1

Michael O’HareGoldman School of Public PolicyUniversity of California, Berkeleyohare@berkeley.edu

Land Use Change and Biofuels Policy

EEA Expert meetingCopenhagen June 10 2008

2

LCA/CTW • Considering any two exclusive policies/practices

A and B:• How will* the world be different in the future

under A compared to under B? – Things with money prices, and things not traded in

markets– Things we can easily see (tractor fuel on corn farm),

and things hard to see (soil and plant C sequestration, N cycle)

– And things we can’t see at all: faraway land use change effected through worldwide food markets

What do we count? How do we measure them? How do we weight them into a scalar measure of merit?

*Not, how might it be different under imaginable policies etc.

3

What could we do with LCA results if we had them?

• Regulatory decisions on specific fuels• Guide research and subsidy investments• Direct attention on ways to improve• Support NGO pressure and lobbying• Highlight need for related (non-GW, extra-

territorial) policy changes

• There is no unitary decisionmaking authority for any of this: “we” is a vague and mushy concept

• Different uses imply different measures and methods (as Bart & Pierpaolo pointed out).

4

“Green” in regulation

• Not a direction, hope, or general quality

• Must be expressed numerically, with– Financially consequential results

• AFCI score• yes/no acceptability

– Court review

5

LCFS liquid fuel concept

+

+ =

=

2010

2020

If this is a biofuel, how green is it?

6

LCFS in practice• For producer j in year t who blends Qi units of

fuel with GHI index Gi, the fine (or sale of credits) when the standard is St will be:

tjttjt

jt

PQAFCISC

iLUCQGQGAFCI

...2211

Policy implementation comprises (mostly) establishing operational definitions for these variables.

Direct LCA

7

LCFS in practice• For producer j in year t who blends Qi units of

fuel with GHI index Gi, the fine (or sale of credits) when the standard is St will be:

tjttjt

jt

PQAFCISC

iLUCQGQGAFCI

...2211

Policy implementation comprises (mostly) establishing operational definitions for these variables.

Direct LCA

8

Fuel

Less food,less meat

Higher Yields(intensity)

Overseas LUCDomestic LUC

Shares determined byprices and elasticities

Displaced food crops induce land use changefar from biofuel growing area

9

Time

GHG

3000

30

140

90

Ethanol withLUC

Gasoline

Physical GHGdischarge/uptake

30yr

Figure1:Physical dischargeof GHGand landusechange.Values roundedfromSearchingeret al.

+

10From Searchinger 2008

11

Net energy and net GHG estimates for 6 studies of corn ethanol, as well as 3 cases. Gasoline is shown for reference. The cellulosic case is switchgrass grown on prime crop land. Adapted from - Farrell et al, 2006

What we found without Land Use Change

12

Considering land use change (LUC)

Net energy and net GHG estimates for 6 studies of corn ethanol, as well as 3 cases. Gasoline is shown for reference. The cellulosic case is switchgrass grown on prime crop land. Adapted from - Farrell et al, 2006 and Searchinger et al, 2008

iLUC is very large for biofuels grown on cropland!

LUC with linear derating

13

Key points

• It doesn’t matter what the biofuel crop is (except for yield)

• It doesn’t matter where you grow it (as long as its on land that could grow food)

• The effect goes both ways (cropland returned to natural)

• Other processes (food demand, BAU yield increases, etc.) are secondary

• “Previously cultivated land” provisions (US, Germany, RTFO, Indonesia) are inconsequential.

14

Time

GHG

3000

30

140

90

Ethanol withLUC

Gasoline

Physical GHGdischarge/uptake

30yr

Figure1:Physical dischargeof GHGand landusechange.Values roundedfromSearchingeret al.

+

How big is this?

How long is this?

LUC for this?

How big is this?

How to count this?

How to count this?

15

GHG intensityGasoline

Prior

•Is the GHG intensity of a biofuel an RV with a PDF?

•If so, what statistic should be used for its GHG index in a regulatory context?

•What does the cost-of-being-wrong function looklike?

Bayesian posterior

P(GWI) < Gasoline

16

17

UC Current Work

• Model uncertainty in iLUC* • Model iLUC with GTAP

– Preliminary results not very different from Searchinger results

• Meta-analysis of iLUC estimates– Other models? (EPA forthcoming)

• Model uncertainty in direct LCA• Account properly for time*• Sustainability in the LCFS context*

*see slides to follow

18(Draft) Monte Carlo Analysis of Searchinger: Plevin & Jones

19

Time

GHG

3000

30

140

90

Ethanol withLUC

Gasoline

Physical GHGdischarge/uptake

30yr

Figure1:Physical dischargeof GHGand landusechange.Values roundedfromSearchingeret al.

+

How big is this?

How long is this?

LUC for this?

How big is this?

How to count this?

How to count this?

What about coproducts?

20

What about time?

• Searchinger (and others) do not discount• Discounting is a complicated issue:

– Economic discounting of events involving goods traded in markets

– “Derating” of physical phenomena: how do we value a ton of C reduction after the Greenland ice cap is in the ocean compared to a ton after? Some effects are irreversible.

• Any recognition of time value increases currently estimated deficits of crop biofuels relative to fossil fuel.

21

Time30 yr

1.0

Discounting(5%)

Calamity

Horizon

Generic

Figure2:Possiblesocial cost of physical GHGrelease functions. Conventionaleconomicdiscountingis shownfor comparison (seetext)

tc

22

Time

GHGreductionbenefit

3000

30

140

90

Physical GHGdischarge/uptake

30 yr

Figure3:Social benefit of reducingphysical dischargeof GHGincludinglandusechanges,withderatingaccordingto theGeneric function (seeFigure2).ValuesroundedfromSearchingeretal.

Simple linear derating increases i LUCcontribution to crop biofuel GW index(relative to gasoline) by about a factor of two.

23

Other considerations for crop biofuels

• Industrial monocrops• Biodiversity, economic diversity• Capital intensive, low-wage labor• Biofuel curse?• Water• Etc.

“Sustainability” comprises a variety of non-GW issues

24

“Sustainability” is [are] another whole can of worms!

Assessment of effectsAssociation with ‘batches’ of fuelLocal enforcement capacityCommensurationApplication in a regulatory environment with

real $ consequences and courtoversight

WTO rules“Goal creep”: LCFS is a GW policy

25

What’s left?

• From waste: ~8% of gasoline– Enzymes to crack cellulose– Thermal gasification + microbes + membrane

separation (eg, Coskata)– Mass burn

• Mixed perennials, oil plants on waste land• Cane, variousols • Algae: too soon to tell, but very expensive now.

Must be on desert (closed system) or open water.

26

Thanks!

Erin PalermoRich PlevinSabrina SpatariDan SperlingBrian TurnerSonia Yeh…and CMU

Alex Farrell

Mark Delucchi

CARB

Kevin Fingerman

Andy Jones

Dan Kammen