Biofuel sustainability The issue of indirect land use change () Presentation at the Annual Danish Environmental Economic Conference 27 August 2013
Content Short introduction to biofuel sustainability Issues when using models to guide policy Regulatory considerations 2
Short introduction to biofuel sustainability
Short introduction to biofuel sustainability Short introduction to biofuel sustainability Biofuels replace fossil fuels and leads to GHG reductions o Led to biofuel mandates Aware of problem of releasing carbon sink to produce biofuel crops (DLUC) o Regulated in EU legislation People started thinking about an indirect effect () o Pressure on politicians to act o Problem that is not observable: Need to use complex economic modelling 4
Short introduction to biofuel sustainability Indirect land use change Difference between and DLUC Forest Forest Ag. food production Ag. food production Biofuel Biofuel Ag. food Direct land use change Indirect land use change Source: Copenhagen Economics 5
Short introduction to biofuel sustainability EU approach The EU has proposed a set of cropspecific factors, cf. table. While not counting towards the fuels sustainability requirement, the factors should be reflected in national reporting of GHG emissions How is such an approach likely to affect industry if implemented in sustainability requirements? Proposed factors Feedstock group Cereals and other starch rich crops Estimated emissions (g CO2eq/MJ) Sugars 13 Oil crops 55 Source: European Commission (2012), Annex VIII to the proposal for amending the fuel quality directive and the RE directive 12 6
Short introduction to biofuel sustainability Impact of factors crucial for industry GHG emissions ass. with production, transport etc. GHG emissions (g CO2- eq./mj) 120 100 80 60 40 20 0 24 26 40 43 55 37 52 58 factor Direct emissions Note: Source: The factors used are from the IFPRI study The pink horizontal line indicates the 35 per cent GHG saving minimum equivalent to 54 g CO2eq/MJ The black horizontal line indicates the fossil fuel comparator Copenhagen Economics based on RED and Annex VIII in Commission proposal 7
Short introduction to biofuel sustainability Impact of factors crucial for industry Including estimated emissions GHG emissions (g CO2- eq./mj) 120 100 80 55 55 60 40 20 0 13 12 24 26 13 12 40 43 12 55 55 37 52 58 factor Direct emissions Note: Source: The factors used are from the IFPRI study The pink horizontal line indicates the 35 per cent GHG saving minimum equivalent to 54 g CO2eq/MJ The black horizontal line indicates the fossil fuel comparator Copenhagen Economics based on RED and Annex VIII in Commission proposal 8
Issues when using models to guide policy
models and policy Precision of results not convincing Estimated factors in 8 models Feedstock factor Minimum values (g CO2 eq./mj biofuel) factor Maximum values (g CO2 eq./mj biofuel) Sugar cane -1-48 19-195 Palm oil -55-45 34-214 Sugar beet 13-33 65-181 Wheat -79-79 -8-329 Maize 5-104 44-358 Soybean 0-92 63-293 Rapeseed -33-80 52-800 Fossil fuel comparator 83.8 Source: Copenhagen Economics (2011), based on Ros et al (2010), Laborde (2011), JRC (2010), E4tech (2010), Searchinger et al (2008), European Parliament (2011) 10
models and policy Even relative ranking of feedstock differ Ranking of feedstock in different models Feedstock Type IFPRI EC Studies AGLIN K- COSIM O ESIM CAPRI ADEME Öko institut CARD GTAP LEITAP Wheat Cereal 4 5 3 3-4 1 1 1 Maize Cereal 2 4 4 2 - - - - 2 Suger cane Sugar 3 1 2-5 2 - - - Sugar beet Sugar 1 - - 1 3 - - - - Rapeseed Oil 6 3 6 5 4 5 2 - - Soybean Oil 8-1 - 2 1 - - - Palm fruit Oil 6 2 - - 1 3 - - - Sunflower Oil 5-5 4 - - - - - Note: Source: In ESIM and CAPRI total land use change from increased biofuel production is measured. In CAPRI only LUC in EU27 is included The models referred to in the table 11
Complex modelling setup A few of the steps needed in modelling Step 1: Which types of biofuels are likely to meet the increased demand in EU? Step 2: Determine the increase in feedstock production, and where on the planet this will take place o Establish crop specific supply curves and level of substitution between crops (production facilities, agricultural conditions: land, water, climate) o Based on historic patterns or simulations? Step 3: Establish carbon content of displaced land o Difficult to use average values. Scientific disagreement about e.g. tropical vs. boreal vs. temperate forests Step 4: Establish response in market for food products o What is the level of substitution between e.g. vegetable oils or sugar products? If palm oil is used to meet biofuel demand it will loose market shares in the market for vegetable oils Step 5: Go back to step 2 and 3 12
Complex modelling setup Few assumptions can move a lot Sources for variation between/within models Source Productivity of marginal land Including co-products and their substitution rate Yield increases derived determined exogenously or endogenously (based on demand) Location of land expansion and land type cultivated Response of food consumption to increases in price of food Estimated carbon stock of different land varies significantly Result If marginal land productivity is 25 pct instead of 75 pct of existing land, GHG emission predictions would increase by app 77 pct. When co-products are taken into account, it may reduce the estimated land requirements by 23-94 pct. An increase in yield productivity by 0.1 per cent per annum can reduce estimated GHG emissions by 72 per cent. Where new crop land expansion takes place is to a very large degree determined by the modeling assumptions. The IFPRI study which the Commission applies e.g assumes that all reduction in oil crops for food consumption will give rise to palm oil expansion on peat lands. This results in high for all oil crops If food consumption is very elastic the food price increases will be limited and there will be no crop land expansion. Getting the food consumption elasticity right (or wrong) will have a large impact on estimated emissions. Some models use estimated carbon stock values seven times larger than other models. This implies a seven times larger. Moreover, the carbon stock may be very different even within the same forest areas. Source: Copenhagen Economics (2011), based on DG Energy (2010), JRC (2010), Prins et al. (2010), Ros et al. (2010), CARB (2009a), CARB (2009b), Dumortier et al. (2009), McDaniel & Balistreri (2002), Keith et al. (2009). 13
Regulatory considerations
Regulatory considerations Should complex models be used to design regulatory instruments? Models illustrate size and potential of tweaking instruments. Instrument design should come from sound economic theory: shadow prices, pricing of externalities etc. Timing What to do when new and better models arrive? o Still new field; models will improve How are new production patterns taken into account? o Evidence that palm oil plantations are growing in Africa on lowcarbon soil What if carbon sink is in fact conserved? o Now and in the future? 15
Indirect way to regulate Main problem: Worried about releasing carbon sink Sustainability of biofuels in EU consumption is a very indirect way to regulate o Punishing producers for lack of regulation elsewhere on the planet Why not regulate at source? Conservation of carbon sinks o There will be no (or very little) if carbon sink is conserved 16
Conclusion Precision of modelling not good enough to guide policy instruments Not consistent regulation over time Potentially huge effect on industry Not convinced that regulation is focused on the underlying issue? 17
Tak for opmærksomheden Kontakt Martin Bo Hansen mbh@copenhageneconomics.com +45 2333 1820 copenhageneconomics.com 18