GHG PERFORMANCE JATROPHA BIODIESEL · 2009-04-28 · Jatropha it will be set conservatively and this project enables (buyers of) D1’s Jatropha oil/biodiesel to report the better

Ecofys bv

P.O. Box 8408

NL-3503 RK Utrecht

Kanaalweg 16-G

NL-3526 KL Utrecht

The Netherlands

www.ecofys.nl

tel +31 (0)30 280 83 00

fax +31 (0)30 280 83 01

e-mail [email protected]

Bart Dehue

Willem Hettinga

2 June 2008

Copyright Ecofys 2008

Ecofys reference: PBIONL073010

Commissioned by: D1 Oils

GHG PERFORMANCE JATROPHA BIODIESEL

GHG-PERFORMANCE JATROPHA BIODIESEL II

Acknowledgements

The authors are thankful to all the Indian team members for demonstrating to us the real

life plantations, providing the required data, their hospitality and openness during the field

trip. Special thanks go to Suresh and Sreenivas for their professional input and pleasant

company. In addition we would like to thank Neil Judd and Kathleen Bottriell from

Proforest for their open cooperation during the project.

GHG-PERFORMANCE JATROPHA BIODIESEL III

Content

1 Introduct ion 1

1.1 Why GHG-performance is important 1

1.2 RTFO default for Jatropha 2

1.3 Overview Jatropha biodiesel supply chain 2

1.4 Structure of the report 3

2 GHG-performance: Base Case 6

2.1 Base Case: Jatropha biodiesel reduces GHG-emissions by 66% to 68% 6

2.2 Base case description and assumptions 8

3 Comparison of Jatropha to other crops 14

3.1 Jatropha biodiesel compared to RTFO defaults 14

3.2 Jatropha biodiesel compared to EC defaults 15

3.3 Understanding the differences 16

4 Impacts of Land Use Change 18

4.1 Using RTFO or EC default values for Land Use Change 18

4.2 IPCC Tier 1 approach for LUC: the effect of climate zones 20

4.3 Using site specific data for Land Use Change 21

5 Opportunit ies to improve GHG-performance 22

5.1 Changes to the production system 22

5.2 Sensitivity analysis 27

5.3 Worst and best case scenario 30

6 Conclusions and discussion 32

6.1 GHG-performance Jatropha versus other crops 32

6.2 Performance under RTFO versus EC methodology 32

6.3 Areas for improvement 33

6.4 Land Use Change 33

GHG-PERFORMANCE JATROPHA BIODIESEL IV

References 35

Annex A Key dif ferences between the RTFO and

EC proposal 36

Annex B Jatropha yield 39

Annex C IPCC defini t ions 40

Annex D LUC calculat ions IPCC Tier 1 43

GHG-PERFORMANCE JATROPHA BIODIESEL 1

1 Introduction

1.1 Why GHG-per formance i s important

Sustainability in production is an important issue for D1 and D1 aims to maximize the

greenhouse gas emission savings of its Jatropha oil/biodiesel. Besides the internal drive to

maximize the greenhouse gas GHG-performance, the following policy and market

developments make GHG-performance an important and valuable parameter in the

biofuel market.

• RTFO. As of April 2008, parties wishing to earn Renewable Transport Fuel

Certificates under the Renewable Transport Fuel Obligation (RTFO), need to report

on the carbon intensity and sustainability of their biofuels. The RTFO starts with a

reporting obligation in which obligated companies such as Shell and BP will be

required to report on the carbon intensity and sustainability of their biofuels. This

information will be made publicly available and it is expected that this will create a

strong moral pressure on these companies to source sustainably produced biofuels.

Recently the Department for Transport even announced that the UK Government:

“Aims to reward biofuels under the RTFO in accordance with the carbon savings that

they offer from April 2010.” This will create a higher value for biofuels with higher

GHG-performance as these biofuels enable fuel suppliers to meet their obligation with

fewer litres of biofuel.

• EU Renewable Energy Directive. The proposal for a Renewable Energy Directive

(RED) from the EC contains a 10% biofuel target with a proposed minimum GHG-

emission saving requirement of 35%. Proposed amendments to this directive show

this minimum GHG-emission saving level may rise in the future.

• EU Fuel Quality Directive. The proposal for a Fuel Quality Directive (FQD)

contains a 10% GHG-emission reduction target for the total transport fuel pool. A

significant amount of these savings will need to come from biofuels. Obligated

companies will need less biofuels to meet this target if they source biofuels with a

higher GHG-performance. Again, this creates a monetary value for biofuels with a

higher GHG-performance.

In summary, with current policy proposals a higher GHG-performance will directly

translate into a higher biofuel value. This makes both the absolute GHG-performance of

Jatropha biodiesel and its relative performance compared to other feedstocks important

parameters for the market value of Jatropha oil/biodiesel.


1.2 RTFO defaul t for Jatropha

Under the RTFO and EC-proposal, obligated parties can use conservative default values

as well as real values for the GHG-performance. Defaults are set conservative to stimulate

parties to report actual data on their GHG-performance. Defaults exist for a limited

number of fuel chains and for Jatropha biodiesel no such default currently exists under the

RTFO or EC-proposal. It is expected that in the next year a default value for Jatropha will

be developed for the RTFO.

The findings of this project can serve as input for the development of a default value for

Jatropha. In addition, this project gives D1 an insight into what the important parameters

are for the GHG-performance of Jatropha biodiesel. Even if a default value is defined for

Jatropha it will be set conservatively and this project enables (buyers of) D1’s Jatropha

oil/biodiesel to report the better performance calculated in this project based on actual

values.

1.3 Overview Jatropha b iodiese l supply chain

Figure 1 on the next page shows the Jatropha biodiesel supply chain as it was defined for

this project. This represents production chains based on D1’s Jatropha plantations in

North and North-east India. The results of this project can not be taken to be valid for

other Jatropha chains although the lessons learned in this project provide valuable insight

for other Jatropha chains as well.


1. Cultivation

2. Drying and storage(in the field)

Fruit

Hull

Seed4.5 t/ha

Shell23%

4. Oil extraction

(expeller in India)

3. Seed transport(field� expeller)

5. Oil transport(expeller� harbour)

(harbour� UK)

6. Transesterification(in UK)

7. Biodiesel transport(outside scope)

Kernel

75%

Seedcake

67%

Oil33%

Oil

Oil

-Glycerine &

-K2SO4 9%

Biodiesel91%

Biodiesel1.0 t/ha

ElectricityShells for steam generationSolvent

Transportation modesFuel efficiencyDistance

Transportation modusDistance Fuel efficiency

Natural gasElectricity

MethanolPotassium hydroxide

Cultivation inputs

Seed

Seed

Shell + dust

23% + 2%

8. Land use change (LUC) Previous land use, climate region, crop type, …

Figure 1 . Jatropha bi odiesel suppl y chain with the yie lds per phase. Words

in i ta l i c represent parameters that have been var ied in

scenarios/sens i t iv i ty analysi s. Inputs are shown on the r ight .

1.4 Structure of the repor t

The remainder of this report is structures as follows:

• Chapter 2 establishes the GHG-performance of the Base Case Jatropha Biodiesel

chain. It does this for both the current RTFO methodology and the RTFO

methodology adapted to fit the methodology from the draft RED from the EC.

• Chapter 3 compares the GHG-performance of the Base Case Jatropha chain with that

of other biodiesel chains with which it will compete in the market.

• The impact of Land Use Change on GHG-performance is discussed in Chapter 4.

• Chapter 5 analyses the possibilities for further improvement of the Base Case as well

as potential risks. This is done by analysing several what-if scenarios as well as

performing a sensitivity analysis.

• Conclusions and recommendations are given in Chapter 6.

• A separate analysis on the value of Renewable Transport Fuel Certificates is included

in Chapter 7.


Figure 2 shows the structure of our analysis as a guidance throughout the report.

Figure 2 . Overview of the st ructure of the GHG-analysi s.

A more technical discussion on the differences in the GHG methodology between the

RTFO and the EC is provided in Annex A.

Terminology

A list of terms used in this report is included below. Figure 3 on the next page clarifies the

names of the main Jatropha products.

EC European Commission

FQD Fuel Quality Directive

GHG Greenhouse Gas

Hull Product from the fruit that encapsulates the seeds

IPCC Intergovernmental Panel on Climate Change

Kernel Product from the seed

LUC Land Use Change

Jatropha Oil Product from the kernel

RED Proposal for a directive of the Parliament and the Council on the

promotion of the use of energy from renewable sources. Version 15.4,

23 January 2008.

RTFO Renewable Transport Fuels Obligation

Seed Jatropha seed (product from the fruit)

Seedcake Product from the kernel

Shell Product from the seed that encapsulates the kernel

S

y

s

t

e

m

EC

RTFO

Without LUC

Without LUC

Yields

Inputs

Transportation

Oil expeller

Base cases Chapter 2

Scenarios and comparison

Chapter 3, 4 & 5

Methodology Throughout report

(details in Annex A)

Best case

Optimal

Chapter 5 & 6

Comparison other crops

Recommendations

LUC

With LUC (IPCC Tier 1)


Fruit

Hull (pericarp)

Seed

Shell

Kernel

Seedcake

Oil

Figure 3 . Overview of the Jatropha product ion cha in with the terminology

that is used throughout the report .


2 GHG-performance: Base Case

Jatropha biodiesel saves 66% of the GHG-emissions compared to fossil diesel, even

when Land Use Change from grassland to Jatropha plantation is taken into account.

GHG-performance of the Base Case Jatropha biodiesel supply chain is nearly similar

under the RTFO-methodology as under the proposed RED-methodology, respectively

1093 kgCO2e and 1040 kgCO2e per tonne of biodiesel. Two Base Cases have been

assessed: one in which no Land Use Change (LUC) has been included and one in

which the GHG-effects of a LUC from grassland to Jatropha has been included. In

subsequent chapters, the Base Case including LUC is used to 1) compare the GHG-

performance of Jatropha biodiesel with other crops and fuels, 2) show the effect of

different types of LUC and 3) analyse how further improvements van be made and what

the main risks are.

S

y

s

t

e

m

EC

RTFO

Without

LUC

Without

LUC

Inputs

Transportatio

n modus

Transportation distance

Oil expeller


Scenarios and comparison Chapter 3 & 4


Best case

Optimal

Chapter 4 & 5

Comparison

other crops

Recommendations

Yields


2.1 Base Case: Jatropha b iodiese l reduces GHG-

emiss ions by 66% to 68%

Figure 4 shows the GHG-performance of the Base Case with and without LUC by using

both the RTFO and EC methodology. Without LUC, the Base Case for Jatropha biodiesel

under the RTFO scores slightly better than under the EC methodology. If we include LUC

in the Base Case the overall GHG-emission savings are reduced by 4.8%-pt and 2.8%-pt

dependent on the methodology. Including LUC, Jatropha biodiesel performs slightly

better under the EC-methodology than under the current RTFO methodology.

A more detailed discussion on the inputs used for the Base Case is given in section 2.2

below.


71% 70%

68%

66%

0

100

200

300

400

500

600

700

800

900

1000

1100

1200

without LUC with LUC (IPCC Grassland) without LUC with LUC (IPCC Grassland)

RTFO EC

GH

G p

erf

orm

an

ce [

kg

CO

2e

/t b

iod

iese

l]

63%

66%

69%

72%

75%

78%

81%

84%

88%

91%

94%

97%

100%

GH

G s

avin

g [

%]

8. Land Use Change

7. Biodiesel transport

6. Transesterification

5. Oil transport

4. Oil extraction

3. Seed transport

2. Drying and storage

1. Jatropha cultivation

Figure 4 . GHG-per formance o f the Base Case Jatropha biodiesel chain with

and without LUC by us ing both the RTFO and EC methodologies.

The r ight axis shows the net GHG-emission saving compared wi th

foss i l diesel . The Base Cases with LUC from grass land to

Jatropha (as calcu lated by IPCC) is used for fur ther analysi s.

2.1.1 RTFO

The Base Case Jatropha biodiesel without LUC has a GHG-performance of 934 kgCO2e/t

biodiesel under the RTFO. This increases to 1093 kgCO2e/t biodiesel if Land Use Change

from grassland to perennial cropland is taken into account by using IPCC Tier 1

calculations. In Chapter 4 it is shown that the impact of LUC becomes larger if plantations

are located in other climate zones and becomes extremely large if default values from the

RTFO or EC are used. Including LUC, Jatropha biodiesel reduces GHG-emissions by

66% compared to fossil diesel.

The largest GHG-emission contributor for the Base Case Jatropha chain is the

transesterification process at 43%. Oil transport (by truck and ship) is the second largest

contributor at 34%. Land Use Change causes 15% of the total emissions. GHG-emissions

from cultivation are zero, as no inputs are being used. Furthermore seed transport, drying

and storage and oil extraction all have very little GHG-emissions.

2.1.2 EC

Under the methodology of the EC proposal, the Base Case Jatropha biodiesel has a GHG-

performance of 945 kgCO2e/t biodiesel. This increases to 1040 kgCO2e/t biodiesel if Land

Use Change from grassland to perennial cropland is taken into account. Biodiesel saves

68% of the GHG-emissions compared to fossil diesel under the EC directive.


Also in the methodology from the EC the transesterification step is most GHG intensive,

overall responsible 48% of the GHG-emissions, followed by oil transport with 38%. Land

Use Change causes 9% of the emissions. Again, the other steps only contribute marginally

to the total GHG-emissions.

2.1.3 Differences between RTFO and EC

The difference in the GHG-performance of the Base Case Jatropha biodiesel chain using

the RTFO methodology and EC methodology is small. The differences that exist are

caused by the fact that co-products are dealt with in a different way in the EC and RTFO

methodologies. This especially influences GHG-emissions from LUC. This is discussed in

more detail in section 2.2.

2.2 Base case descr ipt ion and assumpt ions

The GHG-performance is the result of a fairly limited number of contributors in the

Jatropha biodiesel production chain. Table 1 summarises the key inputs and assumptions

for the Base Case. Where no specific information is given we have used default RTFO

values (e.g. the efficiencies of different transport modes.)

Table 1. Inputs and assumpt ions in the Base Case Jatropha biodiesel chain.

Phase Parameter / assumption

Assumption Remark

1. Cultivation -Seed yield: -Zero inputs (a -Manual harvesting

4.5 t/ha Harvested yield (b

2. Drying and storage -Dried by the sun

3. Seed transport -By truck: 150 km

4. Oil extraction -Mechanical expeller: -Electricity use:

25% oil recovery 6 kWh/t oil

5. Oil transport -By truck: -By ship:

750 km 14,500km

6. Transesterification -Biodiesel yield:

-Natural gas inputs: -Electricity inputs:

-Methanol addition: -KOH addition:

91%

1,690 MJ/t biodiesel 335 MJ / t biodiesel

113 kg / t biodiesel 26 kg / t biodiesel

-Takes place in UK

-Defaults from RTFO

7. Biodiesel transport Outside system boundaries

For Base Case with LUC

Land Use Change Grassland to perennial cropland conversion (c

-climate zone: tropical dry - 0.211 t CO2/ha/yr

IPCC Tier 1 calculation (d

a) In fact, organic manure is applied once. But resulting GHG-emissions are not accounted for in the methodologies, it has therefore been excluded from the analysis.

b) From the seed yield figures provided by D1 Oils, we have calculated an average harvested yield over 20 year plantation lifetime, see Annex B.

c) For LUC, the default has been set at a Land Use Change from grassland to cropland, assuming that D1’s plantations will be established on former grassland and not on forests.

d) We analyse the impact of LUC in the Base Case using an IPCC Tier 1 methodology. Both the RTFO and EC methodology allow producers to do so. More info on this methodology as well as the effects of using the default LUC figures given in the EC and RTFO methodology is discussed in Chapter 4.


2.2.1 Allocation factors

The difference in the outcomes between the RTFO and EC methodology is the result of

difference in the way they deal with co-products.

RTFO

The RTFO methodology uses a substitution approach if possible. If insufficient data is

available for a substitution approach, economic allocation is used. Both are briefly

explained below.

A substitution approach works on the principle that the co-product of the biodiesel chain

replaces another product. This saves the emissions that would have been emitted by

producing the replaced product. For example, rapeseed meal is assumed to replace

soybean meal. The emissions that would have been caused by this soybean meal

production are now avoided and this forms an GHG-credit for the rapeseed biodiesel

chain. The practical difficulty with this approach is that one needs to determine what

product is replaced by the biofuel co-product, in what quantities and how much emissions

are avoided by this. When this information is not available, economic allocation is used in

the RTFO.

Allocation works on the logic that each product is partially responsible for the

environmental impacts which have occurred up to this point in the supply chain and

should be allocated a portion of these impacts. In economic allocation the total emissions

caused to produce two products (e.g. rapeseed oil and rapeseed meal) are allocated to the

two products based on their economic value. The rationale for this is that the product with

the highest value should also carry the highest GHG burden. The practical difficulty with

this approach is that market prices fluctuate which changes the allocation factor over time.

This is why an average market value over a certain period is typically used. Another

challenge is formed where no mature markets exist such as for Jatropha seed cake – in

these cases it is more difficult to determine the ‘market value’.

EC

The EC also uses an allocation approach but allocates the burden to the various products

based on the energy content of these products. This has the advantage of being constant

over time but in general allocates a relatively large part of the GHG burden to residual

products with a low market value.

Allocation factors for the Jatropha chain

The Jatropha biodiesel production chain has three main co-products: seedcake (co-product

from the seed expeller) and glycerine and potassium hydroxide (both co-products from the

transesterification process). All upstream emissions have to be allocated accordingly.

Tables 2 and 3 below summarise the key allocation factors for both methodologies. These

have been used to calculate the results discussed in section 2.1.


Note that agricultural residues such as the hull do not exit the system and therefore no

emissions are allocated to these residues.

Table 2. Al locat ion factors for seedcake in the RTFO (al locat ion by market

value) and EC proposal (al locat ion by energy content) .

RTFO Market value [GBP/tproduct]

Yield [tseedcake/ toil]

Total value [GBP/toil]

Allocation factor [%]

Seedcake 36 2.03 72 14.4% Oil 427 1 427 85.6%

Total 499

EC Energy content [GJ/ tproduct]

Yield [tseedcake/ toil]

Total energy [GJ/toil]


Seedcake 20 2.03 40 51.4% Oil 37.8 1 37.8 48.6% Total 77.8

It can be seen from the table above that allocation by energy content (EC methodology) is

more beneficial for the seedcake allocation factor – compared to economic allocation

(RTFO) more emissions are allocated to the seedcake and therefore fewer emissions are

allocated to the oil. This is caused by the fact that Jatropha seedcake has a relative low

price compared to its energy content. In the sensitivity analysis (Section 5.2) prices have

been varied to determine the effect on GHG-performance.

Table 3. A l locat ion factors for glycer ine and potass ium sulphate in the RTFO

(al locat ion by market value) and EC proposal (a l locat ion by

energy content) .

RTFO Market value [GBP/tproduct]

Yield [tseedcake/ tbiodiesel]

Total value [GBP/tbiodiesel]


Glycerine 345 0.1 34.5 9% Potassium sulph. 75 0.04 3.0 1% Biodiesel 340 1 340 90%

Total 377.5

EC Energy content [GJ/ tproduct]

Yield [tseedcake/ tbiodiesel]

Total energy [GJ/tbiodiesel]


Glycerine 19 0.1 1.9 5% Potassium sulph. Not applicable (a 0.04 0 0% Biodiesel 1 37.2 95%

Total 39.1

a) According to the EC methodology: “Co-products that have a negative energy content

shall be considered to have an energy content of zero for the purpose of the calculation”

For the transesterification co-products, allocation by market value (RTFO) turns out

slightly more beneficial than by energy content (EC). This has two reasons. Firstly,

potassium sulphate has a negative energy content and therefore no emissions are allocated

in the energetic allocation method (EC). Secondly, the RTFO assumes relatively high

glycerine prices compared to the biodiesel price, which leads to a lower allocation of

GHG-emissions to the biodiesel. In the sensitivity analysis (Section 5.2) prices have been

varied to demonstrate the effect on GHG-performance.


Overall the Base Case without LUC scores better with economic allocation (RTFO), the

Base Case with LUC scores better by using energetic allocation (EC). This difference is

caused by the amount of GHG-emissions from LUC. Without LUC, only few emissions

take place that can be attributed to the seedcake, which makes the allocation of glycerine

more important and that is more beneficial under the RTFO. However, if we take LUC

into account this is responsible for a large part of the emissions and suddenly the seedcake

allocation becomes more important, which is more beneficial under the EC.

In other words, if the RTFO has to change its co-product methodology to allocation by

energy content to be consistent with the EC proposal, the GHG-performance of Jatropha

biodiesel including LUC improves slightly.

2.2.2 Land Use Change

Land use change can make or brake GHG-performance

LUC can have a significant impact on the GHG-performance of biofuels and can even

cancel out all GHG-emission savings of a biofuel.

Direct and indirect LUC

Currently both the RTFO and the EC only include emissions caused by direct LUC. In the

RTFO, emissions from LUC that occurred after November 2005 must be included. In the

EC-proposal the reference date is January 2008. This means that existing plantations do

not have to include emissions from LUC. This will normally be a disadvantage for

Jatropha as most Jatropha plantations are established after the reference year while most

competing crops have ample plantations with establishment dates before the reference

year.

Indirect LUC, in which biofuel feedstock production displace other land functions to other

areas where they may cause LUC emissions, are not included in the present RTFO and EC

proposal. However, both the recent review of the 10% target commissioned by the UK

government and the rapport of EP rapporteur Wijkman acknowledge these indirect

impacts. If emissions from indirect LUC are somehow included this could be a relative

benefit for Jatropha as any emissions for indirect LUC would only apply to existing

agricultural land and not to newly established cropland. The remainder of this report will

focus on emissions from direct LUC.

Emissions from LUC in RTFO and EC

The RTFO and EC both present default GHG-emissions from LUC that have to be added

to the GHG-performance of the biofuel, if more detailed information is not available.

Especially the use of default values has very large and negative impacts – see Chapter 4.

However, both methodologies allow producers to provide more detailed data based on

IPCC methodologies. In the Base Case calculations we have included the effects of LUC

using an IPCC Tier 1 methodology because the default numbers from the RTFO and EC

lead to a negative GHG performance.


IPCC Tier 1 approach for the Base Case

The impact of LUC on GHG-emissions in an IPCC Tier 1 approach is dependent on

various variables and assumptions, e.g. region, climate zone, original land use and the

crop replacement type. Relevant assumptions for D1 are:

• Region: We used IPCC default numbers for Continental Asia in our calculations.

• Climate zone: Rajasthan is located in a tropical dry climate whereas the north east

regions have a tropical moist climate. We have used IPCC data for a tropical dry

climate in our Base Case calculations.

• Previous land use: Either forest or grasslands can be chosen as a previous land use

(cropland to cropland in principle does not form a LUC). Grassland has been used in

the Base Case calculations.

• Replacement crop-type: Jatropha is a perennial crop, in contrast to annual crops.

The IPCC Tier 1 separates GHG-emissions associated with LUC from three different

carbon stocks:

• (Above ground) Biomass:

� Below ground biomass is assumed to remain unchanged in the Tier 1 level

methodology.

� Differences in above ground biomass before and after LUC are included in

the calculations. For the above ground biomass of the land use before

conversion we used IPCC defaults for tropical dry grassland, in line with

the above assumptions. Above ground biomass immediately after

conversion is assumed to be zero in a Tier 1 approach.

� For perennial crops, the IPCC Tier 1 approach provides default values for

the amount of annual carbon build up in above ground biomass. However,

in our opinion these numbers are not representative for Jatropha plantations

as they are focussed on (short) rotation forest crops in which the total tree is

harvested periodically. To avoid inappropriate use of IPCC default values

we have taken a conservative approach here and assumed zero carbon build

up in Jatropha in above ground biomass in the Base Case. In order to claim

the GHG-benefits of carbon build up in above ground biomass in Jatropha

plantations we recommend a Tier two or three approach which will provide

more representative results. Applying such an approach can deliver

significant GHG-benefits and is analysed in Chapter 4.1

• Dead Organic Matter (DOM) and litter:

� For grasslands carbon stocks in DOM are assumed to be zero before LUC

(in line with IPCC Tier 1). For the carbon stocks in DOM in forests we

used the IPCC Tier 1 default values for broadleaf deciduous tropical forest.

� All DOM is removed after LUC to cropland, leading to zero carbon stored

in DOM after conversion.

1 The IPCC Tier 1 approach for annual crops assumes annual growth in biomass carbon stock

equals annual losses in carbon stock (harvesting) - the resulting net build up therefore amounts

to zero.


• Soil carbon:

� For perennial crops the Tier 1 approach applies a soil carbon stock change

factor of 1. This implies that zero changes in soil carbon take place in

conversion from grassland or forest to Jatropha.

In Figure 5 the IPCC Tier 1 calculation is displayed for LUC from grassland to Jatropha

plantations in a tropical dry climate.

Total

- 1.15 t C

=

+ 0.211tCO2/ha/y

Above ground biomass- 1.15 t C

Below ground biomass+ 0 t C

Dead Organic Matter (DOM)+ 0 t C

Soil carbon+ 0 t C

Annual C built up (growth) 0 t CAssume

Annual C losses 0 t C

C before LUC 1.15 t CIPCC

C after LUC 0 t CTie 1 definition

DOM / litter stock under old LU+ 0 t CGrassland

DOM / litter stock under new LU:+ 0 t CAssume

Change in carbon stock in mineral soils+ 0 t C

No change (stock change factors for perennials ≠Tier 1)

N2O emissions from mineralised N

(result of carbon stock change, so effect under Tier 1 is zero)

Delta built up+ 0 t C

Delta biomass- 1.15 t C

Not included in Tier 1 approach

(In-)Direct N2O emission factor

0.01 + 0.00225

Amount of N

Dependent on change in management

Reference carbon stock +34.5 t C

IPCC

Stock change factors

No change after LUC All: 1

From Grassland to perennial Jatropha plantations (tropical dry climate)

Figure 5 . Breakdown of the IPCC Tier 1 calcu lat ion in four categories o f

carbon stock and the under ly ing input data for a LUC from

grassland to perennial c ropland in a t ropical dry cl imate.

The above analysis shows how LUC has been included in the Base Case. Chapter 4

includes a more detailed analysis of LUC. It discusses the effects of using the default

LUC numbers from the RTFO and EC, the effects of applying an IPCC Tier 1 approach

with different assumptions (e.g. different climate zone and vegetation type), and the

potential GHG-benefits of using field survey data to claim the carbon build up in Jatropha

plantations.


3 Comparison of Jatropha to other crops

This chapter compares the GHG-performance of the Base Case Jatropha biodiesel

chain with the default GHG-performance of other biodiesel chains. Jatropha

outperforms the default values for all other first generation energy crops. This holds

true even if LUC (from tropical dry grassland) is included for Jatropha.

Sy

st

e

m

EC

RTFO

Without

LUC

Without

LUC

Inputs

Transportatio

n modus

Transportatio

n distance

Oil expeller



Chapter 3 & 4


Best case

Optimal

Chapter 4 & 5

Comparison

other crops

Recomme

ndations

Yields


3.1 Jatropha b iod iese l compared to RTFO defau l ts

Under the default biodiesel production chains supplied by the RTFO2, Jatropha scores best

among the energy crops. Only Used Cooking Oil and tallow have a better GHG-

performance. Figure 6 displays the GHG-performance of Jatropha biodiesel (including

LUC) compared to the other crops.

2 The RTFO also provides defaults for Hydrotreated Vegetable Oils (HVO’s), these have not been

included as the GHG-performance of all these biodiesels increases equally regardless of the

feedstock.


0%

66%

85%

48%

10%

36%

-40

-30

-20

-10

0

10

20

30

40

50

60

70

80

90

100

110

RSO (UK) SBO

(Brazil)

PO

(Malaysia)

UCO &

Tallow

Jatropha

Oil

Diesel

GH

G p

erf

orm

an

ce

[kg

CO

2e

/ G

J f

uel]

Net total

Land use change

For PO: RPO refining

7 - Liquid fuel transport andstorage

6 - Conversion (esterification)

5 - Oil transport

4 - Conversion (crushing)

3 - Feedstock transport

2 - Drying and storage

1 - Crop production

Figure 6 . The GHG-per formance of Jatropha biodiesel compared to other

biodiesels under the RTFO. Some steps are indicated as

negat ive, s ince a subst i tut ion approach is used in these defau l ts .

The net total s of each crop are indicated in grey on the r ight .

The numbers at the top of the bar indicate the GHG-emiss ions

reduct ion compared to foss i l fuel .

Biodiesel from UCO and tallow scores best with 13 kgCO2e/GJ. Jatropha scores much

better than other crops, the next in line is palm oil with 45 kgCO2e/GJ compared to 29

kgCO2e/GJ for Jatropha biodiesel. Compared to fossil diesel, Jatropha biodiesel saves

66% of the GHG-emissions.

3.2 Jatropha b iod iese l compared to EC defau l ts

Figure 7 shows the comparison with the default values of biodiesel from other first

generation feedstocks as included in the EC’s proposal for a RED. Note that these are

default values from the proposed RED directive. These default values are not from the

RTFO. The default values of the RTFO may change if the RTFO has to adopt the EC

methodology for co-product treatment but this will not make them the same as the EC

default numbers. The reason for this is that the EC does not only differ from the RTFO

because of the co-product methodology but it also makes different assumptions on default

parameter values such as yields and fertiliser application rates.

Figure 7 displays two columns for palm oil as the EC provides a default value for if the oil

milling process is not specified and a default for a process in which methane emissions

are captured.


The EC only distinguishes three phases in the total GHG-performance: cultivation,

processing (both milling and transesterification) and transport and distribution. It is not

clear whether biofuel distribution is included in the scope of the EC default values - this is

not included in the RTFO and our analysis of Jatropha biodiesel. Remarkably soybean oil

has not been included in the draft directive.

68%

78%

53%

19%

53%

39%

0%

0

10

20

30

40

50

60

70

80

90

RSO Sunflower PO PO (no CH4

emissions)

Waste

vegetable or

animal oil

Jatropha Oil Diesel

GH

G p

erf

orm

an

ce

[kg

CO

2e / G

J f

ue

l]

GH

G s

av

ing

[%

]

Net total

Land use change

Transport and distribution

Processing

Cultivation

Figure 7 . The GHG-per formance of Jatropha biodiesel compared to other

biodiesels under the EC methodology.

Under the EC methodology, Jatropha ranks the same as under the RTFO leaving all other

crops behind, only waste vegetable and animal oils score better. Jatropha has a GHG-

performance of 28 kgCO2/GJ, whereas sunflower and palm oil (with methane capture)

have 41 kgCO2/GJ. Main difference between the methodologies is that the GHG-

performance of palm oil heavily depends on whether the process is specified or not. In the

EC proposal slightly higher default GHG-emissions are given for waste vegetable and

animal oil compared to the RTFO.

3.3 Understanding the di f ferences

The GHG-performance of the Base Case Jatropha biodiesel chain is better than the default

values of all first generation energy crops in both the RTFO and the draft RE-directive.

The main benefit for Jatropha lays in the very low emissions from cultivation, even if

Land Use Change from tropical dry grassland is taken into account. Using shell as an

energy source in the oil expeller provides another competitive advantage.

The results clearly illustrate the competitive advantage of Jatropha compared to other first

generation biodiesel energy crops in terms of GHG-performance. Assuming technical

suitability of JO, a market which values GHG-performance may lead to higher prices for

Jatropha biodiesel compared to biodiesel from other first generation energy crops.


Several remarks do need to be made here:

• We compared the GHG-performance of Jatropha based on real values with default

values of other crops. These default values for other crops have been set

conservatively to stimulate producers to report real values. In other words, by using

real values, suppliers of biodiesel from other energy crops may well be able to

improve their GHG-performance compared to the default values used here.

Nonetheless, Jatropha biodiesel performs very favourably compared to other energy

crops.

• In line with the above, a default value for Jatropha may be set at a more conservative

level than the Jatropha Base Case which we calculated. Currently no default value

exists in either the RTFO or draft directive. However, we expect a default value for

Jatropha to be included in the coming year in the RTFO and it will be interesting to

see at what level the default value will be set and what the difference is with both

other crops and the value for the Jatropha Base Case discussed here. Note that D1 will

be able to use the Jatropha Base Case number as long as it can demonstrate the inputs

used for the calculation as set out in the previous chapter.

• As discussed in section 5.1.3, the estimated electricity inputs for expelling that we

received from D1 are rather low. If these turn out to be higher, the GHG-performance

of Jatropha will deteriorate, resulting in a GHG-emission saving of 62% in stead of

66% in the Base Case (under the RTFO).


4 Impacts of Land Use Change

Land Use Change is the single most important parameter in the GHG-performance of

D1’s Jatropha biodiesel. If conservative defaults for LUC from the RTFO or EC are

used, this leads to higher GHG-emissions than for fossil diesel.

Both the RTFO and EC proposal allow producers to use more specific data on GHG-

emissions from LUC. The IPCC methodology (2006) is the internationally accepted

methodology for such calculations. Through the values it provides it is possible to make

a more specific calculation without the need for field carbon stock surveys.

Finally, actual data for the specific sites enables D1 to even claim an increase in carbon

stocks from LUC, which significantly improves the GHG-performance. Since

measuring actual carbon stocks was not included in the project, we show the potential

impact on the GHG-performance by using example data.

4.1 Using RTFO or EC defaul t values for Land Use

Change

In the Base Cases LUC has been taken into account by means of an IPCC Tier 1

calculation. How does this relate to the defaults provided in the RTFO and EC proposal?

We have analysed two types of land conversion: grassland to perennial cropland (Base

Case) and forestland to perennial cropland. We compare IPCC Tier 1 outcomes (Base

Case) with the RTFO and EC defaults, see Figure 8.


-183%

-285%

-105%

-224%

-183%

-5%

68%66%

Fossil diesel reference

3214

0

2,000

4,000

6,000

8,000

10,000

12,000

IPCC Tier 1

(economic

allocation)

IPCC Tier 1

(energetic

allocation)

RTFO default

(economic)

EC default

(energetic)

IPCC Tier 1

(economic

allocation)

IPCC Tier 1

(energetic

allocation)

RTFO default

(economic)

EC default

(energetic)

Grassland to Jatropha Forest to Jatropha

GH

G e

mis

sio

ns f

rom

LU

C [

kg

CO

2e/t

bio

die

sel]

8. Land UseChange

7. Biodieseltransport

6.Transesterification

5. Oil transport

4. Oil extraction

3. Seed transport

2. Drying andstorage

1. Jatrophacultivation

Fossil dieselreference

Figure 8 . The impact o f us ing defau l ts for LUC from the RTFO and EC

methodology compared to the Base Case IPCC Tier 1

calcu lat ions. Plus the di f ference in conversion f rom grass land or

forest land to perennial c ropland. For compar ison, the foss i l

diesel re ference GHG-per formance is indicated by the red l ine.

The following conclusions can be drawn from Figure 8:

• None of the defaults from either RTFO or EC could lead to a positive GHG balance

for Jatropha biodiesel. Only by using the more detailed IPCC methodology can LUC

be included without lowering the GHG-performance below that of fossil diesel.

• Conversion from forestland to perennial cropland always results in a negative GHG-

performance for Jatropha biodiesel, also if an IPCC Tier 1 approach is used.

The IPCC default numbers for carbons stocks in forest provide only a single number for

broadleaf forest in a certain region – based on full grown natural forest in that region. No

numbers are given for degraded forest for example. Therefore, as soon as an area is

classified as forest, the full carbon stock of full grown native forest in that region is taken

into account.

The above stressed the importance of the thresholds as used in the definition forest land.

The IPCC does not provide detailed definitions itself but refers to country-definitions of

forest – it does mention internationally accepted definitions such as those by FAO. As an

example, FAO uses a canopy cover threshold of only 10% - everything above that is

classified as forest. In situations where the threshold of the national forest definition are

only just exceeded, the actual carbon stock with be significantly smaller than the default

number given in IPCC (2006).


In these cases it is recommended to assess the actual carbon stock – this actual value can

then be used in the calculations in stead of the (much higher) default number.

The above discussion shows that default numbers for LUC-emissions from the RTFO and

EC can not be used. Therefore either an IPCC Tier (without field survey data) must be

used or a higher Tier (with field survey data). These are discussed below.

4.2 IPCC T ier 1 approach for LUC: the e f fect o f

c l imate zones

In the IPCC Tier 1 calculations for the Base Case, a tropical dry climate was assumed in

combination with grassland as the original land use. The climate zone and original

vegetation affects the carbon stocks such as the amount of carbon stored in above ground

biomass in forest or grassland. Figure 9 shows the effect of using a tropical moist and

tropical wet climate zone in the calculations. Tropical dry represents a large part of India

(see Annex C for an overview of climate zones) stretching from the mid to west side of

the country. Tropical wet can be found in the far east of India, near Bangladesh.

-297%

-169%

-105%

58%58%66%

Fossil diesel reference: 3214

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

11,000

12,000

13,000

Tropical dry Tropical moist Tropical wet Tropical dry Tropical moist Tropical wet

IPCC Tier 1 Grassland to Jatropha IPCC Tier 1 Forest to Jatropha

GH

G-e

mis

sio

ns f

rom

LU

C [

kg

CO

2/t

bio

die

se

l]

LUC emissions

Remaining production chain (RTFO)

Fossil diesel reference

Figure 9 . The impact of c l imate zone on total GHG-emissions from LUC and

the GHG-per formance of Jatropha biodiesel for two types of LUC.

For grasslands there is no difference between tropical moist and wet, both leading to a

decreased GHG-performance of 58% GHG-emission saving. Any forestland conversion is

highly disadvantageous to the GHG-performance, which is increasingly worse with moist

and wet climate zones.

Annex D shows the calculations behind this section per scenario.


4.3 Using s i te speci f i c data for Land Use Change

Perennial crops have an advantage above annual crops in the fact that they store carbon

during the plantation lifetime. The IPCC provides defaults for annual growth in carbon

stocks of perennial systems, but none are especially suited for Jatropha plantations. In this

section, we assess the impact of using site specific data. We assume that Jatropha crops

reach 20 kg fresh weight per tree at the end of the plantation life.

79%

58%

88%

66%

-600

-400

-200

0

200

400

600

800

1,000

1,200

1,400

1,600

Base case Carbon storage Base case Carbon storage

Tropical dry Tropical moist/wet

GH

G-e

mis

sio

ns f

rom

LU

C [

kg

CO

2/t

bio

die

sel]

LUC emissions

Remaining production chain (RTFO)

Net total

Figure 10. The ef fect of tak ing into account the carbon storage in biomass

of a perennial c rop. For al l c l imate zone thi s resul ts i s net

annual carbon capture, s ince the growth in h igher than the

carbon stock of the grassland pr ior to convers ion.

Figure 10 shows the effect of taking the carbon storage into account. In a tropical dry

grassland this leads to the carbon storage of 532 kgCO2 per tonne biodiesel (or 706 kg

CO2e per hectare). The net GHG-performance is then increased to 88% GHG-emission

saving. In a tropical moist and wet climate, the above ground biomass prior to LUC is

higher, so the effect of annual growth is relatively lower. Still 262 kg CO2e is captured per

tonne of biodiesel (or 348 kgCO2 per hectare), increasing the GHG-performance to 79%

saving. The 20 kg fresh weight per shrub is a conservative estimation and if also carbon

build in below ground biomass and soil are taken into account the effect could be larger

still.

Important to note is that an actual carbon measurement should be conducted in order to be

able to assess the carbon stock of a mature Jatropha plantation.


5 Opportunities to improve GHG-

performance

GHG-performance is most sensitive to a number of parameters which have been varied

in this chapter. Firstly, several changes are made to the Base Case scenario in order to

measure the effect on GHG-performance. Variations on cultivation inputs, LUC, and

production system parameters such as transportation modus are made. Secondly, a

sensitivity analysis is conducted on the impact of Jatropha yields, transportation

distances and oil yields on the GHG-performance of D1’s Jatropha biodiesel. Finally, a

best and a worst case scenario are constructed.

S

y

s

t

e

m

EC

RTFO

Without

LUC

Without

LUC

Inputs

Transportatio

n modus

Transportation distance

Oil expeller



Chapter 3 & 4


Best case

Optimal

Chapter 4 & 5

Comparison

other crops

Recommendations

Yields


5.1 Changes to the product ion system

Several changes are made to the production systems assumptions and defaults in order to

quantify the effect on GHG-performance. Changes are modular in this section, in contrast

to the sensitivity analysis in which parameters are varied in a continuous way.

Successively changes are made to the cultivation inputs and yields, transportation modus

and oil expeller.

5.1.1 Cultivation inputs lower the GHG-performance

In the Base Case no fertiliser or other inputs are applied on the plantations. If D1 would

apply amounts of Urea, Phosphate or Lime this would affect the GHG-performance. In

our analysis of cultivation inputs it is assumed that any input will improve the seed yield

to 6.3 t/ha (provided by D1) which should cancel out (part of) the additional emissions

from cultivation inputs.

If seed yield would increase without any additional inputs, the GHG-performance would

increase by only 1%-pt. This is caused by the limited emissions in the cultivation phase

and from LUC. Figure 11 and Figure 12 show the effect of several inputs on the net total

GHG-performance (including LUC). It shows that:


� Applying 125 kg/ha/yr of Urea results in direct emissions and, much larger, indirect

soil emissions and lead to a significant worsening of the GHG-performance to 60%

GHG-emission saving.

� Applying 12.5 kg/ha/yr of Triple Super Phosphate results in almost no increase in

GHG-intensity.

� Applying 1200 kg/ha of Lime (once at plantation establishment) results is almost no

increase in GHG-intensity.

Concluding, by applying lime or TSP there is a net beneficial effect of higher yields. If

Urea is applied, this beneficial yield-effect is outbalanced by resulting GHG-emissions

from Urea production and soil emissions.

59%

66%66%

59%

66%

60%

67%67%

60%

67%

0

100

200

300

400

500

600

700

800

900

1000

1100

1200

1300

1400

Base case Urea TSP Lime All

GH

G p

erf

orm

an

ce

[kg

CO

2e/t

bio

die

se

l]

Net total without yield increase

Net total with yield increase

Figure 11. The effect of di f fe ren t cu lt ivat ion inputs on overa l l GHG-

performance without (purple bars) and wi th (green bars) yie ld

improvements under the RTFO.

Under the EC methodology the effects are less severe: inputs lead to a relative smaller

increase in GHG-intensity. This is caused by a lower allocation factor for oil under the

energetic allocation of the EC because of which fewer of the increased emissions are

allocated to the oil.

The impact of TSP and lime on overall GHG-performance is negligible. However, the

application of nitrogen fertiliser has an important effect. If D1 is to apply Urea, this would

result in an increase of 233 kg CO2e per ton biodiesel. Still, Jatropha is performing better

than other default crops but it is getting closer to palm oil.


In conclusion, TSP, lime and probably also gypsum3 can be applied without any high risk,

but N fertiliser has a negative effect on the GHG-performance.

63%

67%67%

63%

68%

64%

68%68%

64%

68%

0

100

200

300

400

500

600

700

800

900

1000

1100

1200

1300

1400

Base case Urea TSP Lime All

GH

G p

erf

orm

an

ce

[kg

CO

2e

/t b

iod

iesel]

Net total without yield increase

Net total with yield increase

Figure 12. The effect of di f fe ren t cu lt ivat ion inputs on overa l l GHG-

performance without (purple bars) and wi th (green bars) yie ld

improvements under the EC proposal .

5.1.2 Change in transportation modus could increase GHG-performance

Overland oil transport by rail

In the Base Case Jatropha oil is transported by truck from the oil expeller to the closest

harbour over a distance of 750 km. Since in general, diesel truck transport significantly

contributes to the GHG-performance of a biofuel, it is valuable to assess the impact of a

change in transport modus from truck to train. In India the fuel efficiency of trucks is

much lower than for trains: 1.94 MJ/t.km versus 0.19 MJ/t.km. This could lead to an

improvement of the GHG-performance by 111 kgCO2e/t biodiesel under the RTFO and

by 118 kgCO2e/t biodiesel under the EC proposal. The Base Case would then be improved

by 3%-pt under both methodologies, see Figure 13.

Oversea oil transport by ship

The RTFO assumes rather low fuel efficiency for international shipping, we have found

factors that are four times as low (mostly dependent on vessel capacity). Using more

efficient ships results in a significant improvement of 6%-pt under both methodologies,

see Figure 13.

3 No data on the GHG-intensity of gypsum was available.


74%

71%

68%

72%

69%

66%

0

100

200

300

400

500

600

700

800

900

1000

1100

Base case Rail transport Efficient ships Base case Rail transport Efficient ships

RTFO EC

GH

G p

erf

orm

an

ce

[kg

CO

2e/t

bio

die

sel]

8. Land Use Change



5. Oil transport

4. Oil extraction

3. Seed transport



Figure 13. The effects of rai l transport in stead of t ruck t ransport , and the

use of more e f f ic ient sh ips on overal l GHG-per formance.

5.1.3 Variations on the oil expeller

Figure 14 shows the impact on GHG-performance of several variations to the oil expeller,

i.e. using solvent extraction, facing higher electricity inputs and using a mobile expeller.

69%

65%

68%68%68%

62%

67%66%

0

100

200

300

400

500

600

700

800

900

1000

1100

1200

Base case Solvent

extraction

Solvent +

>kWh's

Mobile

expeller

Base case Solvent

extraction

Solvent +

>kWh's

Mobile

expeller

RTFO EC

GH

G p

erf

orm

an

ce

[kg

CO

2e/t

bio

die

se

l]

8. Land Use Change



5. Oil transport

4. Oil extraction

3. Seed transport



Figure 14. The effect of var ious var iat ions on the oi l expel ler: using solvent

extract ion, facing h igher e lectr ic i ty inputs and the use of a

mobi le expel le r.


Solvent extraction

Currently the oil is extracted by using a mechanical expeller. The process requires only

electricity inputs as the shells are being burnt for steam generation. The current oil yield is

0.25 ton toil per ton seed. If solvent extraction is used, we have assumed that the

electricity inputs remain the same and shells are still sufficient to generate enough steam.

Solvent has to be added, but the GHG impact of this is zero in the RTFO. Oil yield

increases to 0.30 ton oil per ton seed if solvent extraction is used.

The higher yield from solvent extraction has only a small positive impact on the GHG-

performance which is increased with 1 %-pt resulting in 1055 kgCO2e/t biodiesel. The

effect is only small because of the low emissions up until the oil expelling. Dividing these

emissions over more oil therefore yields only little benefit.

In the sensitivity analysis we have also varied the oil yield and assessed the impact on

GHG-performance, assuming that oil yield could improve in the future regardless of the

use of a different technology.

Electricity inputs

The 6 kWh per tonne of oil electricity inputs for oil-extraction seems rather low compared

to an input of 410 kWh per tonne of oil for soy bean oil extraction (also using hexane

solvent extraction). We have run a scenario in which we have set the electricity input per

tonne of crushed Jatropha kernel equal to the RTFO default value for soybean crushing.

This leads to an electricity input of 228 kWh per tonne Jatropha oil. This lead to an

increase in the GHG-intensity to 1215 kgCO2e/t biodiesel in the RTFO and 1128

kgCO2e/t biodiesel under the EC (assuming higher oil yields as well). The smaller effect

under the EC methodology is due to the lower allocation factor for oil in this

methodology. These results indicate the importance of verifying the electricity inputs for

Jatropha oil expelling.

Mobile expeller

The use of a mobile oil expeller has several advantages: seedcake can directly be returned

to the farmers/plantation and seed transport is lowered. Oil yield is not likely to increase

with this small capacity and electricity inputs have been assumed to resemble the Base

Case.

In stead of transporting seed, now oil is transported which lowers the GHG-intensity of

this phase by 75% (direct result of the 0.25 ton oil yield). Overall this improves GHG-

performance with 64 kgCO2e/t biodiesel in the RTFO and 38 kgCO2e/t biodiesel in the

EC.


5.2 Sensi t iv i ty analys i s

5.2.1 Parameter variations

The impact of sensitive parameters is assessed on the GHG-performance of Jatropha

biodiesel. The varied parameters are:

• Seed yield;

• Transportation distance (land and freight);

• Oil yield.

For the RTFO results we have added variations on the price of:

• Seedcake;

• Jatropha oil;

• Biodiesel and glycerine price .

These prices affect the (economic) allocation factor. This is not the case in the EC results

since its allocation factor is based on energy contents.

Parameters have been varied with values that are either provided by D1 or by Ecofys

based on our view of how parameters could realistically change in the future ( Table 4).

Table 4. Summary on the var ied parameters and the explanat ions/sources

behind. Bi odiesel and g lycer ine pr ices have on ly been varied in

the RTFO Base Case.

Parameter Low Base case

High Unit Source / remark

Seed yield 3.0 4.5 6.3 t seed /ha D1: higher agronomy Transport dist. oil (land) 250 750 1,500 Km Different harbour

Transport dist. oil (ship) 12,000 14,500 16,000 Km Different harbour Oil yield 0.20 0.25 0.30 t oil /t seed Solvent extraction Jatropha oil price 350 427 550 GBP/t Increased demand Seedcake price 20 36 144 GBP/t D1: use as feed

Biodiesel price (RTFO) 300 340 700 GBP/t FO Licht 2007 Glycerine price (RTFO) 300 345 700 GBP/t Oleonline 2007

Figure 15 and Figure 16 show spider diagrams in which the relative change to a parameter

is shown on the x-axis and the effect on GHG-performance on the y-axis. Although the

sensitivity analysis is a helpful tool, the likelihood of variation is often more important.

The parameter variation given on the x-axis corresponds with realistic values.

5.2.2 General observations

Overall the graphs show that the GHG-performance is highly sensitive to the shipment of

oil. This is caused by the large distances. Although highly sensitive, it is more important

to look at realistic ranges. This indicates that oil transport will not exceed 16,000km, and

therefore the impact is still limited compared to for instance oil transport over land. Here

the ranges are much larger, which makes the overall impact to GHG-performance larger

despite the fact that the curve is less steep.


In general the GHG-performance is sensitive for oil and seed yield, but not as high as one

would expect. This is caused by the few GHG-emissions resulting from the cultivation,

transport and extraction phase compared to emissions from oil transport and

transesterification. For the results from the RTFO, glycerine and biodiesel prices and

especially the ratio between these, could affect GHG-performance significantly in the

future – this allocation factor affects the emissions from all steps up until and including

transesterification. The seedcake and Jatropha-oil price have only a small effect on the

GHG-performance – this allocation factor affects emissions of only the steps up until and

including oil expelling.

Yields: 3 - 6.3 t/ha

Oil transport (land) 250-1500km

Oil expelling: 0.25 - 0.30 t/t

Oil transport (ship) 12000-

16000km

Biodiesel price: 300-700 GBP/t

Glycerine price: 300-700 GBP/t

Seedcake price: 20-144 GBP/t

Oil price: 350-550 GBP/t

950

1000

1050

1100

1150

1200

1250

25% 50% 75% 100% 125% 150% 175% 200%

GH

G p

erf

orm

an

ce

[kg

CO

2e/t

bio

die

se

l]

60.0%

61.7%

63.3%

65.0%

66.7%

68.3%

70.0%

GH

G s

av

ing

[%

]

Figure 15. Spider diagram for several parameters wi th the RTFO methodology.


Yields: 3 - 6.3 t/ha

Oil transport (land) 250-1500km

Oil expelling: 0.25 - 0.30 t/t

Oil transport (ship) 12000-

16000km

950

1000

1050

1100

1150

1200

1250

25% 50% 75% 100% 125% 150% 175% 200%

GH

G p

erf

orm

an

ce

[k

gC

O2

2/t

bio

die

sel]

60.0%

61.7%

63.3%

65.0%

66.7%

68.3%

70.0%

GH

G s

av

ing

[%

]

Figure 16. Spider diagram wi th the EC methodology.

More sensitive for oil yield than for seed yield

Higher seed yield results in relatively less GHG-emissions attributed to cultivation and

Land Use Change. Whereas higher oil yield additionally leads to lower seed transport

emissions and emissions associated with oil extraction, making the GHG-performance

more sensitive for oil yield. The overall effect of seed yield is however larger, as the range

in expected seed yield in wider than for oil yield.

More sensitive for oil transport per ship than by truck

Oil shipping distance is a decisive parameter and affects GHG-performance to a large

extent. In the base case oil is transported over 14,500 km, so any percentual increase leads

to a significant effect. Transport by truck leads to minor increase per percentage

parameter variation, but overall there is high sensitiveness as the expected variations in

truck transport distance are very large.

Glycerine and biodiesel prices affect GHG-performance under the RTFO

Higher prices for glycerine and biodiesel increase or lower the GHG-performance

respectively, since it affects the allocation factor that is based on market values under the

RTFO. In terms of percentage the effects are minor, but as prices are highly volatile

overall effects on the GHG-performance are significant, ranging from 61% to 68% GHG-

emission saving.

As indicated in Section 3.2, the RTFO assumes different prices for biodiesel and glycerine

than are currently applicable. Prices of biodiesel are now generally higher (FO Licht,

December 2007; Oleonline 2007). Glycerine prices are expected to decline in the future as

biodiesel production is increasing. This would lead to overall worse GHG-performance.


Upgrading of seedcake and fluctuating Jatropha oil prices play a minor role

The value of seedcake can be increased by enabling its usage as animal feed. Under the

RTFO, values for seedcake (and Jatropha oil) determine the allocation factors that impacts

the GHG-performance. Overall, the impact is low. Both by percentage of increase and by

taking into account the total expected range, there is little incentive to upgrade the

seedcake. Jatropha oil price has a similar low, but inverse, effect on GHG-performance

under the RTFO.

5.3 Worst and best case scenar io

The findings of the what-if scenarios and sensitivity analysis can be combined into a

worst- and best-case scenario, with the parameters displayed in Table 5. Outcomes are

displayed in Figure 17.

Table 5. Input parameters for the worst and the best case scenar io.

Parameter Worst case Best case Unit

Seed yield 3 6.3 t/ha Cultivation inputs All Zero -

Land Use Change (a Grassland Grassland - Oil transport land 1,500 250 Km Oil transport land- modus Truck Rail - Oil transport sea 16,000 12,000 Km

Glycerine price (RTFO only) 300 700 GBP/t Biodiesel price (RTFO only) 700 300 GBP/t

a) Land Use Change has not been varied, as this would impacts the results that much that other influenced are not visible. But in general we have shown that LUC is the single most important factor in the GHG performance of energy crop based biofuels with potentially large positive and negative effects.

RTFO

A worst case scenario results in a 14%-pt reduced GHG performance of 51%. Especially

the cultivation phase and the oil transport highly contribute to this overall decline in

performance. A best case scenario results in a 9%-pt increase in GHG-performance to

75%.

EC

A worst case scenario results in a 10%-pt reduced of 58%. A best case scenario results in

a 6%-pt increase in GHG-performance to 74%.


66%

51%

75%

68%

58%

74%

0

200

400

600

800

1000

1200

1400

1600

Base case Worst case Best case Base case Worst case Best case

RTFO EC

GH

G p

erf

orm

an

ce [

kg

CO

2e

/t b

iod

iese

l]

8. Land Use Change



5. Oil transport

4. Oil extraction

3. Seed transport



Figure 17. GHG-per formance of the Base Case, worst case and best case.


6 Conclusions and discussion

6.1 GHG-performance Jatropha versus other crops

For the Indian production chain analysed in this project, Jatropha biodiesel saves 66% and

68% of the GHG-emissions under the RTFO and EC respectively. Among all first

generation biodiesel energy crops, Jatropha scores best - even when taking into account

the resulting emissions from Land Use Change.

Main advantages are the zero-input cultivation of Jatropha, which especially avoids

nitrogen fertiliser related emissions. Furthermore the shells are used for steam generation

in the oil extraction phase, which significantly reduces fossil energy inputs. On the down

side, emissions from Jatropha oil transport are considerably higher than for many other

crops due to the relatively long transportation distances.

The comparison is based on default values given in either the RTFO or EC draft directive.

These default values do not necessarily represent the values that will actually be reported

for these alternative feedstocks. Just as for Jatropha, producers can use real data which

would lead to a better GHG-performance. Nonetheless, Jatropha biodiesel performs very

favourably compared to other energy crops.

6.2 Performance under RTFO versus EC methodology

If the UK is to adopt the EC directive this would results in a slightly improved GHG-

performance of D1’s Jatropha biodiesel. Under the RTFO (allocation by market value)

Jatropha biodiesel saves 66% whereas under the EC proposal (allocation by energy

content) Jatropha saves 68% of the GHG-emissions compared to fossil diesel.

The difference in the results is caused by a different allocation method to calculate GHG-

emissions associated with the co-products. The allocation of seedcake is more beneficial

under the EC, as its energy content is relative high compared to its current market value.

This leads to much less emissions from LUC, seed transport and oil extraction attributed

to the Jatropha oil (49% in stead of 86% under the RTFO). Although the allocation of

transesterification co-products under the EC-methodology is less beneficial this does not

outbalance its positive effect of the seedcake allocation.

Although the overall differences are small, Jatropha scores slightly better under the EC

proposal. Future upgrading of the seedcake is not rewarded under the EC as it only

considers the (fixed) energy content.


The use of the EC methodology will also impact all other (default) chains reported under

RTFO. Given the scope of the study, this has not been quantified.

6.3 Areas for improvement

Besides making optimal use of the carbon storage of Jatropha plantations, several

improvements can be made that increase the GHG-performance of Jatropha biodiesel, see

Table 6.

Table 6. Measures that improve GHG-per formance Jatropha biodiese l .

Measure Result to GHG-emission saving RTFO EC

1. Use large efficient ships for international oil transport + 5.8%-pt + 6.7%-pt 2. Oil transport by rail + 3.5%-pt + 3.7%-pt

3. Mobile oil expeller + 2.0%-pt + 1.2%-pt 4. 4x higher seedcake value + 2.4%-pt + 0%-pt 5. Higher yields (by TSP/lime) + 1.4%-pt + 0.9%-pt 6. Solvent oil extraction + 1.2%-pt + 0.4%-pt

7. Site specific carbon stock measurement Approximately + 22%-pt

The choice for oil shipment in large crude carriers is one of the controllable measures that

contributes to a better GHG-performance. Although it seems theoretical, if D1 is able to

control this, GHG-emission saving could be increased to 72% under the RTFO and 74%

under the EC proposal.

Other measures such as solvent extraction and the introduction of a mobile expeller

provide smaller improvements, up to 3.5 percentage points.

Land Use Change can have a significant positive contribution to GHG-performance if real

carbon measurements are used to take into account the annual built up carbon in the

Jatropha plantation. For a tropical dry grassland conversion and above ground biomass of

20 kg dry weight per tree, GHG-performance could be increased to 88% savings.

6.4 Land Use Change

Land Use Change is by far the most important parameter for D1’s Jatropha plantations.

Emissions from Land Use Change must be included as most plantations are established

after to the dates provided in the RTFO and EC proposal. The most important

conclusions with respect to direct LUC are:

• Any default for LUC provided in the RTFO and EC proposal ruins the GHG-

performance.

• Any conversion from forestland results in more GHG-emissions compared to fossil

diesel in the IPCC Tier 1 approach.

• Under the IPCC Tier 1 approach, grassland conversion in a tropical dry, moist or wet

climate (representing the major part of the plantations in India) results is a minor

decrease in GHG-performance, but still saves GHG-emissions compared to fossil

diesel.

• Land use conversion could result in a positive contribution to GHG-performance if

the actual increase in carbon stocks is quantified.


This requires actual carbon stock measurements of the plantation and preferable also of

the original land cover. In some cases this results in an increase in carbon storage. This is

the main advantage that perennial crops have and should be addressed in detail. Actual

field measurements are also of importance in those cases in which national forest

thresholds are just exceeded.


References

E4Tech 2008: Carbon Reporting within the Renewable Transport Fuel Obligation –

Methodology.

EC 2008: Proposal for a directive of the Parliament and the Council on the promotion of

the use of energy from renewable sources. Version 15.4, 23 January 2008.

FO Licht, 2007: World Biodiesel Price Report. Vol 1, No 47, 6 December 2007.

FO Licht, 2008: World Biodiesel Price Report. Vol 2, No 18, 8 May 2008-06-11

FO Licht, 2008: World Ethanol and Biofuels Report. Vol 6, No 13, 10 March 2008

FO Licht, 2008: European Ethanol Price Report. Vol 4, No 1, 31 March 2008

IEA 2007: Energy statistics

IPCC 2006: IPCC Guidelines for National Greenhouse Gas Inventories

Oleonline, 2007: Glycerine Market Report. Nr 79, 15 December, 2007.


Annex A Key dif ferences between the

RTFO and EC proposal

This Annex summarises the key differences between the RTFO methodology and the

methodology as proposed in the EC proposal. Main differences between the two

methodologies are the way co-products are dealt with, default values and Land Use

Change parameters.

Status of the RTFO and draft EC directive

In January 2008, the RFA issued the technical guidance for carbon and sustainability

reporting within the RTFO. As of April 2008 the RTFO came into force, requiring 2.5%

of all road fuels to come from biofuels. Transport fuel suppliers can comply with the rules

by supplying the relevant amount of biofuel themselves, purchasing certificates from

another transport fuel supplier, or by paying a ‘buy out' price in respect of some or all of

their obligation.

Also in January 2008, the European Commission (EC) released its draft directive on the

promotion of the use of energy from renewable sources. This document includes default

values for several biofuel production chains as well as a methodology that treats co-

products by means of allocation by energy content. While still uncertain, it is expected the

directive will be finished in 2009 and will enter into force in 2010.

Calculating GHG-emissions associated with co-products

The impact of co-products must be taken into account when calculating the carbon

performance of a biofuel. Multiple approaches are used to deal with co-products:

� Substitution: this is based on the principle that a biofuel should be attributed with

any consequences (i.e. increased or avoided GHG-emissions) of an increase in

demand of any (co-) product. Any impact that a co-product has on GHG-emissions

should be included within the boundaries of the biofuels’ carbon performance.

� Allocation: each co-product is partially responsible for the environmental impacts

which have occurred up to this point and should be allocated a portion of these

impacts. The allocation can be carried out on the basis of a range of characteristics

of the co-product, the relevant methods are given below:

o Market value.

o Energy content.

The reader is directed to E4Tech (2008) for further explanations and discussions on co-

products treatment.


The approach taken in the RTFO depends on the co-products and its use. In principle, co-

products must be accounted for by using the substitution approach where possible. Where

there is no sufficient data to undertake the substitution approach, the co-products must be

accounted for by using allocation by market value. For default fuel chains in the RTFO it

has already been indicated how to address co-products and fixed credits have been

determined for most of the different uses of the co-products. Jatropha biodiesel is not

included in the RTFO default chains and no default uses of the main co-products are

provided. Therefore, allocation by market value is the preferred approach, since detailed

information on the co-products usage is lacking.

In the EC proposal all co-products are dealt with by allocation on energy content.

Defaults

The RTFO calculation methodology uses default values that provide estimates of the

carbon intensity of different fuel chains. This enables suppliers with specific information

about their supply chain to provide additional qualitative or quantitative data to improve

the accuracy of the calculation. High level default values (where little is known about the

origin of the supply chain) represent conservative GHG-emission savings; but typical

default factors (where the calculation includes more detailed information) are less

conservative in order to encourage the supply of information. This is illustrated in the

figure below.

The EC directives in principle proposes the same approach but provides default values at

only one level,


Land Use Change

Where information on previous land use has been supplied the calculation includes the

effect on overall GHG-emission savings. RTFO default values for specific Land Use

Changes are based on IPCC guidelines and are specified per climate region, crop type and

vegetation type before conversion. For all plantations established after November 30,

2005 emissions from Land Use Change must be included.

In the EC directive, only one single default value for emissions from LUC is given

without any differentiation between regions, crop types or original vegetation. The

reference date in the proposal is set at January 2008.

Both the RTFO and EC-proposal allow for the use of more detailed data based on IPCC

methodologies.


Annex B Jatropha yield

Base case yields

D1 Oils provided us with Jatropha seed yields for various stages in the plantation lifetime.

An average yield figure has been used to calculate the GHG-performance of Jatropha

biodiesel. Yield has been averaged over the plantation lifetime, in which only harvested

yield is considered. This is in line with FAO statistics which also give average yields for

‘harvested area’. We assume a plantation lifetime of 20 years, excluding the first three

unproductive years.

Jatropha yields no seeds in the first three years. Only in year 4 yields start and grow until

year 9, when yields remain constant until year 23.

Year Yr 1-3 Yr 4 Yr 5 Yr 6 Yr 7 Yr 8 Yr 9-23

Yield 0 1.875 2.5 3.125 3.75 4.375 5 tseed/ha

Average harvested yield is then: 4.53 t/ha.

High yields

A higher yield series has been provided by D1 Oils as well to represent better agronomy

figures. These yields are indicated below.

Year Yr 1-3 Yr 4 Yr 5 Yr 6 Yr 7 Yr 8 Yr 9-23

Yield 0 2.5 3.75 5 5.625 6.25 6.875 tseed/ha

Average harvested yields equals 6.31 t/ha.

Average Base case

yield

Average Improved

higher yield

0

1

2

3

4

5

6

7

8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Years

Ja

tro

ph

a s

ee

d y

ield

[to

n/h

a]

Base case yields

Improved higher yields


Annex C IPCC definitions

Land use categor ies (IPCC, 1996)

� Forest Land

This category includes all land with woody vegetation consistent with thresholds used to

define Forest Land in the national greenhouse gas inventory. It also includes systems with

a vegetation structure that currently fall below, but in situ could potentially reach the

threshold values used by a country to define the Forest Land category.

� Cropland

This category includes cropped land, including rice fields, and agro-forestry systems

where the vegetation structure falls below the thresholds used for the Forest Land

category.

� Grassland

This category includes rangelands and pasture land that are not considered Cropland. It

also includes systems with woody vegetation and other non-grass vegetation such as herbs

and brushes that fall below the threshold values used in the Forest Land category. The

category also includes all grassland from wild lands to recreational areas as well as

agricultural and silvi-pastural systems, consistent with national definitions.

� Wetlands

This category includes areas of peat extraction and land that is covered or saturated by

water for all or part of the year (e.g., peatlands) and that does not fall into the Forest Land,

Cropland, Grassland or Settlements categories. It includes reservoirs as a managed sub-

division and natural rivers and lakes as unmanaged sub-divisions.

� Settlements

This category includes all developed land, including transportation infrastructure and

human settlements of any size, unless they are already included under other categories.

This should be consistent with national definitions.

� Other Land

This category includes bare soil, rock, ice, and all land areas that do not fall into any of

the other five categories. It allows the total of identified land areas to match the national

area, where data are available. If data are available, countries are encouraged to classify

unmanaged lands by the above land-use categories (e.g., into Unmanaged Forest Land,

Unmanaged Grassland, and Unmanaged Wetlands).


Cl imate zones (IPCC, 1996)


Eco logi ca l zones (source: IPCC, 1996)


Annex D LUC calculations IPCC Tier 1

Total- 1.15 t C

=

+ 0.211tCO2/ha/y

Above ground biomass- 1.15 t C



Soil carbon+ 0 t C

Annual C build up (growth) 0 t CAssume



C after LUC 0 t CBy definition




No change (management factors for perennials ≠Tier 1)

N2O emissions from mineralised N(result of carbon stock change, so effect under Tier 1)

Delta built up+ 0 t C



(In-)Direct N2O emission factor0.01 + 0.00225

Amount of NDependent on change in management

Reference carbon stock +34.5 t C

IPCC

Management factorsNo change after LUC All: 1

From Grassland to perennial Jatropha plantations (tropical dry climate)

Total-69 t C

=

+ 12.6tCO2/ha/y

Above ground biomass- 65 t C


Dead Organic Matter (DOM)- 3.65 t C

Soil carbon+ 0 t C


C before LUC 130 t biomassIPCC


DOM / litter stock under old LU- 3.65 t C

IPCC Forest land


Delta growth+ 0 t C

Delta biomass- 65 t C


(In-)Direct N2O emission factor

0.01 + 0.00225

Reference carbon stock +34.5 t CIPCC


From Forest to perennial Jatropha plantations (tropical dry climate)







Total- 3.1 t C

=

+ 0.57tCO2/ha/y

Above ground biomass+ 3.1 t C



Soil carbon+ 0 t C






Delta growth+ 0 t C




Reference carbon stock +55 t C

IPCC


From Grassland to perennial Jatropha plantations (tropical moist climate)






Total-90 t C

=

+ 16.3tCO2/ha/y

Above ground biomass- 90 t C


Dead Organic Matter (DOM)- 3.65 t C

Soil carbon+ 0 t C


C before LUC 180 t biomassIPCC


DOM / litter stock under old LU- 3.65 t C

IPCC Forest land


Delta growth+ 0 t C

Delta biomass- 90 t C



Reference carbon stock + 55 t C

IPCC


From Forest to perennial Jatropha plantations (tropical moist climate)






Documents

GHG PERFORMANCE JATROPHA BIODIESEL · 2009-04-28 · Jatropha it will be set conservatively and this project enables (buyers of) D1’s Jatropha oil/biodiesel to report the better