramatically improved between week 4 and week 7 and this continued through to week13. After

week 7, the toxicity was improved even at 95% leachate concentration (400l leachate and 20l artificial

sea water and 20-30 number of Artemia salina). Even 24 hours after addition of the column treated

leachate, artemia could still be observed swimming strongly within the cells of the microplate indicating

a more favourable, less toxic environment as a result of the column filtration.

Leachate

before

column

Column1

Sandard mix

Column 2

Local mix

Leachate

before

column

Column1

Sandard

mix

Column 2

Local mix

pH 7.39 6.74 7.62 7.34 6.82 7.54

conductibity

(mS/cm)2.28 2.21 2.05 1.85 1.80 1.62

N-NH4 (mg/L) 119.79 3.37 57.99 87.61 6.17 85.54

N-NO2&NO3

(mg/L)0.00 107.97 1.86 2.16 95.77 10.20

Leachate

before

column

Column1

Sandard mix

Column 2

Local mix

Leachate

before

column

Column1

Sandard

mix

Column 2

Local mix

pH 7.08 6.87 7.8 7.83 7.43 7.52

conductibity

(mS/cm)2.03 2.17 1.87 2.70 2.52 2.21

N-NH4 (mg/L) 152.05 6.77 7.25 139.77 2.35 0.24

N-NO2&NO3

(mg/L)2.10 95.11 100.47 0.00 165.66 172.64

Week2 Week4

Week7 Week13



148

Table 11: Biological toxicity test from week 2 to week 13

After column treatment, five heavy metals: nickel, iron, lead, chromium and cadmium were effectively

removed from the leachate to below detection limits of the instrumentation. Figure 4 shows the results

from week1 to week7 for these 5 elements. It was observed that arsenic was released from the column

materials with concentrations peaking at week 2 before decreasing gradually with the concentration at

week 7 having declined to 28g/l, similar to the standard mix. Column 2 also reduced manganese (c)

and zinc (d) effectively. Filter materials seemed to release copper, too. But it decreased gradually

removed from filter materials.

Figure 4: Changes in Inorganic pollutant concentrations

(a) Arsenic

(b) Lead

Leachate

before

column

Column1

Sandard mix

Column 2

Local mix

Leachate

before

column

Column1

Sandard

mix

Column 2

Local mix

0% leachate 3 1 5 0 0 3

48% leachate 3 0 2 0 0 2

71% leachate 11 1 0 5 0 4

83% leachate 23 0 2 42 0 26

89% leachate 61 1 21 65 2 73

95% leachate 85 42 71 98 22 93

Leachate

before

column

Column1

Sandard mix

Column 2

Local mix

Leachate

before

column

Column1

Sandard

mix

Column 2

Local mix

0% leachate 0 2 1 3 0 2

48% leachate 0 0 0 0 3 1

71% leachate 7 0 3 29 0 0

83% leachate 58 2 3 82 3 2

89% leachate 77 14 2 87 10 3

95% leachate 100 23 8 100 34 4

% Artemia

immobility

% Artemia

immobility

Week2 Week4

Week7 Week13



149

(c) Manganese

(d) Zinc

(e) Copper

All organic pollutants were below the detection limit as standard untreated leachate were analysed for

the presence of 21 PAHs and 7 PCBs prior to column filtration.

Conclusion

This study only ran for 6 months in order to try to determine the potential of implementing the Laqua

system for the on-site management of landfill leachate in Northern Ireland. There is strong evidence

from 24 hour shaking tests followed by controlled column filtration tests, that locally sourced materials

have the potential to function effectively and similarly to the tried and tested standard Laqua systems

filtration media. In order to build on-site landfill leachate treatment filters, a scaled up quantity of filter

material will be required and indeed recharged after a number of year’s operation hence the importance

of the local and sustainable sourcing filter media materials.

It is apparent that some prefiltration filter conditioning is required in order to remove inorganic pollutants

(copper and arsenic) as well as to activate the microbiological populations for eg ammonium nitrification.

This study has indicated that 4-5 weeks will achieve this.

Only one combination of ash and peat was examined by the column test in this initial study but other

combinations showed similar results from the 24 hours shaking experiments. Further column testing

will be undertaken to investigate these identified peat/ash mixtures. Additionally, an on-site pilot trial

using 1m3 cipax containers is being implemented in order to test the durability and real-world

functionality of a scale up. It is considered that if the column experiment results can be replicated in

these, then a full-scale pilot at a landfill site would be a plausible solution for sustainable leachate

management.



150

Acknowledgements

The Authors acknowledge the financial support of Northern Ireland Environmental Agency (NIEA),

Government challenges Ideas from business Innovate solutions (SBRI) and InnovateUK and Strategic

Investment Board.

References

nidrect government services, 2017. Available from: https://www.opendatani.gov.uk/dataset/niea-

authorised-landfill-sites/ [accessed 7 August 2017]

British Standards Institution (2002) Characterisation of waste - Leaching – Compliance test for

leaching of granular waste materials and sludges, BS EN 12457-2:2002

MET Office, 2017. Available from:

http://www.metoffice.gov.ukhttp://www.metoffice.gov.uk/climate/uk/regional-climates/ni#rainfall

[accessed 14 August 2017]

Laqua Treatment AB Sweden, 2017. Available from:

http://www.laqua.se/index_en.php?m=fs_intro&p

Svensson,B.M., Mathiasson, L.,Mårtensson and Bergström, S.(2005) Artemia salina As test organism

for assessment of acute toxicity of leachate water from landfills Environmental Monitoring and

assessment 102, 309-321.

Svensson,B.M., Mathiasson, L.,Mårtensson and Kängsepp,P. (2011) Evaluation of filter material for

treatment of different Types of Wastewater Journal of Environmental Protection 2,888-894.

JOHN MONASH science school , 2014. Available from:

https://sites.google.com/a/jmss.vic.edu.au/scifair2014-web0004/conclusions-and-references

[accessed 15 August 2017

https://www.opendatani.gov.uk/dataset/niea-authorised-landfill-sites/

https://www.opendatani.gov.uk/dataset/niea-authorised-landfill-sites/

http://www.metoffice.gov.ukhttp/www.metoffice.gov.uk/climate/uk/regional-climates/ni#rainfall

https://sites.google.com/a/jmss.vic.edu.au/scifair2014-web0004/conclusions-and-references

Documents

EUROPEAN WASTE WATER MANAGEMENT - AquaEnviro · PRACTICAL APPLICATION OF MODULAR OFF-SITE BUILD: A COMMISSIONING PERSPECTIVE 65 Baird, A., WPL Ltd, UK LIVERPOOL WWTW SBR CARBONACEOUS