Summary
Joint Implementation (JI) projects that capture carbon through sequestration are believed by investors to be more risky than other projects that reduce Greenhouse Gas emissions. This is because of their long lifetime, complicated nature, numerous poorly defined input parameters and perceived high costs of monitoring. Whereas the first three factors are true, sensitivity analysis can help reduce these costs by focusing one’s attention on the important parameters. Secondly, sensitivity analysis can be used to improve program design. And finally, by applying simple probability theory, one can manipulate the important variables to create a distribution of possible outcomes for the project. From this distribution, the mean and standard deviation can be calculated. These values have more meaning than a single point estimate of the project’s possible outcome.
The carbon flux model proposed by Schlamadinger and Marland (1) is used to calculate the amount of carbon sequestered by a forest management project. Using techniques developed by Strategic Decisions Group (2) the model has been extended to create tornado diagrams and probability distributions. Analysis of these data have led to focusing on estimates of important variables, an understanding of the time-value of money and the possibility of project redesign by the operating Non Government Organization (NGO).
The project used in this discussion is a forestry management program supervised by Grupo Ecológico Sierra Gorda, I.A.P. Their goal is to create a sustainable forestry practice in the Sierra Gorda Queretana, Mexico. It is a 25 year project involving seven staggered eighteen year cycles. Each cycle converts marginal cropland to forest by replanting 1000 ha/year and natural reforestation of a further 1000 ha/year. These two components of the project sequester 1.24 million tonnes of carbon (4.8 million tonnes of CO2). A forestry protection program sequesters a further 0.75 million tonnes of carbon (2.8 million tonnes of CO2) at marginal cost. The project is anomalous for a sequestration project in that it makes money and has a 13.8 % rate of return.
It is shown that the influential parameters are dependent on what happens to the wood after harvest. Parameters difficult to estimate (root & litter and soil sequestration, average lifetimes of wood products) have little effect on carbon sequestered. It suggests that effective monitoring costs could be smaller than earlier submissions suggest.
Of the parts of the project each cycle of planted afforestation has ~ 12% error, so the entire forestry portion also has ~ 12% error (144 000 MgC). The MFP program has a larger percentage error and also a larger total error (~240 000 MgC). Clearly our efforts should be concentrated here. The average area per year lost to fires can easily and cheaply be verified by reviewing satellite images, air photos, and site visits. The maximum carbon stored per hectare, also a key parameter, is more expensive to estimate.
To alleviate an investor’s concerns about risk it is better to quote the predicted amounts of carbon captured as a mean with a standard deviation than just a point estimate. It is recommended that JI projects hold at least one standard deviation in reserve as deposit insurance. These excess credits can be sold after completion of the project if they in fact exist. In this way the investor is guaranteed the carbon promised. Finally we investigate project redesign. A continuous, selective harvest might be better than the originally planned program. It sequesters more carbon, but it returns less value. A continuous harvest is preferred as it involves the community in an ongoing fashion, and creates a sustainable resource. Both assure continued survival of the project in the future.
References
Back to Woodrising Consulting Inc.'s Homepage
Last updated: 01-Apr-97
This webpage was written and designed by D. Neil Bird.
1997 Copyright © Woodrising Consulting Inc.. All rights reserved