Using LCA to understand the climate change effects of bioenergy

Dr Miguel Brandão

Outline

• Life Cycle Assessment as a decision‐support tool

• Policy Context• Unresolved methodological issues in the LCA of bioenergy systems – Reference systems & carbon stocks– Biogenic carbon & time accounting– iLUC, modelling approaches & allocation

• Prospects

Counter‐intuitive lessons from hard systems analysis

•What’s better?

– Composting vs Incineration– Reusable nappies  vs  Disposable nappies– Recycled bags vs Plastic bags– Recycled paper  vs  Virgin paper– Organic food vs  Conventional food– Local food vs Imported food– Bioenergy vs Fossil fuels

Food Miles

Greenhouse gas emissions from the food chain: 2004 in United Kingdom

Source: SEI (2006) in defra (2007)

Life Cycle Assessment compared with process or site-specific environmental analysis

Extraction and

processing of raw materials

Manufacture Use, reuse and/or

maintenance

Distribution and retailing

Waste management

Raw materials and energy

Product(s); solid waste; emissions to air, water and land

ENVIRONMENT

Source: Hodgson et al. (1997) in Cowell (1998).

Environm

entalmechanism

(impact pathw

ay)

Environm

entalmechanism

(impact pathw

ay)

Midpoints

NOx, Cd, CO2, CH4, dioxins, energy, coal, silver ore, land use… and other emissions and resource flows... Inventory

Summer sm

og

Ozone

 layer

Acidificatio

n

Clim

ate change

Carcinogen

s

Land

‐use

Ecotoxicity

Eutrop

hicatio

n

Resource dep

letio

n

Respira

tory inorganics

Radiation

HumanHealth

Naturalenvironment

Natural resources

Area of Protection

Toxicity

weighting

Single indicator (overall environmental impact )

weighting

Aggregation to single score indicator

· Integrated Product Policy Communication (IPP), 2003:

“LCA is the best framework for assessing the potential environmental impacts of products, but the debate is ongoing about good practice”

· Sustainable Consumption and Production Action Plan, 2008:

“To implement this policy, consistent and reliable data and methods are required to asses the overall environmental performance of products …”

European Commission response to the need for consistent and quality‐assured life cycle data and assessments

European Platform on Life Cycle Assessment

Policy Context: EC policy needs

Necessity of Life Cycle Thinking in Policy and Business· Avoid shifting of burdens: 

∙ from one stage to another ∙ among countries∙ across different environmental and health impacts and resources use∙ from one generation to the next

· Fair basis for comparisons:∙ scope of assessments (stages, impacts, …)∙ same functionality

EU Renewable Energies Directive Implementation of Directive 2009/28/EC on the Promotion of the use of energy from renewable sources 10% share of biofuels in transports by 2020

The Directive addresses the production and use of biofuel in a sustainable context (Article 17): Minimum GHG savings, relative to fossil fuels of 35% (50% from 2017) Preservation of biodiversity on land used for biofuel production Preservation of lands with high carbon stocks Requirement for good agricultural and environmental conditions

• ILUC effect refers to the GHG emissions from land‐use conversions arising as a consequence of increased demand for land‐based products

Life cycle methodology –RED’s Annex V

04/03/2015 IEA Bioenergy Task 38 Webinar ‐ The Application of Life Cycle Assessment in quantifying the climate change effects of bioenergy

Typical and default values for biofuels

Biofuel production pathway Typical greenhouse gas emission saving

Default greenhouse gas emission saving

sugar beet ethanol 61 % 52 %

wheat ethanol 32 % 16 %

corn (maize) ethanol 56 % 49 %

sugar cane ethanol 71 % 71 %

rape seed biodiesel 45 % 38 %

sunflower biodiesel 58 % 51 %

soybean biodiesel 40 % 31 %

palm oil biodiesel  36 % 19 %

pure vegetable oil from rape seed 58 % 57 %

European Commission Renewable Energies Directive (2009)

Large variability in published GHG emissions

-1000

-500

0

500

1000

1500

2000

Ded

icat

ed F

eeds

tock

s

Res

idue

s

With

Avo

ided

Em

issi

on C

redi

ts

Pho

tovo

ltaic

s

Con

cent

ratin

g So

lar P

ower

Geo

ther

mal

Hyd

ropo

wer

Oce

an E

nerg

y

Win

d

Nuc

lear

Nat

ural

Gas

Oil

Coa

l

Life

Cyc

le G

reen

hous

e G

as E

mis

sion

s (g

CO

2e /

kWh)

Biopower (no LUC)

Count of Estimates 105 96 8 124 42 6 28 10 126 120 82 24 169 Count of References 29 30 6 26 16 5 11 5 49 31 35 10 49

Technologies powered by renewable resources

Technologies powered by non-renewable resources

Source: Heath et al. (2010) Special Session on Meta-Analysis of Energy LCAs. InLCA X

Applying LCA in quantifying the climate effects of bioenergy systems• Urgent need for replacing fossil fuels in order to mitigate climate change

• Bioenergy systems promising strategy• Quantified benefits of bioenergy systems very variable and depend on methodological choices – Reference systems & carbon stocks– Biogenic carbon & time accounting– iLUC, modelling approaches & allocation

• How important are these choices?

Reference Systems & Carbon Stocks

Carbon stocks in UK 

Source: : Azeez (2009), IPCC (2003; 2006)

101 10166 77 82 101

21 0

90

5 2

0

50

100

150

200

250

Forest

Grasslan

dAnn

ual C

ropland

Organic

Cropland

Set-Asid

ePere

nnial

Cropla

nd

t C / ha

Above-Ground BiomassDead Organic MatterSoils

Change in carbon stocks(20 years after conversion)

Source: IPCC (2003; 2006)

Biogenic Carbon & Time Accounting

Radiative forcing (W/m2), t CO2‐eq. and t C*yr

Global Warming Potential: the extended Moura‐Costa approach

Global Warming Potential (CO2-eq)

20 years 100 years 500 years

Carbon Dioxide 1 1 1

Methane 72 25 7.6

Nitrous Oxide 289 298 153

Carbon Dioxide Sequestration (tonne-years)

1/14.6 = 0.074 1/47.8 = 0.021 1/157.3 = 0.006

Carbon tonne-years

0.074*44/12==0.27

0.021*44/12==0.08

0.006*44/12= =0.022

Calculating impacts from LU and (i)LUC

2/20*)( finini ECSECS

ECSini

Timetfin

EcosystemCarbon

Stock

tini

ECSfin

ECSdeficitECS inifinfinini ttECSECS

tini+20tini+1

Lashof Approach

Indirect Land Use Change (iLUC)

What is iLUC?dLUC vs iLUC

Crop displacement (Edwards et al. 2010)

Source: Edwards (2009)

Critical issues in iLUC modelling

1. GHG emissions from LUC– Land requirements– Reference system

2. Ascribing LUC to their drivers– Amortisation

3. Dealing with modelling uncertainties– Treatment of by‐products– intensification vs. expansion vs. displacement

Where does increased supply come from?• Existing agricultural land

1. Displacement decreased food availability or iLUC

2. Yield increases increased input use

• Newly‐converted agricultural land3. Direct Land Use Change

– Increased carbon emissions

Alternative modelling approaches

Economic‐equilibrium models

• Price elasticities

• Black boxes

Biophysical models

• Usually no constraints considered, full price elasticities

• Transparent physical relations

Comparison of iLUC models

31

Schmidt J H and Brandão M (2013), LCA screening of biofuels ‐ iLUC, biomass manipulation and soil carbon. Concito, Copenhagen

iLUC factors from different modelsSource iLUC Factor (g CO2‐eq./MJ)

Searchinger et al., 2008 106

Audsley et al., 2010 1.43 t CO2‐eq./ha of agricultural land used

Cederberg et al., 2011 n/a

EPA RFS2 30.1

Dumortier et al., 2009 118

Tyner et al., 2010 13.9

JRC 36

JRC ethanol 4 to 20

JRC biodiesel 36 to 60

Are models suitable for determining ILUC factors?

Source: Berndes, G., Bird, N. & Cowie, A., 2011. Bioenergy, land use change and climate change mitigation. Background Technical Report, Paris: International Energy Agency (IEA).

New EU proposal (October 2012) to reduce iLUC risk• Amends both the Renewable Energy (2009/28/EC) and Fuel 

Quality (2009/30/EC) Directives• Avoids incentives for the continued displacement of food 

crops for fuel– A 60% GHG‐saving for new biofuels installations (from 1 July 2014)– Including iLUC factors in the reporting by suppliers and MS– A 5% cap on the amount of biofuels in the EU’s 2020 transport mix– An end to public subsidies for 1st generation biofuels after 2020 unless 

they can demonstrate “substantial GHG savings”– Incentives for 2nd and 3rd generation biofuels– A review of policy and scientific evidence on ILUC (in 2017)

04/03/2015 IEA Bioenergy Task 38 Webinar ‐ The Application of Life Cycle Assessment in quantifying the climate change effects of bioenergy

When the new ILUC proposal was launched on 17 October 2012, EU Climate 

Commissioner Connie Hedegaard said that some biofuels currently receiving EU 

subsidies were “as bad as, or even worse than the fossil fuels that they replace.”

Legislative milestones• 2008: EP asked EC to look into iLUC• 2010: importance of iLUC recognised but doubts about appropriate 

methodology• October 2012: proposal for considering iLUC with factors for 

different crop groups and a 5% cap• June 2014: EU Council reaches agreement:

– No subsidies for worst offenders post 2020– 7% cap from 1st generation biofuels– 0.5% target for advanced biofuels

• February 2015: Approved by EP Environment Committee (374 proposed amendments on iLUC!)– 6% cap– Boost advanced biofuels (1.25%)– Reduce iLUC (EU biofuels policy budget of €10 billion / year)

Estimated indirect land‐use change emissions from biofuels• Cereals and other starch rich crops ‐ 12 g CO2‐eq/MJ• Sugars ‐ 13 g CO2‐eq/MJ• Oil crops ‐ 55 g CO2‐eq/MJ

• Fossil fuel comparator ‐ 83.8 g CO2‐eq/MJ (2008)– 2012: 89.2 for diesel– 2012: 90.7 for gasoline– 2012: 90.3 for weighted average (3:1)

New factors for biofuels • Palm Oil – 105 g CO2‐eq./MJ• Soybean – 103 g CO2‐eq./MJ• Rapeseed – 95 g CO2‐eq./MJ• Sunflower – 86 g CO2‐eq./MJ• Palm Oil with methane capture – 83 g CO2‐eq./MJ• Wheat (process fuel not specified) – 64 g CO2‐eq./MJ• Wheat (as process fuel natural gas used in CHP) – 47 g CO2‐eq./MJ• Corn (Maize) – 43 g CO2‐eq./MJ• Sugar Cane – 36 g CO2‐eq./MJ• Sugar Beet – 34 g CO2‐eq./MJ• Wheat (straw as process fuel in CHP plants) – 35 g CO2‐eq./MJ• 2G Ethanol (land‐using) – 32 g CO2‐eq./MJ• 2G Biodiesel (land‐using) – 21 g CO2‐eq./MJ• 2G Ethanol (non‐land using) – 9 g CO2‐eq./MJ• 2G Biodiesel (non‐land using) – 9 g CO2‐eq./MJ

No GHG saving!

<35% GHG saving

50-60% GHG saving

35‐50% GHG saving

>60%

Modelling Approaches and Allocation

Attributional and consequential modelling

• UNEP/SETAC (2011). Shonan LCA database guidance principles:– Attributional approach: System modelling approach in which 

inputs and outputs are attributed to the functional unit of a product system by linking and/or partitioning the unit processes of the system according to a normative rule.

– Consequential approach: System modelling approach in which activities in a product system are linked so that activities are included in the product system to the extent that they are expected to change as a consequence of a change in demand for the functional unit.

Consequential

Attributional and consequential modelling (Weidema, 2003)

Attributional

42

System boundary ‐ Dairy cow example

System boundary for substitution

System boundary for substitution and allocation

ISO/CLCA(substitution)

Allocation– impossible activities are created

Milk Meat Milk Meat

Allocation

Unallocated milking cow(per 100kg DM feed)

9.3kg Milk 2.2kg Meat

100kg DM feed

2.0kg CH4

28.3kg C in CO2

23.2kg Manure

35.0kg respiratory water

Milk: 77% of economic turnoverMeat: 23% of economic turnover

outputs = 100kg

04/03/2015 IEA Bioenergy Task 38 Webinar ‐ The Application of Life Cycle Assessment in quantifying the climate change effects of bioenergy

“Meat” and “Milk” from the multi‐output system “Dairy cow”

Original systems System expansion Dry matter allocation Economic allocation

Dairy cow Meat cattleMilk = 

Dairy Cow‐Meat cattle

Meat = Meat cattle Milk 81% Meat 19% Milk 77% Meat 23%

Feed DM 100.0 26.0 74.0 26.0 81.0 19.0 77.0 23.0

Milk DM ‐9.3 ‐9.3 ‐9.3 ‐9.3

Meat DM ‐2.2 ‐2.2 ‐2.2 ‐2.2 ‐2.2

CH4 ‐2.0 ‐0.5 ‐1.5 ‐0.5 ‐1.6 ‐0.4 ‐1.5 ‐0.5

Manure DM ‐23.2 ‐4.7 ‐18.5 ‐4.7 ‐18.8 ‐4.4 ‐17.9 ‐5.3

C in CO2 ‐28.3 ‐8.1 ‐20.2 ‐8.1 ‐22.9 ‐5.4 ‐21.8 ‐6.5

Respiratory water ‐35.0 ‐10.5 ‐24.5 ‐10.5 ‐28.4 ‐6.6 ‐27.0 ‐8.1

Mass balance: 0.0 0.0 0.0 0.0 0.0 0.0 ‐0.4 0.4

Feed C 46.5 12.0 34.5 12.0 37.66 8.84 35.8 10.7

Milk C ‐5.3 ‐5.3 ‐5.3 ‐5.3

Meat C ‐1.0 ‐1.0 ‐1.0 ‐1.0 ‐1.0

CH4‐C ‐1.5 ‐0.4 ‐1.1 ‐0.4 ‐1.22 ‐0.29 ‐1.2 ‐0.3

Manure C ‐10.4 ‐2.2 ‐8.2 ‐2.2 ‐8.42 ‐1.98 ‐8.0 ‐2.4

C in CO2 ‐28.3 ‐8.4 ‐19.9 ‐8.4 ‐22.92 ‐5.38 ‐21.8 ‐6.5

C balance: 0.0 0.0 0.0 0.0 ‐0.2 0.2 ‐0.5 0.5

Allocated milking cow(economic allocation: milk 77%)

48

9.3kg Milk 2.2 Meat

77kg DM feed

1.5kg CH4

21.8kg C in CO2

17.9kg Manure

27.0kg respiratory water

outputs = 77.5kg

Milk: 77% of economic turnoverMeat: 23% of economic turnover

The co‐product algorithmCase study: Small additional demand for protein feed

Soybean mill

Barley to 

generic feed 

market

Market for generic vegetable 

oil

Soybean production

Land Use

Palm oil mill Soybean 

meal to generic feed 

market

1 kg

Market for 

protein feed, crude

1 kg

20.5 MJ energy feed

-20.5 MJ

-0.26 kg crude protein

Barley production

2.13 kg meal

0.53 kg veg. oil

‐0.53 kg veg. oil

‐2.77 kg barley

‐0.55 kg

Oil palm production

‐4. 9 kg FFB

2.77 kg soybean

To compare: An attributional system

Case study: Small additional demand for protein feed

Soybean meal to generic feed 

market

Market for 

protein feed, crude

Barley to 

generic feed 

marketprotein

Palm oil millmeal

Corn productionproteinRape oil 

mill

Sunflower oil mill

meal

meal

Wheat flour millbran & germ

Rape production

Oil palm production

Soybean productionSunflower 

production

Wheat production

Barley production

Land Use

Results for 1 kg soybean meal[kg CO2‐equivalents]Activity: Consequential AttributionalTotal of all activities -0.101 0.889iLUC 3.980 0.338Soybean cultivation 0.275 0.139Natural gas 0.082 0.042Light fuel oil 0.042 0.033Electricity BR 0.025 0.008Other chemicals -0.004 0.013Lorry -0.016 0.001Farm capital equipment -0.038 0.078Farm services -0.045 0.092Oil mill -0.658 0.014Diesel -0.675 0.107Oil palm cultivation -0.897 <0.001Barley cultivation -0.985 <0.001Fertiliser -1.130 0.002Other activities -0.057 0.021

04/03/2015 IEA Bioenergy Task 38 Webinar ‐ The Application of Life Cycle Assessment in quantifying the climate change effects of bioenergy

Case Study

UK bioenergy Crops Substitution effects Marginal producers

-Ethanol

-Biodiesel

-Combined Heat andPower

WheatSugar Beet

Oilseed rape

MiscanthusWillow SRCScots Pine

Gasoline + Animal feed*(DDGS)

Diesel + Animal feed* (rapecake)

Coal and/or Natural Gas(marginal feedstocks)

Argentina (soymeal),Indonesia (vegetable oil)and Canada (wheat)

*A combination of palm oil, soymeal and feedwheat (and associated land use!)

Bioenergy on Current Cropland in UK

‐12 000

‐10 000

‐8 000

‐6 000

‐4 000

‐2 000

0

2 000

4 000

6 000

8 000

 FtFuelW  FtFuelOSR  FtFuelSB

kgCO2‐eq./ha

 UK ‐ Fossil

 UK ‐ Biogenic

 Substitution ‐ Food

 iLUC ‐ Food

 Substitution ‐ Feed and vegetable oil

 iLUC ‐ Feed and vegetable oil

 Substitution ‐ Fossil Fuels

Bioenergy on Set Aside in UK

‐20 000

‐15 000

‐10 000

‐5 000

0

5 000

10 000

15 000

 UK ‐ Fossil

 UK ‐ Biogenic

 Substitution ‐ Food

 iLUC ‐ Food

 Substitution ‐ Feed and vegetable oil

 iLUC ‐ Feed and vegetable oil

 Substitution ‐ Fossil Fuels

kgCO2‐eq./ha

Bioenergy on Grassland in UKkgCO2‐eq./ha

‐20 000

‐15 000

‐10 000

‐5 000

0

5 000

10 000

15 000

 UK ‐ Fossil

 UK ‐ Biogenic

 Substitution ‐ Food

 iLUC ‐ Food

 Substitution ‐ Feed and vegetable oil

 iLUC ‐ Feed and vegetable oil

 Substitution ‐ Fossil Fuels

Critical parameters1. Reference System & Carbon Stocks

– Grassland 2,000‐7,000 kg CO2‐eq/ha higher than Set Aside– Carbon emissions from LUC (UK): 3 – 145 t C/ha

2. Biogenic carbon & Time accounting– Extended Moura‐Costa– Alternatives possible (Lashof, BioGWP)

3. Indirect Land Use Change, Modelling approaches & Allocation– European Commission factors: 

• Cereals and other starch rich crops ‐ 12 g CO2‐eq./MJ• Sugars ‐ 13 g CO2‐eq/MJ• Oil crops ‐ 55 g CO2‐eq/MJ

– Attributional vs Consequential• Audsley et al. (2010): 1.43 t CO2/ha• CLCA: 1.32 t CO2‐eq./ha*year‐eq.• ALCA: 32 kg CO2‐eq./ha*year‐eq.

Schmidt, J.H., Weidema, B.P., Brandão, M., 2015. A Framework for Modelling Indirect Land Use Changes in Life Cycle Assessment. J. Clean. Prod. (accepted for publication)

Precision vs. Accuracy

High PrecisionLow AccuracyBiased results

High AccuracyLow PrecisionRepresentative Results

Source: Brandão et al. (2014) The Use of Life Cycle Assessment in the Support of Robust (Climate) Policy Making. Journal of Industrial Ecology

¨It is much more important to be able to survey the set of possible systems approximately than to examine the wrong system exactly. It is better to be approximately right than 

precisely wrong.” 

Tribus and El‐Sayed (1982)

Conclusions and Outlook• LCAs relevance for policy support growing

• Several international developments and ongoing initiatives• Better methods, databases, software• Lack of consensus• Increasing harmonisation needed in LCA practice

• LCA of bioenergy systems• Methodological choices in LCA determine results• Biomass/biofuels not necessarily better• Indirect effects important• All models are wrong, some are useful

• Non‐shifting of burdens

¨It is much more important to be able to survey the set of possible systems approximately than to examine the wrong system exactly. It is better to be approximately right than precisely wrong.” 

Tribus and El‐Sayed (1982)

Thank you!