Estimating Leakage from Forest and Agricultural

Carbon Sequestration Projects

Presented byBrian C. MurrayRTI International

Presented at3rd USDA Symposium on Greenhouse Gases & Carbon Sequestration in

Agriculture & ForestryMarch 23, 2005Baltimore, MD

RTI International is a trade name of Research Triangle Institute

3040 Cornwallis Road ■ P.O. Box 12194 ■ Research Triangle Park, NC 27709Phone 919-541-6468 e-mail bcm@rti.orgFax 919-541-6683

2

Funding and Collaborators

Funding: US EPA, Climate Change Division

Collaborators: RTI: Brent Sohngen* Texas A&M: Bruce McCarl, Dhazn Gillig,

Heng-chi Lee EPA: Ken Andrasko, Ben DeAngelo

* On sabbatical from Ohio State University

3

Project Based Approaches to GHG Mitigation

Projects involve intentional activities or actions to reduce GHG’s

The product of these projects may (may not) be used to produce GHG emission offsets

Mitigation projects are voluntary, not required by law

Development of mitigation projects contain nuances that are location and sector specific

4

What is Leakage?

Leakage: Emissions that occur outside the project boundaries as a result of the project activities themselves It is caused by the shifting of emitting activity

elsewhere in response to reductions (sequestration) in the project area Spatial

Local: aka “primary” Distant: aka “secondary” or “market”

Sectoral/life cycle: GHG effects up and down the supply chain

5

* Important Point about Leakage

Leakage is only a problem if the “leaked” (shifted) emissions fall outside some accounting framework, E.g., from a capped or monitored sector or region

to an uncapped/unmonitored sector/regionFrom a monitored project to an unmonitored

activity

Otherwise, its captured in the accounting and does not undermine net emissions reduction

6

Why do we care about leakage at the project level?

It erodes the GHG benefits of a project

Can be difficult to measure

Difficult to enforce due to incomplete contracts

Potential to undermine a project-based offset system

Accounting for Leakage

8

Leakage as an issue in forestry and agriculture projects

Induced by economic forces: Supply/demand supplanted by the project is met elsewhere

Formal markets Other institutional arrangements

Leakage is not unique to forest and ag projects

But, features of forestry and agriculture make them somewhat susceptible to leakage Fixed land base: Land use change has spillover

effects Commodity markets are often broad in scope

(regional, national, global)

9

Emissions Shifting as a Spatial Concept

Project

“Secondary”

“Primary”Local shifting: observable and contractable

Regional, National, Global Markets

10

How to Address Leakage at the Project Level

Action Primary Secondary

1. Minimize leakage through project design

Expand contract to include local activity near project boundaries

Select activities that are not likely to have a lot of market leakage

2. Measure what you can’t minimize

Extend monitoring to include area proximate to project boundaries

Survey of local stakeholders

Market modeling

11

Project Leakage in a Market Context

Project

S0

S1

Q0Q1Q’0

P0

P1

Leakage

Market A (Commodity i, region x)

Market B Market C Market D

12

Estimating Leakage through Market Modeling

Simple comparative statics of individual market equilibria

L´ = R

N

C)]*1(*Ee[

C**e*100

Where e, E, γ, Φ, and Ci are market parameters

Sector modelsForest (e.g., Sohngen, Sedjo, Mendelsohn)

Forest and Ag (e.g., FASOMGHG)

CGE models

13

Myth: Leakage only happens when projects are big enough to affect the market price

Reality: Leakage can happen any time that a project involves goods and services exchanged in a market.

In fact, leakage is proportionately larger for small projects thanfor large projects or policies

Myths and Reality

14

Leakage Estimates from Market Models

International emissions leakage/energy: ~10-20% of targeted reductions are offset by leakage (from the literature)

Forest carbon leakageAfforestation Program Leakage Estimates by Region (All Quantities Are Percentages)

Region Leakage Estimate (%)

Northeast 23.2

Lake states 18.3

Corn Belt 30.2

Southeast 40.6

South-Central 42.5

Source: Murray, McCarl, Lee. 2004. Estimating Leakage from Forest Carbon Sequestration Programs. Land Econ: 80(1):109-124

15

Leakage Estimates from Market Models (II)

Forest preservation (avoided deforestation, no harvesting*)

Region Leakage %

PNW – East 8.9

Northeast 43.1

Lake States 92.2

Corn Belt 31.5

South-central 28.8

Source: Murray, McCarl, Lee. 2004. Estimating Leakage from Forest Carbon Sequestration Programs. Land Econ: 80(1):109-124

* Leakage is moderately lower if harvesting is allowed

16

Some recent leakage results comparing different forestry and agriculture activities*

Leakage Estimates by Mitigation Activity at a GHG Price of $15/t CO2

Eq. All quantities are on an annualized basis for the time period 2010–2110.

Targeted Mitigation Activities

AGHG

Effects of Targeted Payment (Tg. CO2

Eq.)

BNet GHG Effects of

All Activities(Tg. CO2

Eq.)

CIndirect

GHG Effects from

Nontargeted Activitya

(Tg. CO2

Eq.)

DLeakage

Rateb

(%)

Afforestation 137 104 –33 24.0

Agricultural Soil Carbon 154 145 –9 5.7

Biofuels 84 83 –1 0.2

aIndirect effects: C = (B – A).bLeakage rate: D = –(C/A) * 100; rounding occurs in table.Note: Negative leakage rate in D refers to positive leakage (i.e., additional mitigation outside targeted activity region).

* Ongoing work, Murray and McCarl

17

Regional dimensions of leakage *

Regional Leakage Flows for Afforestation-Only Payment Scenario: $15/t CO2 Eq.

-40

-20

0

20

40

60

80

100

120

140

GH

G M

itiga

tion

inTg

CO

2 E

q., A

nnua

lized

, 20

10–2

100

SCAfforestaion

All OtherAfforestation

SC FossilFuels/Crops

SC ForestMgmt.

SE ForestMgmt.

SC Ag. Soil CSequestration

All OtherActivities and

Regions

Afforestation Induced by Payments Leakage Effects

* Ongoing work, Murray and McCarl

18

Leakage over Time

Leakage time

horizon

AGHG Effects of Targeted

Payment(Tg. CO2 Eq.)

BNet GHG

Effects of All Activities

(Tg. CO2 Eq.)

CIndirect GHG Effects from Nontargeted

Activitya

(Tg. CO2 Eq.)

DLeakage

Rateb

(%)

10 decades 137.4 104.4 –33.0 24.0%

5 decades 170.7 129.7 –41.0 24.0%

2 decades 208.5 127.7 –80.8 38.8%

Effect of varying the time horizon over which leakage is quantified.

Afforestation program paying $15 per t CO2

19

How is leakage being handled in project accounting protocols?

WRI/WBCSD: Screening, mitigation, quantification of primary and secondary leakage are prescribed but no specific requirements in place

California Registry Draft Protocol (2004)

Leakage category Assessment Quantification

Primary (onsite) Required Required

Primary (offsite) Required Strongly encouraged

Secondary (market) Strongly encouraged Strongly encouraged

Lifecycle Required Encouraged

Chicago Climate Exchange: leakage not explicitly considered

1605(b) guidelines still in development

20

Conclusions

A project-based offsets/trading system seeks assurance that the emissions allowance correctly corresponds to the reduction by the project

For Carbon sequestration projects, the main factors that may disrupt this correspondence are Permanence Additionality Leakage

Methods are now being developed to address each of these factors, but there is debate about how far to go in terms of reporting standards, stringency, etc…

21

Conclusions (II)

Early empirical evidence suggests leakage could either be trivial (~0) or enormous (over 90%) depending on the activity, location, and time period considered

Q: Is this enough to make these investments uneconomic?

Depends on the price and on the discounts applied to other offset credits

First: design projects to minimize leakage

Centralized efforts needed to harmonize approaches to address and quantify leakage