Options For Project Emission Baselines

COM/ENV/EPOC/IEA/SLT(2000)8
OECD ENVIRONMENT DIRECTORATE

AND
INTERNATIONAL ENERGY AGENCY
OPTIONS FOR PROJECT EMISSION BASELINES
INFORMATION PAPER
Organisation for Economic Co-operation and Development 2000

International Energy Agency
Organisation de Coopération et de Développement Economiques
Agence internationale de l'énergie
FOREWORD
This document was prepared by the OECD and IEA Secretariats in October 1999 at the request of
the Annex I Expert Group on the United Nations Framework Convention on Climate Change.
The Annex I Expert Group oversees development of analytical papers for the purpose of
providing useful and timely input to the climate change negotiations. These papers may also be
useful to national policy makers and other decision-makers. In a collaborative effort, authors
work with the Annex I Expert Group to develop these papers. However, the papers do not
necessarily represent the views of the OECD or the IEA, nor are they intended to prejudge the
views of countries participating in the Annex I Expert Group. Rather, they are Secretariat
information papers intended to inform Member countries, as well as the UNFCCC audience.
The Annex I Parties or countries referred to in this document refer to those listed in Annex I to
the UNFCCC (as amended at the 3rd Conference of the Parties in December 1997): Australia,
Austria, Belarus, Belgium, Bulgaria, Canada, Croatia, Czech Republic, Denmark, the European
Community, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan,
Latvia, Liechtenstein, Lithuania, Luxembourg, Monaco, Netherlands, New Zealand, Norway,
Poland, Portugal, Romania, Russian Federation, Slovakia, Slovenia, Spain, Sweden, Switzerland,
Turkey, Ukraine, United Kingdom of Great Britain and Northern Ireland, and United States of
America. Where this document refers to “countries” or “governments” it is also intended to
include “regional economic organisations”, if appropriate.
ACKNOWLEDGEMENTS
This paper was prepared by Jane Ellis (OECD) and Martina Bosi (IEA) under the guidance of Jan
Corfee-Morlot (OECD) and Jonathan Pershing (IEA). The authors thank Mark Stevens
(Australia), Satender Singh (Canada), Mette Nedergaard (Denmark), Jean-Jacques Becker
(France), Tiit Kallaste (Estonia), Murray Ward (New Zealand), Anne Johanne Enger (Norway),
Anne Arquit-Niederburger and José Romero (Switzerland) and Maurice LeFranc and Shari
Friedman (US) for the information, comments and ideas they provided. The authors are also
grateful to Debbie Stowell and Axel Michaelowa for their suggestions. Jean Cinq-Mars,
Stéphane Willems, Gene McGlynn and Thomas Martinsen (OECD), and Kristi Varangu, Richard
Baron, John Paffenbarger, Jeffery Piper and Maria Argiri (IEA) also provided advice and
comments.
Questions and comments should be sent to:
Jane Ellis
Administrator
Environment Directorate
Organisation for Economic Co-operation and Development
2 rue André Pascal
75016 Paris, FRANCE
Tel: +33 1 4524 9697
Fax: +33 1 4524 7876
E-mail: [email protected]
2
TABLE OF CONTENTS
EXECUTIVE SUMMARY .......................................................................................................................... 5
1. INTRODUCTION................................................................................................................................ 9
1.1 OBJECTIVES AND APPROACH ................................................................................................. 9
1.2 CONTEXT.................................................................................................................................... 10
2. BASELINE APPROACHES DESCRIBED..................................................................................... 12
2.1 PROJECT-SPECIFIC BASELINES............................................................................................. 15
2.2 MULTI-PROJECT BASELINES ................................................................................................. 16
2.3 HYBRID PROJECT BASELINES............................................................................................... 17
3 CROSS-CUTTING THEMES .......................................................................................................... 20
3.1 DYNAMIC VS STATIC BASELINES........................................................................................ 20
3.2 EMISSIONS TIMELINE ............................................................................................................. 21
3.3 DATA SOURCES ........................................................................................................................ 22
4 QUANTITATIVE COMPARISON OF BASELINE APPROACHES.......................................... 23
4.1 SURVEY OF EXPERIENCE WITH DIFFERENT BASELINE APPROACHES ...................... 23
4.1.1 Conversion of heat plants in the Czech Republic.................................................................. 23
4.1.2 Boiler conversions in Estonia ............................................................................................... 24
4.1.3 Other projects ....................................................................................................................... 25
4.2 SURVEY OF EXPERIENCE WITH DIFFERENT BASELINE ASSUMPTIONS .................... 27
4.3 CASE STUDY SIMULATIONS WITH MULTI-PROJECT BASELINES................................. 28
4.3.1 Electricity Generation in Brazil............................................................................................ 29
4.3.2 Electricity Generation in India ............................................................................................. 31
4.3.3 Simulation Results ................................................................................................................ 33
5 QUALITATIVE COMPARISON OF DIFFERENT BASELINE APPROACHES..................... 34
5.1 DATA, MONITORING AND REPORTING REQUIREMENTS ............................................... 34
5.2 BASELINE DEVELOPMENT COST ......................................................................................... 35
5.3 TRANSPARENCY AND EASE OF THIRD PARTY VERIFICATION .................................... 36
5.4 ENVIRONMENTAL EFFECTIVENESS .................................................................................... 37
5.4.1 Project-level environmental additionality ............................................................................ 37
5.4.2 Overall environmental effectiveness of JI/CDM ................................................................... 39
6 CONCLUSIONS ................................................................................................................................ 42
ANNEX A: POSSIBLE REPORTING REQUIREMENTS FOR BASELINES .................................. 45
ANNEX B: DATA USED TO ASSESS DIFFERENT POSSIBLE BASELINE ASSUMPTIONS IN

INDIA AND BRAZIL................................................................................................................................. 50
REFERENCES ........................................................................................................................................... 54
GLOSSARY ................................................................................................................................................ 57
3
LIST OF TABLES
Table 1: TERMINOLOGY OF DIFFERENT BASELINE APPROACHES..................................................... 14

Table 2: POTENTIAL FOR STANDARDISING DIFFERENT COMPONENTS OF EMISSION
BASELINES........................................................................................................................................ 18
Table 3: REPLACING OLD POWER PLANT WITH NEW PLANT USING RENEWABLE
ENERGIES – EXAMPLES OF SUGGESTED BASELINES AND TIMELINES ............................. 19
Table 4: VARIATION IN CALCULATED CARBON OFFSETS UNDER DIFFERENT
BASELINES: SIMULATED SWISS/CZECH PROJECTS ............................................................... 24
Table 5: COMPARING FOUR ESTIMATES OF THE EMISSION BENEFITS FROM
VILJANDI BOILER CONVERSION AIJ PROJECT......................................................................... 26
Table 6: SUMMARY AND COMPARISON OF DIFFERENT BASELINE APPROACHES......................... 41
Table A1: Estimated Energy Use and Carbon Emissions in Baseline [at time of project
registration].......................................................................................................................................... 47
Table A2: Estimated Gross Changes in Energy Use and Carbon Emissions from Project [at time
of project registration].......................................................................................................................... 48
Table A3: Estimated Net Changes in Energy Use and Carbon Emissions from Project [at time of
project registration].............................................................................................................................. 49
Table B1: C02 Emissions from Electricity Generation, Brazil ............................................................................. 50

Table B2: Cost of Electricity Generation, Brazil.................................................................................................. 51
Table B3: C02 Emissions from Electricity Generation, India ............................................................................... 52
Table B4: Cost of Electricity Generation, India ................................................................................................... 54
LIST OF FIGURES
Figure 1: REPRESENTATION OF DIFFERENT BASELINE APPROACHES ............................................... 13

Figure 2: COMPARISON OF DECIN PROJECT BASELINES ........................................................................ 27
Figure 3: EFFECT OF DIFFERENT BASELINE ASSUMPTIONS ON BASELINE LEVELS,
BRAZIL............................................................................................................................................... 30
Figure 4: EFFECT OF DIFFERENT ASSUMPTIONS ON BASELINE LEVELS, INDIA .............................. 32
Figure 5: POSSIBLE EFFECT OF BASELINE STRINGENCY AND COMPLEXITY ON
PROJECT NUMBERS AND PROJECT’S ENVIRONMENTAL ADDITIONALITY...................... 40
4
Executive Summary
The Kyoto Protocol establishes two project-based mechanisms: the clean development
mechanism (CDM) and Joint Implementation (JI). Emission baselines for JI and CDM projects
aim to quantify “what would have happened” in terms of greenhouse gas emissions in the absence
of those projects. Actual emissions from JI or CDM projects are measured against baseline
emissions, and, if lower, can generate emission credits. Baselines are, by definition, hypothetical
reference cases and are subject to a number of uncertainties.
There are differing views about whether or not CDM and JI require the same framework for
setting baselines. This paper does not attempt to prejudge the outcome or potential implications
of political negotiations on this issue. The paper focuses on the technical aspects of baselines.
Once established, rules and/or guidance for JI and CDM under the Kyoto Protocol are expected to
cover many aspects of the mechanisms and not just the baseline. The linkages between the
emissions baseline and other aspects of project-based mechanisms, such as eligibility criteria and
monitoring and reporting requirements means that the rules for one may influence others.
There are different ways to set up an emissions baseline. Ideally, baselines should be credible,
transparent, simple and inexpensive to set up. In practice, drawing up baselines is likely to
involve tradeoffs among these criteria. Moreover, the level of baseline aggregation and
standardisation can vary widely. Baselines can be established so they apply to one project only
(“project-specific”), aggregated to a sub-sector or sector level (“multi-project”), or be in a grey
area somewhere between these two levels of aggregation (“hybrid”). The literature also mentions
the possibility of aggregated nationwide baselines (“top-down”), but this paper focuses on the
three other approaches.
Project-specific baselines are the least aggregate type of emission baselines, and evaluate
emission reductions generated from one particular project (rather than a group of similar
projects). These baselines are established by using project-specific assumptions, measurements,
or simulations for all key parameters: they tend to take output levels into account implicitly rather
than explicitly. Most AIJ projects to date have used project-specific baselines when calculating
the emission benefits of those projects.
Multi-project baselines seek to standardise emission levels or rates, and are designed to be
applicable to multiple projects of a similar type. Individual projects would be measured against
these baselines to see whether or not they were eligible for emissions credits, and, if so, how
many credits they would generate. Multi-project baselines may be calculated based on
assumptions about the emissions rate (e.g. g CO2/kWh) as well as on an absolute emissions level.
Multi-project baselines can be highly aggregate and be applied to many projects, or fairly
disaggregated and applied to a smaller range or number of projects.
Hybrid baselines would be designed for projects that do not quite “fit the mould” for a multi-
project baseline, or would simplify the process of collecting and monitoring data for projects
which are unique and thus require individually tailored baselines. Hybrid baselines would be
more aggregate and standardised than project-specific baselines and less aggregate and
standardised than multi-project baselines. Hybrid emission baselines would, like project-specific
baselines, be made up of various components. But the parameter values, or the methodology used
to determine one or more of the baseline’s underlying data points, would be standardised. This
5
would mean that establishing a hybrid baseline would be a more streamlined process than
establishing a project-specific baseline.
Some cross-cutting issues are relevant to all baseline approaches. These include the length of time
emission credits can accrue, and whether or not the baseline is fixed at the start of the project
(static) or revised during the project operation (dynamic). Static baselines are predictable and
reduce the uncertainty surrounding the level of credits generated from a project. Dynamic
baselines may better reflect actual trends, but would need to be re-estimated and re-reported at
certain intervals. A number of analyses suggest that dynamic baselines may be more appropriate
for some project types.
The total number of credits generated by a project is very sensitive to the length of time over
which they can accrue (the emissions timeline). If a standard methodology to calculate an
emissions timeline could be agreed, it would increase the comparability between projects, and
also offers a potentially simple way of limiting the effects of free riders and gaming.
The approach used to determine an emissions baseline for a JI or CDM project has consequences
for the project’s transaction cost, transparency and administrative feasibility (including data,
monitoring and reporting requirements) as well as for its environmental additionality. This study
considers each of the main baseline approaches against these performance “criteria”.
Data, monitoring and reporting requirements are important because they affect the costs and
administrative feasibility of project preparation and review. These requirements vary across
different baseline approaches. Project-specific baselines have relatively heavy data requirements
and may require some monitoring of current activities before the actual JI/CDM project or
activity starts. Using multi-project baselines requires less or no monitoring of the pre-JI/CDM
project situation for project participants. However, data are required to establish these baselines.
Different baseline approaches also have different cost implications1. Using a baseline approach
that incorporates either standardised methodologies or assumptions will be cheaper and easier
than developing a project-specific approach. While there are costs involved in developing all
baselines, who pays for their development may differ depending on the approach. By lowering
the costs of project preparation, an agreement on standardised approaches could increase the
number of JI and CDM projects. In turn, this could help increase the effectiveness of JI and
CDM by increasing their contribution to cost-effective emission abatement.
The transparency of a baseline also varies with different baseline approaches. In general, the
more assumptions related to an individual project or to a system that are included in a baseline,
the more documentation is needed with that baseline to make it transparent. Increased
transparency may help to increase participation, and may also facilitate any verification and
certification.
The environmental additionality of a JI/CDM project can be affected by the baseline approach as
this can influence the potential level of gaming, free riders and leakage. Increased levels of
gaming, free riders and leakage would artificially inflate the number of credits resulting from a
project. For the CDM, this would mean that not all credits accruing from CDM projects would
represent actual emission reductions and may lead to higher total Annex I emissions. Artificially
1
The cost of establishing an emissions baseline is one component of the transaction costs
associated with JI and CDM projects, but should ideally be kept as low as possible to encourage
investment through these mechanisms.
6
high credit levels for JI projects will not affect the overall Annex I assigned amount2, although it
could make it more difficult for some host countries to meet their individual Kyoto commitments.
The potential for gaming may be particularly high for CDM project-specific baselines. Multi-
project baselines may be more vulnerable to potential free riders in some cases, although the level
at which the baseline is set – and this is true for all baseline approaches- is crucial in this regard.
The environmental additionality of individual projects is clearly correlated to the level or

stringency of the baseline. The stringency of an emissions baseline varies depending on the
assumptions used in setting it up. This paper presents a preliminary survey of analyses done on
the effect that different baseline approaches and/or assumptions have on the level of credits
generated by a particular project. This paper also presents case studies that illustrate the influence
of national circumstances on the absolute level of the baseline and the relative level (compared to
other technologies/processes). For example, if both India and Brazil use the same multi-project
baseline approach to set a baseline for new electricity projects at the level of their current average
emissions, gas-fired electricity projects could generate certified emission reductions (CERs) if
they were undertaken in India, but not in Brazil.
The assessments presented in this paper indicate that assumptions used to develop baselines
and baseline approaches (e.g. project-specific, multi-project) are independent variables.
Both can influence the level of credits for a particular project. In the examples examined
here, the range in different possible assumptions within one baseline approach can be as large as
the range in assumptions between different baseline approaches. However, not all approaches
may be equally appropriate in all circumstances, and different approaches may be viewed more or
less suitable for different types of projects (e.g. forestry, electricity or landfill).
Both baseline assumptions and approaches will have an impact on the overall
environmental effectiveness of the mechanisms3. The baseline approach will influence the
effectiveness of the mechanism through its impact on the complexity of setting up an emissions
baseline. All other things being equal, more complex methods are likely to limit the number of
projects initiated by adding to the transaction costs associated with developing a CDM or JI
project. Within each approach, baseline assumptions are also likely to have an impact on the
effectiveness of the project-based mechanisms through their impact on the baseline stringency,
which affects the level of credits, and through this the number of projects initiated. Regardless of
the approach used, stringent baselines could limit the number of projects initiated due to their
effect on the cost of credits while lax baselines could obviously have a negative environmental
effect.
In circumstances where different baseline approaches are plausible, the independence of baseline
stringency and approaches suggests that maximum environmental effectiveness across the
project-based mechanisms (as opposed to individual projects) is likely to be achieved by
optimising baseline stringency and reducing baseline complexity. In practice, this means: (i)
seeking to minimise baseline complexity, as long as the ability to determine "what would have
happened otherwise" is not compromised; and (ii) optimising the baseline stringency so that it
2
As long as each Annex I country meets its emissions commitment.
3
Unlike the CDM, JI is a zero-sum game in which transfers and acquisitions of emission credits
will not affect the total Annex I emissions, as allowed by the Kyoto Protocol. The implications
on environmental effectiveness are thus different.
7
maximises the overall global environmental effectiveness from the project-based mechanisms.
The optimal strategy takes into account that a high volume of projects will be needed to deliver
strong environmental effectiveness from the mechanisms: a greater number of good projects will
be more beneficial for the environment (in terms of total GHG reductions) than a lower number
of individually better projects.
Progress on technical issues would be greatly facilitated if policy makers decided which baseline
approach(es) are to be used for each JI and CDM project/project type. Such an agreement would
allow analysis to focus on how such approach(es) could be applied in a comparable manner or, in
the terminology of this paper, on finding assumptions that provide the balance between the
environmental effectiveness and encouraging participation given different circumstances. An
agreement would also facilitate the resolution of other outstanding questions (e.g. reporting
formats) which could also help in the development of the mechanisms. In addition,
understanding of the issue would be improved if common definitions and a common vocabulary
for key baseline-related parameters could be reached.
A decision on baseline approach(es) would open the door for assessment and eventual agreement
on other credit-related aspects of JI and CDM projects, such as:
• how to determine the value of key assumptions;

• how long any project should be allowed to generate emission credits for (the emissions
timeline);
• whether or not emissions baselines should be static or dynamic;
• maximising the environmental effectiveness of the project-based mechanisms by minimising
the level of leakage, free riders and gaming; and
• assessing how to determine the environmental additionality of a climate-friendly project in a
country that has low emission levels.
8
1. Introduction
The Kyoto Protocol (KP) establishes two project-based mechanisms: the Clean Development
Mechanism (CDM) and Article 6 projects (usually referred to as Joint Implementation, JI).
Article 12.5(c) of the KP states that CDM project activities should result in “reductions in
emissions that are additional to any that would occur in the absence of the certified project
activity”. Article 6.1(b) of the KP states that JI projects should provide “a reduction in emissions
by sources, or an enhancement of removals by sinks, that is additional to any that would
otherwise occur”.
Estimating business-as-usual reference scenarios (i.e. what would happen otherwise) is thus
necessary to evaluate the environmental “additionality” (in terms of emission reductions) of JI
and CDM projects. These reference scenarios are defined as “baselines”.
The international community has not yet decided how these baselines should be established,
reported and verified. However, baselines are an important issue within the ongoing negotiations
on the Kyoto Protocol to the UNFCCC.
Activities Implemented Jointly (AIJ) was established in 1995 at the first Conference of the Parties
(COP1) as a pilot phase of Joint Implementation. Some experience with emission baselines has
been gained during the AIJ pilot phase. This experience is mainly with emission baselines that
have been established on a project-specific (case-by-case) basis. However, these emission
baselines have been relatively expensive and time-consuming to draw up. Lack of international
guidance on how exactly to draw up these emission baselines has also meant that they are often
neither transparent nor consistent.
There are other possible options by which to develop emission baselines, and many countries
have expressed an interest in examining in more detail how such options would function and what
would be their implications. This paper seeks to define and explain the different baseline
approaches and to examine how the various approaches can impact the assessment of project
environmental additionality and the overall environmental effectiveness of the project-based
mechanisms. A survey of experiences with baseline approaches under AIJ, as well as quantitative
case studies, are included to illustrate the potential use and implications of different baseline
approaches.
1.1 Objectives and approach

The objectives of this paper are:
• to define and assess how different baseline approaches could function for JI and/or for CDM
projects;
• to determine whether the use of a particular baseline approach (over another approach) could
influence the evaluation of environmental additionality of a particular project; and,
• to determine whether and how the use of a particular baseline approach could influence the
incentive to participate in CDM and JI projects (i.e. the potential effect on the number of
projects).
9
This assessment is then used to examine the potential effect of baseline approaches on the overall
effectiveness of the project-based mechanisms4 in helping meet the objectives of the Kyoto
Protocol, i.e. contributing to sustainable development5 and helping Parties achieve their Kyoto
commitments most cost-effectively. In addition, this paper outlines the potential data, monitoring
and reporting requirements for each baseline approach, the likely cost implications of these
requirements, and assesses the applicability of each approach to different project types.
Implications of using different baseline approaches in the verification process of emission
reductions from CDM and JI projects are also examined.
This paper assesses6 different baseline approaches under a number of criteria including:
• the environmental credibility of the baseline (which is related to the potential levels of
gaming, free riders and leakage);
• the relative costs of developing and using the baseline;
• the data, monitoring and reporting requirements of the different baseline approaches; and
• transparency.
The variation in these criteria is likely to influence the transaction costs associated with a project,
the number of projects initiated under the project-based mechanisms, each project’s
environmental additionality, and therefore, the environmental effectiveness of JI and CDM as a
whole7.
The paper also looks at the potential implications of the assumptions used to develop baselines
(using different baseline approaches) on the level and stringency of the baseline. In addition, the
paper provides an initial assessment of the relationship between the assumptions and the different
baseline approaches as well as between the stringency and the complexity of baselines.
1.2 Context
Emission baselines for project-based activities aim to quantify “what would have happened” in
the absence of those activities. Baselines are estimated reference cases for likely future emission
pathways, and are therefore hypothetical8. If actual emissions from JI or CDM projects are lower
than the baseline, they are viewed as additional9 and can be used to generate emission credits.
4
As explained later in section 1.2, there are different views on whether or not baseline-setting
framework should be the same for JI and CDM.
5
Article 12.2 of the Kyoto Protocol explicitly states that one of the purposes of the CDM is to
assist non-Annex I Parties (i.e. the developing countries) “in achieving sustainable
development”.
6
This assessment is in section 5.
7
The implications on environmental effectiveness may be different for JI and CDM.
8
Even an emission baseline based on factual data is hypothetical because it is based on the
hypothesis that future and past performance will be the same – which may or may not be the
case.
9
In this paper, “additional” is defined solely based on the impact of the project on GHG
emissions. No financial criteria are taken into account though this is another important area of
consideration.
10
It is important to note that there are different views on whether or not rules and guidelines for
setting emissions baselines for JI and CDM projects should be the same. Whereas JI can be
viewed as a “zero-sum” exercise where the total assigned amount for Annex I countries does not
change as a result of the transfer and acquisition of ERUs, the CDM can be viewed as a “plus-
sum” exercise.
Global emissions remain unaffected by ERUs from JI projects (as long as the JI host country
meets its Kyoto commitments), regardless of the uncertainty surrounding estimated
environmental additionality. On the other hand, CERs generated from CDM projects are added
to the Annex I assigned amount. The inherent uncertainty associated with the estimation of CERs
resulting from CDM projects can lead to positive or negative impacts on the level of global
greenhouse gas emissions. For these reasons some Parties are of the view that CDM projects and
JI projects should be treated differently. In this view, rules about baseline-setting would be
appropriate for CDM projects, while only guidance about baseline-setting would be necessary for
JI projects. Other Parties, however, do not support such a distinction between the two project-
based mechanisms. As the Kyoto Protocol (Articles 6 and 12) states that emission reductions
resulting from a CDM or JI project must be “additional to any that would” occur in the absence of
that project, these Parties are of the view that both mechanisms should have a consistent treatment
of environmental “additionality.” Emissions baselines are important to this debate because they
are used to assess environmental additionality.
While this paper notes these differing views and their potential implications, no attempt is made
to prejudge the outcome of political negotiations on this issue. Instead, the paper focuses on
technical aspects of baselines, their relationship with environmental additionality, participation in
the mechanisms (i.e. potential number of projects), and more broadly with environmental
effectiveness. Environmental effectiveness, while recognising that implications may be different
for JI and the CDM, is defined in the paper by the environmental additionality of each individual
JI and CDM project, (compared to what would have happened without such projects), and the
likely number of projects undertaken.
Leaving aside the debate on whether or not setting CDM and JI baselines require the same
framework, an ideal emissions baseline should:
• be environmentally credible (to ensure long-term benefits greater than what would happen
otherwise);
• be transparent and verifiable by a third Party;
• be simple and inexpensive to draw up (low transaction costs); and
• provide a reasonable level of crediting certainty for investors.
In practice, any baseline approach is likely to involve tradeoffs among the criteria above.
If the baseline level is higher than the emission level that would “happen otherwise”, an
artificially high number of emission credits would be generated per project. A high baseline level
would also result in more projects being eligible for JI/CDM status and could increase the number
of free riders as well as lower the cost of emissions credits. On the other hand, if the project
baseline is very stringent and set at a level lower than the emission level that would “happen
otherwise”, the amount of emission credits per project would be artificially small. A low baseline
will by definition limit the number of eligible projects, and may even disqualify some potentially
11
climate-friendly projects. A low baseline is also likely to reduce the number of profitable projects
and to increase the cost of emissions credits from those projects that do qualify10.
Uncertainties in emission baselines cannot be eliminated as it is not possible to prove what would
“happen otherwise” in these hypothetical reference cases. However, there are concerns about the
potential incentives to create greater inaccuracies in the baselines, e.g. by biasing them upward in
order to generate a greater number of credits for the CDM/JI project participants (i.e. “gaming”).
Apart from the impact of the baseline level, this paper considers the impact of different baseline
approaches and assumptions on the environmental effectiveness, as defined above, of the CDM
and JI. The analysis recognises the need to strike a balance between maximising the
environmental credibility of the baseline and minimising the costs of developing the baseline in
order to encourage participation and a high level of investment in emission-reducing projects.)
2. Baseline approaches described

There are many different ways by which emission baselines could be set up for JI and CDM
projects. This paper assesses three main types of options:
• project-specific: baseline determined on a case-by-case basis, with project-specific

measurements or assumptions for all key parameters;
• “multi-project” (which can be at many levels of aggregation such as by technology or
sector): baseline is equivalent to an “activity standard11” or policy target that is aggregated at
a certain level; and
• hybrid: baseline determined in a hybrid fashion, with some key parameters project-specific,
and others standardised (the number and level of the standardised parameters will vary for
each different project category).
These baseline approaches are outlined in Table 1.
The literature also mentions baselines drawn up at the national level, referred to as “top-down”
baselines (e.g. Puhl 1998). These top-down baselines are highly aggregated, reflect national
government objectives/targets and policies and have therefore been most often referred to in the
context of JI or AIJ projects. These baselines could be used to assess emission reductions
resulting from policy initiatives (e.g. Puhl et al.1998) and not only concrete project activities, as
is the case for the other baseline approaches. Because of this fundamental difference with the
other baseline approaches considered in this paper, top-down baselines12 are not treated further in
this paper.
There are many possible variations to the three baseline groups assessed in this paper. For
example, the time over which emission credits can accrue (the “emissions timeline”) can vary.
10
Estimating whether or not a hypothetical baseline was overestimated (lax) or underestimated
(stringent) is likely to be difficult, uncertain and best carried out only several years after the start
of the project (e.g. if the assessment is to be based on data regarding comparable, but non-
JI/CDM, projects).
11
This type of baseline includes those variously referred to as “benchmarks”, “activity standards”,
“project-category”, “technology matrix” and “sectoral”.
12
In some cases, top-down baselines refer to sectoral baselines based on national data. In this
paper, such baselines would fit under the “multi-project” typology.
12
These, and other cross-cutting issues, such as whether or not baselines should be static or
dynamic, are discussed in section 3 of this paper.
Figure 1: Representation of different baseline approaches
Project-specific Hybrid Multi-project

(many components, (some components, (key assumptions
key assumptions some assumptions standardised and
project-specific) standardised/ aggregated, e.g. to one
aggregated) figure)
Standardisation/aggregation
Terminology and baseline approaches
The vocabulary surrounding emission baselines is under development. At present, there appears
to be no international consensus either on terminology, or on the definition of baseline categories.
In some cases, one term is used to describe more than one concept. For example, some use
“dynamic” baselines to refer to a baseline that is not at a constant level throughout the crediting
life of the project, e.g. because of projected changes in output from that project. Some use it to
mean a baseline that has been retroactively altered. Others use the same term to refer to a
baseline that is revised during the course of a project. This paper uses the latter definition for
dynamic.
In other cases, there are many different labels describing one concept, such as using one
aggregate factor to determine the emissions credits from a project. In this case, the aggregate
factor has been variously referred to as an activity standard, benchmark, project-category, or
intensity indicator. This paper uses “multi-project”.
In addition, the approach used to determine a baseline’s level can be distinguished in different
ways, e.g. by level of aggregation, whether underlying data were historical or projected, and to
what extent models were used. Since many different data points are needed to establish a
baseline, and since each of these data points can vary in level of aggregation etc. any
categorisation is likely to be blurred. This paper distinguishes baseline approaches by the level of
aggregation and standardisation.
13
Table 1: Terminology of different baseline approaches
Baseline approach Description Calculation of annual Credits

Project-specific Baseline determined on a case-by-case basis, with Difference between measured project emissions and
project-specific measurements or assumptions for key estimated baseline emissions.
parameters.
Multi-project (various designs):
- technology level Baseline emissions are specified per technology, e.g. on Difference between measured project emissions and
a rate basis such as t CO2/GWh. inferred baseline emissions.
- (sub-)sector level Baseline is equivalent to a “performance standard” (or Projects/activities would only qualify for credits if
intensity indicator) that is aggregated at a certain level emissions (per unit activity or output) were under
(e.g. sub-sector X in country Y, or sector P in country the performance standard. Credits would be based
region Q). on the difference between project emissions and the
performance standard.
Hybrid Baseline determined in a hybrid fashion, with some key Difference between measured project emissions and
parameters project-specific, and others standardised (the estimated baseline emissions.
number and level of the standardised parameters will
vary for each different project category).
14
2.1 Project-specific baselines

Project-specific baselines are the most disaggregate type of emission baselines. They are
developed to evaluate emission reductions generated from one particular project (as opposed
to a group of similar projects). These baselines are established by using project-specific
assumptions, measurements, or simulations for all key parameters, such as fuel and
technology characteristics and changes in output/fuel/technology over the life of the project.
The project-specific baseline approach was the most commonly used in AIJ pilot phase
projects, and is therefore the baseline approach for which there is most experience. These
baselines were drawn up on a case-by-case basis for a particular project. For project-specific
baselines, each key component used to establish the emissions baseline is individually
assessed to determine what would have happened in lieu of that particular project. These
different data components of a project-specific baseline may be measurements, estimations,
simulations or assumptions, and may be unique to a particular project (e.g. distance from
current electricity or gas grids) or applicable to a group of projects (e.g. the technical lifetime
of the equipment installed).
Despite the project-specific nature of this baseline approach, it may still be subject to
considerable uncertainty and thus lead to uncertain estimates of the “environmental
additionality” of JI or CDM projects. One analysis (Begg et. al. 1999) indicates that using a
project-specific baseline to estimate emission reductions from refurbishment-type energy
sector projects results in uncertainties of the order of + 80% - and these type of projects are
those usually thought of as the “easiest” projects for which to draw up emission baselines!
The same analysis indicates that the largest component of this uncertainty is the hypothetical
nature of the baseline, i.e. the choice and timing of baseline fuel and technology options.
Most project-specific baselines in AIJ projects were established using measurement and
expert judgement for key parameters, though model simulations or comparisons were also
used to a lesser extent. These baselines were often drawn up by a consultant acting on behalf
of the investor, before being reviewed by a team in the investor’s country and agreed to by the
host country. Because of the lack of rules or guidance governing what needed to be measured
and/or reported in order to set up an emissions baseline, AIJ project reports did not include a
consistent set of data in their elaboration of the emissions baseline (Ellis 1999). Moreover,
the detail of data and underlying assumptions presented in the project reports submitted to the
UNFCCC varied significantly between different AIJ projects. The costs of establishing
tailored emission baselines for some JI-type projects have been estimated at 1-8% of total
project costs (Puhl 1998).
Project-specific baselines could be made more consistent and transparent than those used to
date. This would be achieved if each project type had a uniform reporting format in which to
report specific information (such as fuel and technology characteristics) from that project
type. This could encourage an agreement on methods on how to calculate emissions
baselines. Such a baseline could then be classified as hybrid or multi-project in some cases13.
Data requirements for determining tailored emission baselines vary by project type. Initial
reporting requirements for selected project types have already been suggested by some
analysts (e.g. MacDicken 1997, Vine and Sathaye 1999, Ellis 1999) and are listed in Annex
A. These suggested reporting requirements are quite data intensive, and may in some cases
13
The current Uniform Reporting Format for AIJ projects does not request the reporting of
specific data, but “estimated emissions without the activity (project baseline)” and
“estimated emissions with the activity”. Consistent methods would result in greater
transparency irrespective of baseline type, and are discussed further in section 2.5.
15
necessitate monitoring and reporting fuel and/or technology use over an extended period of
time before the project is in place, as well as during the life of the project. For example, the
current Dutch approach to establishing a baseline requires information “for twelve successive
months prior to the start of the project” (JIRC 1997). The tables in the UNFCCC’s current
Uniform Reporting Format imply that annual monitoring and reporting of project emissions is
required for AIJ projects.
Project-specific emission baselines are elaborated on a case-by-case basis and could, by

definition, be used for all project types in all sectors.
2.2 Multi-project baselines

Multi-project baselines14 indicate the emissions level associated with certain activities, often
at a sectoral or sub-sectoral level. Potential CDM or JI projects would be measured against
these aggregate baselines to see whether or not they were eligible for emissions credits, and, if
so, how many. In the energy and industry sectors15, these baselines would probably be based
on an emissions rate (e.g. t CO2/GWh) rather than on an emissions level16. In comparison,
project-specific baselines are usually, but not always, based on an emissions level.
Multi-project baselines are an extremely broad category of baseline. They could encompass
baselines drawn up at differing levels of geographical or sectoral aggregation, and baselines
using each level of aggregation could be based on historical data, trends or projected data. At
a disaggregate level, e.g. specific to a particular technology, sub-sector and country, multi-
project baselines can approach the level of detail of a hybrid project baseline17. Alternatively,
highly aggregate project categories could be drawn up, e.g. sectoral level. The AIJ pilot
phase includes at least one project (cement manufacture in the Czech republic18) where the
emission benefit was calculated by using a sectoral baseline.
Multi-project baselines are potentially simple, transparent, predictable and low-cost. They
may also reduce “micro” gaming (gaming at the project level) compared to other baselines, if
they were developed by a (neutral) third party.
Opinions differ on whether multi-project or other baseline approaches give a more plausible
estimate of the hypothetical emission level that would occur without the JI or CDM project.
Multi-project baselines are defined and applied to a group of projects. This means that the
environmental additionality of each project credited under a multi-project baseline has not
been assessed on what would have happened in lieu of that particular project, although multi-
project baselines can nevertheless represent a plausible estimate of the overall situation that
would have happened otherwise. Ultimately, projects assessed against multi-project baselines
can generate the same number, greater or fewer emission credits than if they were assessed
14
Many different alternative labels for this type of approach have been used, including
“benchmarks”, “intensity indicators”, “activity standards”.
15
Little work has been done to date on the possibility of using benchmarks for biotic projects.
However, some initial thinking on the issue (see Trexler 1999a) indicates that a binary
(yes/no) determination of additionality may be more appropriate than a rate-based
determination, at least for some LUCF sub-sectors.
16
Using a baseline based on an emissions rate allows changes in output to be taken into
account.
17
Further disaggregation is possible, e.g. for multi-project electricity sector baselines which
could separate fossil/non-fossil sources, peak/off-peak etc.
18
See http://www.unfccc.de/program/aij under the Cze/Fra project
16
under project-specific baselines, depending on the assumptions used for the two baseline
approaches.
Multi-project baselines could be set up at a fairly disaggregate level, such as by technology

in a particular country (or part of a country). In this case, emissions from e.g. a gas-fired
plant installed as a JI/CDM project would obtain emissions credits if the per kWh emissions
were lower than that designated for that project category.
Of course, the key issue would be where to set the baseline level. Should this level reflect the
host country average; the regional average; the country or regional average for recently
installed technology; the marginal technology; or the best equivalent system already installed
in the host country? Alternatively, should the baseline level be determined by the best
economically attractive system? These different choices in assumptions underlying the
emissions baseline have a significant effect on the number of credits that could be generated
for a project, and are illustrated in section 4.3.
However, once the exact technology characteristics that will form the baseline have been
chosen, the baseline could be relatively simple to establish, for example if it was determined
relative to the emissions performance of best available technology (BAT), e.g. 10% greater
emissions than BAT.
Alternatively, aggregate multi-project baselines could also be drawn up, e.g. at a sector level.
As for more disaggregate multi-project baselines, they could be based on historical data or on
projected performance of the sector.
2.3 Hybrid project baselines

Project-specific emission baselines are made up of many different components. Standardising
the value of one or more of these components, or the method by which they are elaborated,
would increase the transparency and comparability between the emission baselines of
different projects and would also reduce the time and cost of establishing an emissions
baseline19.
“Hybrid” project baselines could therefore be designed to be applicable to a small number of

projects. These baselines would therefore include a greater degree of standardisation for one
or more of the key baseline components than project-specific baselines, but would be less
aggregated and less standardised than e.g. multi-project baselines designed for particular sub-
sectors. The local and regional variability of the different baseline components will determine
the extent to which standardisation is feasible. For example, a JI/CDM project that reduces
leaks of methane from a natural gas distribution system is likely to need to use a project-
specific value for leakage rate as this rate likely to be specific to the system concerned.
Whether or not a component used to calculate emission baselines can be easily standardised
depends on the component itself. These are examined in Table 2. The components that are
bolded are those likely to have the greatest impact on a baseline’s level.
19
A hybrid baseline would not necessarily result in similar projects generating identical
emission credits because of project-specific variations in certain parameters and countries’
different circumstances. Nevertheless, it could reduce the wide divergence of emission
reductions for similar projects undertaken in different circumstances that was notable for AIJ
projects.
17
Table 2: Potential for standardising different components of emission

baselines
Baseline Project types for Potential for Comment

component which assumption standardisation
is important
All High Standardised methodologies could be agreed on how
to determine the timeline (standardising the actual
timeline length would be more difficult, although
possible at a disaggregate project category level).
These methodologies could be further subdivided by
project category, and by project type (e.g.
greenfield/replacement). It may also be possible to
obtain political agreement on the absolute timeline,
such as 10 years for power plants.
Baseline fuel “Greenfield” and Medium-high Some standardisation may be possible for some
used refurbishment (short-term). countries (e.g. those with a predominant fuel use in a
energy/industry Low (longer particular sector and limited changes in fuel trends).
projects term). Longer-term fuel use more difficult to determine
because of the possibility of autonomous fuel
switching (e.g. coal/oil to gas).
Autonomous D/S-side energy Medium Arbitrary efficiency improvements could easily be
energy efficiency factored into baseline/credits, although different
efficiency projects. Fuel factors may need to be elaborated for different
measures transportation countries/sectors.
Accounting S/D-side projects Medium/High Standardisation could be easy in principle (e.g. by
for energy- that reduce using average C/kWh values for displaced
efficiency demand for fossil electricity)20, but difficult to determine at the start of
measures fuels, or that a project because of the influence of a changing fuel
reduce mix. Lack of data may also prove to be a barrier to
demand/displace determining average values. Values may need to be
fossil electricity “deflated” in order to take any rebound effect into
account. The actual effects of energy efficiency
measures may, however, deviate significantly from
“average”.
Fuel emission Energy/industry High (for CO2), Default emission factors (e.g. as in the IPCC
factors projects Low/Medium guidelines) could be used for CO2 from fuel
(other gases) combustion. Standardisation more difficult for other
gases/sectors, often being site or region specific (e.g.
CH4 from mining, landfills or rice production).
Carbon Biotic projects Medium Site-specific differences and natural variabilities
content result in variations in the order of + 25% (WBGU
1998).
Sequestration Forestry projects High Best used in local or regional areas.
rates
Host country All N/A (except at a May be a critical component for baselines in some
policies country level) sectors.
Development of standardised components for hybrid or multi-project emission baselines

might best be carried out by experts working within the framework of an accepted process
(such as the process used by the IPCC/OECD/IEA inventory programme to elaborate methods
for calculating national emission inventories).
20
Average national values for this could be calculated by using national or international
statistics on electricity productions and associated carbon emissions (see e.g. IEA 1998c and
1998d).
18
Michaelowa (1998) makes some recommendations for methods on how to calculate hybrid
emission baselines for some project categories. These include methodological
recommendations for some energy supply project types, summarised in Table 3.
Table 3: Replacing old power plant with new plant using renewable energies -
examples of suggested baselines and timelines
Baseline Emission Project lifetime Case in which baseline applies

reduction
Efficiency level of old E (old) Commercial life of Plant still operating
plant old plant economically
(approximate value:
depreciation value)
Average level of Commercial life of Commercial life of old plant
efficiency in new plant E(mean) new plant has run out, reference is
operating on old fuel (approximate value: previous fuel since autonomous
built in the last 5 years depreciation period) fuel substitution cannot be
assumed
Efficiency level of the E (old) 10 years Transitional solution for out-
old plant dated power stations; to start 3
years at the latest after the JI
mechanism comes into force.
Source: Michaelowa 1998
The level of emission credits is highly dependent on a few key assumptions (e.g. timeline,
fuel use, technology type). Reaching agreement on adequate default value(s) for these
assumptions is crucial, and may be difficult, as there are likely to be different views that will
need to be considered.
19
3. Cross-cutting themes
There are certain issues of relevance to many different baseline options. These include:
• whether or not the baseline should be re-estimated over the life of the project (i.e. whether
the baseline is “static” or “dynamic”);
• the emissions timeline (the length of time over which a project can generate emissions
credits); and
• how data are used to calculate the emissions baseline are generated (i.e. whether data are
historical, observed by comparing against a control group, or simulated/projected e.g. via
models).
These cross-cutting issues are discussed in this section. Other cross-cutting issues, such as
whether or not a distinction should be made between refurbishment and “greenfield” projects,
the importance of the renewal rate of technology, implications of inclusion/exclusion of non-
CO2 greenhouse gases in emissions baselines and pre-combustion emissions, are not
discussed here.
3.1 Dynamic vs Static baselines

Emission baselines can either be fixed at the start of the project for the lifetime of the project
(“static”) or revised during the project operation (“dynamic”). Static baselines are
predictable, and therefore reduce the uncertainty surrounding the level of credits generated
from a project. They therefore give investors a greater level of certainty than dynamic
baselines. Static baselines are also less of an administrative, monitoring and reporting burden
than dynamic baselines because they require only one estimate of a baseline. They thus result
in lower transaction costs. Some Parties explicitly favour static baselines, at least for certain
project types (e.g. DISR 1999).
Dynamic baselines will need to be re-estimated (and therefore re-reported, and the amount of
future expected emission credits revised) at certain intervals during the project’s life. This re-
estimation will allow the baseline to reflect more closely changing “best estimates” for the
key parameters used in drawing it up. Dynamic baselines may therefore reflect more
accurately than a static baseline what would have happened in the absence of the project.
Dynamic baselines may therefore ensure the continuing environmental additionality of a
project more consistently than static baselines, as they would enable a baseline to be adjusted
downwards if the environmental performance of the sector/process improves. Dynamic
baselines result in a greater level of investor uncertainty in the number of credits for a
particular project than static baselines. However, this uncertainty can be reduced if investors
know exactly when, after what time interval, and upon what factors the baseline is to be re-
calculated21.
Both policy and technology factors could lead to the need for baseline changes. For example,
Costa Rica has announced that they would phase out non-renewable sources of electricity
supply by 200122. If this policy is successfully implemented, average per kWh emissions will
decrease to zero (within the lifetime of any proposed AIJ/CDM electricity supply project).
An electricity project proposed before the policy announcement may have assumed a
21
For example, if investors know that a baseline is fixed for eight years, and subsequently
recalculated, this dynamic baseline may nevertheless provide sufficient certainty to
encourage investments. However, if a baseline may be recalculated at any point during the
project’s lifetime, uncertainty will be much greater, and investment likely to be less.
22
It is unclear, however, if this objective will be achieved even without CDM projects.
20
continued proportion of fossil fuels in electricity supply. Such a baseline would be

significantly higher than one taking the policy into account.
Technological improvements or process changes can also be relatively sudden and

unpredictable. Such changes could make even relatively recent emission baselines outdated.
For example, recent process changes in adipic acid production have resulted in a ten-fold drop
in the N2O emission factor (although some allowances might need to be made for a lag-time
for existing plants to be retrofitted). In both cases, a dynamic baseline would help maintain
the environmental effectiveness of projects that had previously been initiated in these sectors
by adjusting the baseline during the course of the project to one that is a more realistic
projection of what would have otherwise occurred.
Most experience to date with emission baselines for project-based activities is with static
baselines, which were used in the majority of AIJ projects. However, at least one AIJ project
has a dynamic baseline: the project on cement production in Cizkovice (see the France/Czech
project under http:/www.unfccc.de/program/aij) has a baseline that will be re-estimated after
the first five years of project operation. In addition, many of the Swedish AIJ projects revised
their (static) emission baselines between the first and subsequent reports to the UNFCCC.
Some AIJ projects, such as the Romania/Swiss one, used a combination of (static) baselines23.
A number of analyses have suggested that dynamic baselines may be appropriate for some
project types. Begg et. al. (1999) suggest that the best way of dealing with baseline
uncertainty is to check and revise the baseline every 8-12 years. They clarify that any
revisions should apply to the remainder of a project’s lifetime only, and should not be applied
retroactively. Ernst Basler & Partners suggest that either a dynamic baseline, with
reassessment after approximately 8 years or a short timeline is needed in projects that require
long-term assumptions about a system’s electricity production. No analysis of the cost
implications arising from the use of dynamic versus static baselines has been undertaken – but
clearly, the responsibility for reassessing credits during the life of a project will add to the
transaction costs.
Both static and dynamic baselines can be set up in such a manner as to make them
environmentally effective. The effectiveness of a baseline is not only determined by the level
at which it is set (which in turn is influenced by which baseline option is used), but the length
of time over which it allows credits to accrue to the investor. The emissions timeline is
assessed in the following sub-section.
3.2 Emissions timeline

The total number of emissions credits generated by a JI or CDM project will be extremely
sensitive to the time during which emissions credits are allowed to accrue. The investing
party in both JI and CDM projects will therefore have an interest in obtaining the longest
possible “emissions timeline” in order to maximise the emission reduction units (ERUs) or
certified emission reductions (CERs) from their investment. Given the absence of emission
commitments for CDM host countries, many potential CDM hosts may wish to accept a
timeline proposed by the investing party, particularly as a longer timeline may increase the
likelihood of CDM investments occurring in their country. However, JI host parties may
favour shorter timelines that would limit the number of ERUs transferred.
The AIJ projects examined in a previous paper (Ellis 1999) had widely differing emission
timelines, even for projects of a similar nature. Creating a standard methodology to calculate
23
Available at http://www.admin.ch/swissaij/in_meth_baselinestep.html.
21
an emissions timeline would have many advantages. It would increase both the transparency
of emission baselines and the comparability between projects. A standardised methodology
would also reduce the time and costs of setting up an emissions baseline, as the method would
be one factor that the developer would not have to develop themselves. If this standardised
methodology was conservative (stringent), it could also help to ensure the environmental
integrity of the projects by aiming to err on the side of caution.
However, there are many different views on how an emissions timeline could be set. These
include:
• numerical limits, e.g. 5 years for energy-sector JI-type projects (NEFCO, quoted in Puhl
1998) and 99 years for biotic projects (the Dutch FACE foundation);
• project-category formulae, e.g. 10 years for replacing an old power plant with one based
on renewable energies, or the commercial life (depreciation period) of a new power plant
(Michelowa 1998); or
• the technical or economic lifetime of the project.
Because the effect of the timeline on total credits from a project is so important, standardising
a methodology to assess a project’s emissions timeline offers perhaps the easiest way of
limiting the negative effects of free riders and gaming. A standardised timeline methodology
would therefore offer a simple way of working to ensure environmental effectiveness when
calculating emission baselines and emission credits from a possible project. Moreover, all
emission baselines will need an emissions timeline, so a standardised methodology could be
used to affect all projects. Nevertheless, applying one standard methodology to all projects is
unlikely to be appropriate. At a minimum, the methodology would need to vary between
projects that reduce emissions and those that increase sinks.
3.3 Data sources

Different data sources could be used when drawing up an emissions baseline. For all
approaches, data could either be based on past performance (historical trends), a snapshot of
current performance (e.g. by comparing the project to current investment trends in a particular
sector or to a control group), or a simulation/projection of future performance. The use of
these different data types will influence the perceived credibility of the baseline, as well as its
data, monitoring and reporting needs, transparency, and the levels of free riders and gaming
associated with a particular baseline.
There are advantages and disadvantages associated with using the different types and sources
of data. For example, projections may be viewed as a more plausible reflection of what
would happen under a future business-as-usual scenario than baselines based on historical
data. On the other hand, the inherent speculative nature of some of the information needed to
establish baselines based on projections means that these baselines may not be not be
transparent unless extensively documented. They may also be more open to gaming and basic
error than baselines constructed using other data sources.
The availability (or not) of the data sets needed may in turn influence the approach used to
calculate an emissions baseline24, as the baseline approaches assessed in section 2 could be
based on any of these data types.
24
For example, if data are available for a comparable project a comparison-based project-
specific baseline could be used. If sectoral data are not available, it would be difficult to use
a multi-project baseline at the sectoral level.
22
4. Quantitative comparison of baseline approaches

This section examines whether and how different baseline approaches could affect the level of
emissions baselines for JI and CDM projects. Sections 4.1 and 4.2 are based on an analysis of
available literature. This literature is mainly on electricity and heating plants in EIT
countries. Section 4.1 examines the effect of different baseline approaches on projected credit
levels from AIJ or AIJ-type projects. Section 4.2 examines the effect of different baseline
assumptions on crediting levels.
Section 4.3 simulates case studies that examine the impact of different assumptions on the
amount and potential value of emission credits under a multi-project baseline in the electric
power generation sector in Brazil and India. By examining the implication of varying baseline
assumptions on the cost of CDM credits, it is possible to understand better the significance of
these decisions on the potential volume of CDM project activity.
Variation among countries’ industry structure, fuel mix, land-use, investment patterns,
policies, evolution of supply and demand all affect the assumptions underlying any emission
baseline. This means that the relative levels of emission baselines established by a common
approach may be different depending on the country and region. The resulting emission
credits would also vary between regions and sectors. This supports the need to establish
baselines at a level where circumstances are similar, which may be at the “sub-country”,
country or regional level, depending on the country/sector examined. Thus, while some
baseline approaches are found to result in more or less generous crediting levels in certain
cases than other baseline approaches, this does not necessarily hold for all project types in all
countries. Different baseline approaches may be appropriate in different circumstances.
4.1 Survey of experience with different baseline approaches

Some analysis has been done on how the level of emissions baselines would vary under
different baseline approaches. This section summarises relevant results.
4.1.1 Conversion of heat plants in the Czech Republic

The following assessment is based on an analysis of Swiss-Czech co-operation in the
conversion of heating plants in the Czech republic (Ernst Basler & Partners 199925). The
analysis assesses what potential credits could be for two different project types and four
different emission baselines. The projects examined were fuel switching (mainly from coal or
coke to natural gas) in 70 existing heating plants, and cogeneration. The results for the four
different baselines examined are shown in Table 4.
The relative credit levels from the different baseline approaches are due to the different
underlying assumptions. Technology-standards (a disaggregate form of multi-project
baselines) gave relatively high levels of offsets for both project types because the reference
technologies used were coal26. However, other assumptions could have also been valid that
25
The projects analysed were not designed as AIJ projects; official acceptance, endorsement or
approval of them as AIJ projects was not sought; and the Swiss government is seeking no
emission credits for these projects as potential JI projects. Nevertheless, these projects fulfill
the AIJ criteria specified internationally (UNFCCC 1995) as well as the development criteria
specified by the Czech government, and illustrate issues relating to emission baselines as
well as an officially endorsed AIJ project.
26
Coal was chosen as the reference technology as the majority of Czech heating plants are
coal-fired.
23
would result in a lower baseline (and crediting levels). The most extreme example in this case
would be if the reference technology were the marginal unit. Since the Czech Republic has
recently approved construction of a nuclear power station, and since nuclear power emits no
CO2, the marginal unit would in this case give a baseline of zero emissions.
Table 4: Variation in calculated carbon offsets under different baselines:

simulated Swiss/Czech projects
Project Types Offsets under different baselines (t CO2)* - see text for baseline descriptions
Project-specific Technology- Sectoral Top-down##
based standards# standards#
Fuel switch 175000 430000 180000 195000
Cogeneration 41000 43000 21000 22000
Source: adapted from Ernst Basler & Partners, 1999
* The figures represent calculated offsets for the entire project life.
# Both these baselines would be included under the “multi-project” category described in this paper.
## This baseline is based both on current sectoral performance and future emission targets.
The size of the difference in offsets under different the baselines in this Swiss-Czech analysis
is due to the difference in per kWh emissions between gas and coal-fired technologies (which
is in turn a factor of the differing carbon contents of gas and coal, and the relative efficiencies
of gas and coal technologies). The highest number of offsets will occur if comparing against
100% coal, and the least if comparing against 100% gas. Any baseline that mixes the two,
e.g. by comparing to the relative importance of gas and coal, will give offsets somewhere in
between the two extremes.
The top-down and sectoral baselines gave lower potential credit levels because they assumed
that the starting point was a mixture of coal and other fuels, and that per unit emissions would
gradually decrease between the start and end of the project. The project-specific baseline
gave a relatively low number for the fuel switch projects because the baseline assumed that all
units would become gas-fired within 15 years. However, the project-specific baseline gave a
relatively high level of potential credits for the cogeneration project because it assumed that
the electricity displaced by the CHP plant was generated by coal. The differences in the credit
level are thus a result of the underlying reference case more than the baseline approach per se.
4.1.2 Boiler conversions in Estonia

Some recent (June 1999) AIJ project reports for Swedish-funded projects in Estonia
calculated emission reductions using both a project-specific and a multi-project approach
(http://www.unfccc.de/program/aij). The project reports examined in this section involved
converting a heating boiler from fossil fuels to biomass. The multi-project approach
examined the whole country by sector and compared competing energy production options
(taking into account government policy such as targets for increased renewable energy
supply, and the emission commitments agreed to at Kyoto). The project-specific approach
examined the likely scenario in production and emissions from a particular heating plant with
and without the AIJ project. Earlier reports for the same projects had also drawn up baselines
on a project-specific basis, although they were slightly different from those presented in the
June 1999 baselines.
24
The results of the four separate baseline studies for one project (e.g. Viljandi boiler
conversion) are outlined in Table 5. The projected emission benefits of the same project with
the four different baselines vary by 23% over the lifetime of the project when comparing the
benefits calculated with one timeline27. (One of the suggested baselines had a shorter timeline,
which resulted in significantly lower expected emissions reductions from the project when
compared to longer timelines).
4.1.3 Other projects

One Dutch AIJ project report (installation of a micro-hydro plant in Bhutan) states that it will
use two different forms of project-specific approaches to calculate the project’s baseline
(using both a project-specific approach as well as a baseline based on the evolving
performance of a comparable situation). However, because the comparison-based approach
was dynamic in its nature, it was not presented in the report submitted to the UNFCCC, and
quantitative comparison between the two is not currently possible.
27
Because of the different relative shapes of the different baselines, the variation in projected
annual emission benefits varied by less than this amount at the beginning of the project, and
by more at the end of the project’s emissions timeline.
25
Table 5: Comparing four estimates of the emission benefits from Viljandi boiler conversion AIJ project
Report Baseline Timeline Annual emissions (t CO2) Cumulative

(y) y1 y2 y3 y4 … Final emission
year reductions*
1997 Project-specific: status quo 10 9800 9800 9800 9800 9800 9800 98000
1998 Project-specific: status quo energy 15
production but mazout replaced by
natural gas after 2 years
- projected reductions 9800 9800 9800 9800 … 9800 147000
- actual reductions 3245 8500 8100 n/a n/a (137445)
1999p Project-specific: status quo 15
- projected reductions 4146 10280 10280 10280 … 10280 158345
- actual reductions 3379 8899 10598 10512 n/a (159531)
1999t Multi-project, including consideration of 15
government policy, such as increased use
of renewables: 122516
- projected reductions 3629 8670 8583 8496 … 7317
- actual reductions 3022 7790 8776 8776 n/a
* Emission reductions = baseline emissions – project emissions
Source: http://www.unfccc.de/program/aij/
26
4.2 Survey of experience with different baseline assumptions

Other studies have also compared the effect of different assumptions on baseline level (and
therefore on the level of emission benefits) from selected AIJ projects.
For example, the Centre for Clean Air Policy (CCAP) has undertaken a qualitative estimation
of different baselines possible for the AIJ project situated in Decin, Czech Republic, Figure 2,
(CCAP 1999). This AIJ project converted a coal-fired district heating plant to one fired on
natural gas. The project also included building a new co-generation unit that would supply
electricity locally and feed surplus into the grid.
The CCAP exercise demonstrates that, as also illustrated above, different valid assumptions
for baselines can significantly alter the level of credits generated by a project. The exercise
analysed two forms of “benchmark” (multi-project) approaches. The “historical benchmark”
baseline was based on emissions from average heat and electricity production facilities, and
continuing energy efficiency improvements. The “forward looking benchmark” assumed
further improvements in energy efficiency and a greater penetration of natural gas in heating
and electricity sectors as well as some new nuclear generation. The “forward looking
benchmark” baseline also sought to take the country’s Kyoto commitment into account, as
well as sectoral development plans, when assuming gradual improvements in emissions per
unit from heating and electricity production. (The actual emissions baseline used for the
Decin project is in fact none of those below, although it follows a similar pattern to the
historical benchmark line).
Figure 2: Comparison of Decin Project Baselines
Comparison of Decin Project Baselines
40,000
35,000
30,000
25,000 Hist Bnchmk

tCO2/yr
20,000 Top-Down
Frwd Bnchmk
15,000
Proj Emssns
10,000
5,000
0
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
12
19
19
19
19
20
20
20
20
20
20
20
20
20
20
20
20
20
Year
Source: CCAP 1999
Some AIJ projects presented multiple baselines in their report to the UNFCCC (e.g. a
Dutch/Russian horticultural project in Tyumen). The Dutch/Russian analysis in the Tyumen
project report includes five different baselines, all calculated using a project-specific baseline
approach but with different assumptions for fuel use, technologies and project boundaries.
The emission benefits of the project (presented on a per unit basis) vary by more than a factor
of two between the different baselines, and the lowest baseline was used to calculate the
emission benefits of the project.
27
Other AIJ projects have had more than one baseline drawn up for them, with one chosen as
the most likely scenario by the project participants and then submitted to the UNFCCC.
Some of these alternative baselines are public (e.g. for the Norway/Slovakia project - see
Yager et. al. 1998), and others are not (e.g. for the US/Russia Rusafor project - see
Michaelowa 1998). The three project-specific baseline scenarios with different underlying
assumptions drawn up for the Norway/Slovakia project (converting a coke-fired heating
boiler at Jochy to biomass) result in emission benefits that vary by a factor of nine from the
same project28. Even discounting the baseline scenario judged unrealistic (a 100% switch to
biomass in the absence of the AIJ project), the other two baselines vary by almost a factor of
two (estimating emission benefits from the 2 * 300 kW boiler project as 20 or 35.6 kt CO2).
4.3 Case study simulations with multi-project baselines

This section examines examples of hypothetical multi-project baselines for new electricity
generation29 projects in two potential CDM host countries: Brazil and India. The simulations
are developed to illustrate the implications of some alternative multi-project baseline
assumptions on the potential amount of emission credits that could be generated by such
projects. The section also considers the differences in investment and operation cost estimates
of different power generation alternatives, and how the relative economic feasibility of these
alternatives could be affected by the potential value and amount of emission credits under
different multi-project baseline assumptions30.
Assumptions and Caveats
Various sources were used to obtain emission and cost data and estimates for these case studies although
most data were obtained from publications of the IEA and NEA (NEA/IEA 1998 and IEA 1997b).
Consequently, the data presented may not always be fully comparable.
Cost estimates, types of energy sources, technology and resulting emissions vary according to each
country’s specific national circumstances. In particular, cost estimates are highly dependent on assumptions
for discount rates and future prices of energy. In this section, the discount rate used is 10 percent. It is also
important to note that costs may be underestimated in some cases, as they were estimated based on a stable
political and socio-economic context, which may not reflect the real investment conditions. Cost estimates
for renewable energy sources are difficult to compare as some, such as wind and solar, are intermittent and
are thus not comparable to sources used for baseload generation; while others, such as hydro and biomass,
are very site-specific making cost comparisons between sites quite imperfect.
For the purpose of these case studies, CERs are calculated based on projects generating emissions lower
than a set baseline standard, i.e. using a range of multi-project approaches, as described in section 2, at the
sectoral level. No judgements are made to determine whether such CERs would meet an environmental
“additionality” test or whether they would meet the host country’s sustainable development priorities,
although both are of course critical aspects of the CDM.
28
However, the lowest baseline - corresponding to an autonomous fuel switch to biomass - was
not considered as a viable alternative because of investment constraints.
29
Electricity generation was chosen for this case study for several reasons. It is a large and
growing source of GHG emissions in many countries; it is a sector with a homogenous
output; and relevant data (e.g. on emission characteristics of current technologies and BAT,
and demand projections) were relatively easy to obtain.
30
The assessment of economic feasibility is based on the amount of emission reductions (under
a particular multi-project baseline assumption) divided by the difference in investment and
operation costs between two potential electricity generation options.
28
The different multi-project baseline assumptions simulated in this section include: weighted
average emissions from all sources, average emissions from each fuel, 25% improvement on
average emission performance, planned new technology emission performance and BAT
emission performance (see tables in Annex B). Alternative sectoral multi-project baselines
are possible, e.g. using the emission performance of marginal units constructed, and different
sectoral aggregations, which may be viewed as a closer representation of business-as-usual.
However, these are not examined here because of current lack of data.
These case studies are intended to provide an illustration of the potential practical
implications of different simple multi-project baseline assumptions in different countries.
They should not be taken as a model for the development of baselines for CDM projects in
India and Brazil.
4.3.1 Electricity Generation in Brazil

According to the projected energy market scenarios of the Brazilian National Electric Energy
Plan 1993/2015 published in 1994 (NEA/IEA 1998), Brazil’s consumption of electricity is
expected to grow to 534.1 TWh by 2015 (i.e. an increase of 95% from 1993) under the lowest
scenario. Under the highest scenario, it is projected to grow to 743.3 TWh (i.e. a 171%
increase from 1993). It is important to note that scenarios31 on Brazil’s electricity generation
indicate a decrease in its reliance on hydropower, which represented about 92% of its total
electricity generation in 1996 (see tables in Annex B), and an increase in thermoelectric
sources (including oil and coal, but mostly natural gas). Although there is still significant
hydro potential remaining in Brazil, most economic sites are already exploited. Furthermore,
about 50% of the remaining potential is located in the Amazon area, which may not be
considered as appropriate for reservoir development. This area is also a significant distance
from the most developed region of the country (i.e. Rio de Janeiro and São Paulo), where the
main source of growth in electricity demand is expected to occur.
The predominance of hydro in Brazil’s electricity generation mix has important implications
for the construction of an emissions baseline. For example, if a single multi-project emission
baseline for Brazil’s power sector were set at the level of the weighted average of emissions
from all sources (i.e. 43.9 t CO2/GWh), this emissions baseline would preclude all fossil fuel
sources generating CERs from CDM projects. In addition, under a weighted average
emissions baseline, the amount of CERs generated from even non-GHG emitting projects
would be quite low compared to similar projects in other countries (see below) that have a
more fossil-fuel intensive electricity sector. Under a multi-project baseline set at the weighted
average emissions level, new hydro capacity could qualify32 as a CDM project and could
therefore generate some CERs. However, hydro is not expected to be the cheapest option for
future new electricity generating systems as most economical sites are being exploited and
large up-front investment capital is required. Moreover, further development of hydropower
may be subject to other constraints such as those related to transmission infrastructure or a
reluctance to developments in particular areas.
Thus, under a weighted average emissions baseline in Brazil, even the most efficient new gas-
fired plants would not generate any CERs. However, a multi-project baseline set at this level
would mean that a new wind project could generate around 44 CERs/GWh (assuming that 1
CER = 1 t CO2).
31
Information from the Electrobras website:
www.embratel.net.br/infoserve/electrobr/evindice.html
32
Such projects would also need to meet Brazil’s sustainable development objectives.
29
Given that projections indicate a significant growth in electricity generation, it is interesting to

examine what these potential emission baselines and credits could imply for future
investments in electricity generation projects in Brazil. Different electricity generation
alternatives have different costs33 for investors, as shown in table B2. However, the potential
value of the CERs under different multi-project baseline assumptions resulting from
electricity generation projects may provide different investment incentives than would
otherwise be the case when CDM is not taken into account. For example, wind electricity is
currently more costly than gas-fired electricity in Brazil, but could become an economic
alternative to a gas plant if CERs are worth more than 340 USD/t CO234, and an economic
alternative to a coal plant if CERs are worth more than 130 USD/t CO2.
Given that scenarios for Brazil’s future electricity generation indicate a significant increase in
the use of thermoelectric sources (and not hydro), it may be deemed appropriate to design
baselines based on these technologies. Multi-project baselines set at 25% better (i.e. 25% less
CO2-intensive) than the current average for thermoelectric sources, or at the emissions level
from best available technology of various sources, could provide incentives to invest in more
GHG-friendly thermoelectric sources and technology than would otherwise be the case. The
baselines and resulting crediting levels for different multi-project baseline assumptions are
illustrated in Figure 3. The length of the arrow is proportional to the number of credits
obtained per GWh.
Figure 3 shows that, for example, a wind (or biomass, or other renewable-based) CDM
project could generate 44 CERs per GWh (represented by the length of the arrow at the left of
figure 3) if using a baseline set at the level of the weighted average emissions of all plants.
This would increase to 382 CERs/GWh if the baseline were set at the level of BAT gas plants,
755 CERs/GWh if it were set at the level of weighted average thermal generation and 882
CERs/GWh if it were set at the level of average fossil fuel emissions. (See table B1 for other
examples).
Figure 3: Effect of different baseline assumptions on baseline levels, Brazil
Of course, the actual

Possible baseline assumptions:
Weighted average (fossil) investments in such projects
800 will very much depend on
Average (current gas plants) whether the value of CERs
cover any increased expense
600 (compared to business-as-
Emissions (t CO2/GWh)
usual investment plans) that

Planned new gas would need to be incurred to
400
Average BAT (gas) generate electricity emitting
GHG emissions below the
200
baseline level. For example, if
a baseline is set at 25% better
Weighted average (all sources)
than average gas emission
New wind plant levels (i.e. 556 t CO2/GWh),
Wind New gas the investment in BAT natural
Possible credits under different baseline assumptions
33
Investment and operational costs are provided at a national level.
34
The difference between the cost of electricity generated by wind in Brazil (4.47 US
cents/kWh) and gas-fired electricity (3.38 US cents/kWh) is 1.49 US cents/kWh or 14,900
USD/GWh. Based on this cost differential, investment in wind is economic if a CER is
worth at least 340 USD per ton of CO2 (i.e. 14,900 USD/GWh divided by 43.9tCO2/GWh).
30
gas plants (i.e. more efficient CCGT) may become an economic alternative to the gas plants
currently planned if CERs are worth more than 79 USD/t CO2. Under this same emission
baseline (i.e. 556 t CO2/GWh), wind generation would become an economic alternative if
CERs were worth more than 27 USD/ t CO2.
It is important to note that while these examples examine the potential implications of
alternative unique multi-project baseline assumptions for new projects in Brazil’ s electricity
sector, another possibility may be to consider developing more than one multi-project
baseline for the electricity sector. For example, one for thermoelectric sources and a separate
one for non-GHG emitting sources. However, this possibility is not examined further in this
paper.
4.3.2 Electricity Generation in India

India is the world’s largest coal producer. Coal is used in the generation of more than 70% of
Indian electricity, and is expected to remain the main fuel source for power generation to at
least the year 2020. Coal is also the most economical electricity generation option, as shown
in table B4 (see Annex B), and benefits from the most developed infrastructure, including
mine-mouth electricity stations. However, the current average efficiency of coal plants is
quite low relative to other countries, particularly OECD countries: average generation
efficiency was 27% in 1996 (IEA 1998), although this masks significant variations in average
efficiency at the state level. The low efficiencies are due to various factors, such as the age of
power plants, poor maintenance and lack of appropriate quality of coal.
India’s power sector has been experiencing significant peak and supply shortages. Shortfalls
in the investment and construction of new power plants, poor-quality transmission lines and
theft are the main reasons for the supply shortages. These are exacerbated by the lack of an
inter-connected grid across the nation, as states with surplus power do not transfer that surplus
to states facing power shortages. Plans to bridge the gap between increasing electricity
demand and supply include, in addition to greater use of coal, increasing India’s nuclear
power generation capacity, as well as (on a smaller scale) increasing the use of renewable
energy. India’s Ninth five-year plan (1997-2002) includes a target for non-hydro renewable
capacity of 3000 MW.
Figure 4 shows various multi-project baseline assumptions for India’s electricity sector, along
with their associated crediting levels.. For example, based on 1996 data provided in Table
B3, if a sectoral multi-project baseline was set at the weighted average of current generation
in India, it would equal 945 t CO2/GWh. In effect, this would mean that electricity generated
by all sources other than coal (using current Indian technology) would be below this emission
baseline, including coal plants using the best available technology (i.e. clean coal). Under
such a scenario, new wind (or other renewable) plants would generate 945 t CO2/GWh. This
would decrease to 786 t CO2/GWh if the sectoral multi-project baseline were based on BAT
coal emissions, and decrease further to 382 t CO2/GWh if the baseline was based on BAT gas
plants. Planned new coal plants would not be eligible for any credits at all if the multi-project
baseline assumption chosen was the weighted average emission for all generation sources,
BAT coal, or of course, BAT gas.
31
Figure 4: Effect of different assumptions on baseline levels, India
However, the lower

emission level of coal BAT
Possible baseline assumptions:
plants is associated not only
Average (current coal plants) with the better technology,
1200
Weighted average (fossil) but also with higher levels
1000 Planned new coal of maintenance than
Weighted average (all sources) currently observed in India.
Emissions (t CO2/GWh)
800 This results in the

Average BAT (coal)
investment and operational
600 cost of these plants being
higher than that of actual
400 Planned new gas planned coal plants in India
by about 0.92 US
200 cents/kWh (Table B4).
Although various factors are
New wind plant taken into account in
Wind New gas New coal investment decisions, the
Possible credits under different baseline assumptions
incremental cost associated
with investing in BAT coal plants means that the investment attractiveness, from a strict
economic perspective, of BAT coal plants is less than that of currently planned coal plants. To
create the incentive to switch from planned plants to the best available technology for coal
plants, under a baseline set at 945 t CO2/GWh, the revenues from 159 CERs would have to at
least match the difference between the cost difference between the two types of plants. This
means that each CER would need to be worth a minimum of 58 USD/t CO2 to make BAT
coal plants an economic option, compared to planned coal plants, under this multi-project
baseline assumption.
However, setting a sectoral multi-project baseline at the level of the weighted average for all
sources in India would mean that all planned gas plants35 would be able to generate CERs if
they were initiated as a CDM project. According to the cost estimates in Table B4, installing
a new BAT gas-fired plant 36 would cost about 5,400 USD more per GWh than the cost of
installing a “planned” coal plant in India. In terms of emission reductions, the BAT-type gas
plant would generate emissions reductions totalling about 563 CERs. In this example, the
additional investment cost of a BAT natural gas plant would be economical, compared to the
cost of a planned coal plant, as long as a CER was worth more than 9.6 USD per tonne of
CO2. A wind electricity plant would be economic compared to a new coal plant if CERs were
worth 23 USD (assuming that a unique sectoral multi-project baseline was based on the
weighted average emissions of all sources).
As mentioned previously, only scenarios involving a unique multi-project baseline for India’s
electricity sector are examined here. However, scenarios with more than one sectoral multi-
project baseline could also be envisioned.
35
Emissions associated with natural gas electricity generation may be underestimated as only
CO2 emissions are taken into account in this study.
36
BAT figures in this study are derived from data on planned plants in OECD countries
(NEA/IEA, 1998). In the case of natural gas, there is not a large difference between planned
CCGT plants in India (at 50% efficiency) and OECD planned CCGT plants (at 52.9
efficiency).
32
4.3.3 Simulation Results

This case study only takes a very preliminary look at the potential implications of using
different sectoral multiple-project baselines in the electricity sector. More work in this area
may be desirable in order to understand better the effects associated with different multi-
project baseline assumptions.
This analysis shows that a key issue in determining the emissions credits associated with a
project is the stringency of the baseline level. Multiple-project baselines can be designed to
be more or less stringent within one country. However, if different countries choose to use
the same multi-project baseline assumption, it could result in different baseline levels and
different implications for potential CDM projects.
For example, a multi-project baseline in the electricity sector set at weighted average
emissions would equal 945 t CO2/GWh in India, and 44 t CO2/GWh in Brazil. This would
mean that if a unique multi-project baseline based on weighted average emissions was chosen
for the power sector in both Brazil and India, all clean coal-fired projects would be below the
baseline in India, whereas no thermoelectric plants would be in Brazil. The implications for
CERs, in such a scenario, would consequently be different in both countries: new gas plants
could generate 542 CERs/GWh if they were installed in India, but none in Brazil. The current
diversity in conditions in the electricity sector among different countries (and sometimes,
between different regions within a country) is expected to continue. It may therefore be
useful to consider the applicability of different multi-project assumptions for different
countries, as well as perhaps more than one sectoral multi-project baseline within the same
country.
In all cases, a less stringent multi-project baseline level results in a greater amount of credits
being generated by each project, as well as potentially leading to a greater total number of
projects by increasing the range of “eligible” projects. The tradeoff for lax baselines is that
eligible projects may not be, on their own, strictly environmentally “additional.”.
With a less stringent multi-project baseline the value of each emissions credit can be
relatively low while still making the additional CDM investment feasible. With a more
stringent baseline (where the level of the emission baseline is lower) fewer credits will be
generated from a particular project, so the value of each emission credit will need to be higher
to recoup the same incremental investment costs.
33
5. Qualitative comparison of different baseline approaches

The different baseline approaches and their application outlined above may differ in:
• data, monitoring and reporting requirements;

• baseline development or transaction cost;
• transparency; and
• environmental effectiveness, including a project’s environmental additionality (related to
the potential for gaming, free riders and leakage).
Using these criteria, this section compares and assesses each of the different baseline
approaches.
5.1 Data, monitoring and reporting requirements

The data needed to establish a baseline for a JI/CDM project will vary by project type, but
will also vary depending on the approach used to establish the baseline. Higher data
requirements are also likely to increase the cost of setting up an emissions baseline. This
section assesses the data requirements and monitoring and reporting needed to establish an
emission baseline under different baseline approaches37.
Project-specific baselines
Project-specific baselines are relatively data intensive. Experience from the AIJ pilot phase
shows that establishing these project-specific baselines, in the absence of any centralised
guidance on what needs to be monitored and how, is often costly and time-consuming.
Project-specific baselines may require some monitoring of current activities before the
JI/CDM project or activity starts in order to be set up. This type of information would vary
by project type, and would also depend on whether the proposed project is a refurbishment
project or at a greenfield site. For example, some of these data, such as the age of the current
installation (in the case of a refurbishment project), are likely to be readily available.
However, other data, such as a plant’s thermal efficiency, may need measuring specifically in
order to set up the emissions baseline. Examples of information that may need to be gathered
and reported for selected project types are shown in Annex A. Project-specific baselines can
in theory be applied to all JI and CDM project types.
Multi-project baselines
Using a multi-project baseline does not require monitoring if it has already been developed.
As with all approaches, of course, a certain amount of project-specific data will need to be
reported (such as project size and output) during the project operation.
Multi-project baselines can be based on the performance of current installations or the

projected performance of future installations. To establish such a baseline requires
information on:
• current technology characteristics of sector (i.e. the current distribution of technologies

used in that sector); and
• the emission characteristics of each technology type.
37
This section does not explore the data, monitoring and reporting requirements needed to
monitor the actual performance of a project.
34
Multi-project baselines at the technology level could also be used where there is not enough
data on a sectoral level to construct a sectoral baseline. For example, a technology-level
baseline could be set at an emissions level 10% higher than the best available technology of
that type.
The development of multi-project baselines would require some resources and agreement at a
level other than that of the participants (e.g. governments, international organisations). The
development of these baselines may require balancing competing interests, e.g. between
investors and hosts for JI projects.
Constructing multi-project baselines may not be appropriate or possible for all sectors in all
countries. For example, not all countries may have complete data coverage for particular
sectors. However, there are a number of straightforward possibilities, as demonstrated in
section 4.3.
Hybrid baselines
Hybrid baselines are between project-specific and multi-project baselines on the continuum of
possible baseline approaches (Figure 1). They are therefore likely to need more data,
monitoring and reporting than multi-project baselines, and less data, monitoring and reporting
than project-specific baselines. However, how much more or less would depend on how
many of the baseline components were standardised (which is in turn dependent on project
type).
5.2 Baseline development cost

The cost of developing an emissions baseline is related not only to the data requirements,
outlined above, but to the number of projects it, or some of its components, could be used for.
In the case of multi-project baselines, or hybrids of this approach, the cost in time to achieve
international consensus on the basic approach and rules for its application must also be
considered. There is a trade-off here with the ease of implementation once agreement is
reached. The cost of using a baseline will therefore not be the same for each baseline
approach. If the international process can be assumed to move quickly to reach consensus on
basic starting points, using a standardised methodology, or an off-the-shelf multi-project
baseline will be cheaper than developing a project-specific baseline from scratch.
The cost of an emissions baseline is also related to its complexity. For example, a simulation
baseline that needs a system model will be expensive and time-consuming to set up and run.
The cost of developing such a model may mean that this approach is only used in future for
either very large-scale projects, such as forestry, or where many similar projects are likely to
occur, such as in fast-growing sectors such as electricity or cement production.
Limiting the time needed to establish and approve the baseline will also help to limit the cost
of the baseline. This can be done if agreement is reached on:
• what data to monitor and report;
• how to monitor/report this data;
• calculation methods; and
• value(s) of the baseline (e.g. for multi-project baselines).
35
Many potential emission mitigation projects are small scale, e.g. increased household use of
renewable energy and energy efficiency technologies. In order to maintain small-scale
projects as an attractive JI and/or CDM opportunity, the up-front costs for these projects need
to be low relative to the cost of the project. Having a simple emissions baseline or crediting
formula for small-scale projects would be one way of keeping the costs related to construction
of the baseline low. Of course, for both large and small-scale projects, the proportion of total
project costs used to establish the baseline will also need to be kept low in order not to create
barriers to investment under JI and CDM.
The level at which baselines are aggregated (e.g. sub-sectoral or sectoral level) will influence
the number of projects for which the baseline could be used, and ultimately the cost per
project. The degree of standardisation between different baseline approaches also has cost
implications. More standardised methods and reporting formats, irrespective of the level of
aggregation, would mean that each project developer would not have to spend time and
money developing methodologies or reporting formats but would instead apply those already
developed. However, reaching agreement on standardised calculation methods, parameters,
and/or reporting requirements could have significant up-front time and cost implications.
However, setting up an emissions baseline does not represent the only costs involved in
establishing a JI or CDM project. Other costs include the costs of identifying the potential
project, administrative costs (for both the host and investor), ongoing costs related to
monitoring, reporting and verification, and, for the CDM, contributions to funding adaptation.
5.3 Transparency and ease of third party verification

A transparent baseline is one where the construction is clearly explained, where the
parameters and methods used in the baseline are referenced and traceable, and where the
baseline can be reconstructed by a third party. Use of transparent baselines for JI and CDM
projects would help to “demystify” emission baselines; would help potential project
developers better assess CDM or JI projects, and associated emission credits; and would
facilitate any verification of a project’s emission baseline. They would therefore increase
confidence in the effectiveness of the project-based mechanisms.
The transparency of a baseline will vary with the baseline approach used, and will also
depend on how baseline assumptions were arrived at (e.g. model simulations, expert
judgement) and how these assumptions and supporting information were reported to the
UNFCCC.
Many baselines used for AIJ projects were not transparent. This is partly because which data
had to be reported, how, and in what detail was not established38. Thus, many project reports
do not reference or explain key assumptions, and do also not report all relevant underlying
data and information. Transparency could be improved for future reports of project-based
activities if guidance was available on the assumptions to be reported.
Another reason behind the lack of transparency of AIJ baselines is due to the project-specific
nature of many of those baselines. Since project-specific baselines are made up of many
components, listing each of these components and explaining why they (rather than
alternatives) were chosen would be a time-consuming exercise. Given the uncertainties
surrounding the underlying components of emission baselines, and the importance of expert
judgement in drawing up project-specific baselines, a degree of subjectivity is inevitable.
38
A draft revised version of the Uniform Reporting Format has been prepared and is available
in document FCCC/SB/1999/5/Add.1
36
While emission baselines based on projections may be viewed as a closer representation of

what would happen otherwise, basing emission baselines on projected data (simulations) is
also likely to result in baselines that are not as transparent because of the inherent uncertainty
associated with projections. Lack of transparency is due to the importance of expert
judgement in projections, and also because any simulation would need to be explained fully in
order to enable a reconstruction.
Baseline transparency is likely to increase if there are fewer components to an emissions

baseline (e.g. in the case of multi-project baselines), or if the components and/or methodology
used to set up the baseline are standardised to some extent. In addition, standardisation could
help to group similar projects together39 and thus facilitate centralised, and presumably more
streamlined, verification (e.g. by governments, by the UNFCCC, the CDM’s Executive Board
or another body) of emissions credits by allowing easy identification of “outliers”.
5.4 Environmental effectiveness

There are at least two components to environmental effectiveness. The first is at the project
level, i.e. the environmental additionality of the project. The second is the performance of JI
or the CDM as a whole with respect to whether they contribute to achieving the objectives of
the Protocol by limiting global emissions in a manner that is least cost.
5.4.1 Project-level environmental additionality

The difference between a project’s actual emissions and emissions baseline is a measure of
the environmental additionality of that project. Of course, the emission baseline needs to
reflect a credible business-as-usual scenario in order to assess emissions “additionality”. This
additionality is also affected by the extent of free-riders, gaming and leakage (see glossary
and below) associated with a project and its baseline40. This section assesses how and if the
approach used to establish an emissions baseline can affect these different factors.
Project-specific baselines enable an individualised assessment of environmental additionality

based on what could have happened in the absence of that particular project. In contrast,
hybrid and multi-project (both aggregate and disaggregate) baselines are not developed
expressly for one particular project and therefore represent an average approximation of
“what would have happened otherwise”.
Gaming can have a substantial effect on the level of baselines for individual projects, and
therefore the level of credits associated with a single project. For project-specific baselines,
gaming could take place for each of the many components that make up the emissions
baseline. This is because the assumptions used to develop a project-specific baseline might
be expected to vary significantly from project to project even for those that appear to be
similar (sometimes referred to as “micro” gaming). Gaming would have the effect of
weakening the stringency of the baseline, and therefore increasing the level of credits
associated with that project. The relatively high potential for gaming was a significant
drawback noted to the project-specific baselines used in AIJ pilot projects.
Gaming might also extend across whole project categories (sometimes referred to as “macro”
gaming), e.g. if gaming was prevalent in establishing agreement on the levels of multi-project
39
This is not necessarily the case in the AIJ pilot phase, where the different categories are so
wide that there not all projects within each category are comparable.
40
As mentioned previously, the implications for global GHG emissions may be different for
CDM and JI.
37
baselines. However, the potential for gaming can be expected to be lower in the case of
standard assumptions being developed by bodies not having specific economic interests in the
projects.
As the examples in sections 4.1 and 4.2, and in other sensitivity analyses (e.g. Ellis 1999),
demonstrate, baseline levels are more sensitive to some assumptions than others. For
example, for energy projects emission benefits are highly dependent on e.g. the emissions
timeline and the baseline fuel/technology. Gaming is possible when assessing and deciding
both of these assumptions.
In the case of a coal-fired power plant, emissions per kWh are approximately double those of
gas-fired electricity (due to gas’ lower carbon content and the higher combustion efficiency of
gas-fired stations). Therefore, comparing a proposed JI or CDM project against a coal-fired
baseline could double the ERUs or CERs compared to a gas-fired baseline. It may be
appropriately argued that a coal-fired baseline is a valid assumption for a new plant if coal-
fired generation makes up a substantial proportion of baseload generation in the proposed host
country. However, coal may not be the type of plant that is presently being used for new
baseload facilities in that country. Nevertheless, assuming that a plant currently fired by coal
would continue to be so could be considered a valid assumption for a baseline in a proposed
refurbishment-type JI or CDM project, even in a country where policies aim for a
decarbonisation of electricity supply.
Gaming could significantly alter the credits associated with a particular project, especially
with respect to fuel input assumptions. This in turn could limit or cancel the environmental
effectiveness of the project. Clear rules on how to choose specific baseline assumptions in
different situations could limit the amount of possible gaming.
Free riders would obtain emission credits for whole projects that would have gone ahead in
the absence of the project-based mechanisms. Without clear rules on baseline selection and
thus project eligibility, free riders could inflate the number of projects obtaining credits. Free
riders could therefore artificially inflate the estimated “environmental benefits” or credits
arising from JI and CDM projects.
“Free-riding” is essentially a matter not of the baseline approach, but of baseline stringency.
For example, if projects are deemed automatically eligible for credits by fulfilling certain
criteria (e.g. by having per unit emissions below a certain level), a lax baseline could
systematically result in free riders41. Project-specific baselines may also give rise to free
riders. This could occur as a result of the potential for gaming in developing project-specific
baselines (e.g. in the case that a project developer falsely stated that the project would not
have gone ahead in the absence of JI/CDM), or because of uncertainty in the emission
baseline (e.g. if a developer estimates a baseline in good faith but is proved wrong at a later
date).
Some have suggested that multi-project baselines could also be set in such a way to allow
crediting only for low-emitting project types or categories (Friedman 1999b). This has been
referred to as “forcing”, and could reduce the risk of free riders. However, if “forcing” led to
low baseline levels, it may inhibit some projects that may nevertheless be environmentally
beneficial (e.g. a gas-fired plant where coal is being widely used).
41
However, assessing that some technologies are a priori additional may result in lower or
higher numbers of “free rider” credits depend on which technologies were assessed as
automatically additional. For example, allowing all non-hydro renewable projects to obtain
credit is still likely to result in only a small number of “free rider” credits if only a limited
number of projects that are small-scale are initiated.
38
Free riders and gaming may pose more of a problem for CDM projects than for JI projects:
CDM hosts potentially gain investment in CDM projects and do not lose anything if a CDM
project is demonstrated to be a free rider. However, the global atmosphere may lose from
free-riding in the CDM. Similarly, free rider projects in a JI host country could make it more
difficult for that country to meet its emission commitment.
Leakage occurs if the actual emission reduction or increase in sinks from a JI or CDM project
results in emissions increasing elsewhere. Leakage could occur because system boundaries
may be drawn in such a way as to ignore emission increases from the proposed JI or CDM
project.
Potential sources of leakage vary according to project type, and which emission sources
and/or effects are taken into account (or not) in a project’s baseline. This may be influenced
by the type of baseline used. Multi-project baselines could result in relatively low levels of
leakage if the baselines account for all emissions in a sector/country. Project-specific
baselines could result in higher or lower levels of leakage, depending on where the boundary
for assessing the emissions baseline is drawn.
5.4.2 Overall environmental effectiveness of JI/CDM

The foregoing discussion points out that different baseline approaches result in different
potentials for gaming, free riders and leakage. However, it is difficult to assess the aggregate
effect of these three factors without knowing the number and mix of projects that would be
eligible and initiated under each of the different approaches.
The overall environmental effectiveness of JI/CDM42 will be a factor of both the

environmental additionality of projects and the total number of projects undertaken under the
mechanisms. Environmental performance will be greatest if the mechanism is set up so that a
high number of clearly additional projects are established. Performance will be lowest if
many projects, none of which are additional, are established. It will be positive, but not
optimal, if a few projects with a high additionality are undertaken, or if many projects with
low additionality are undertaken. The challenge is to balance the number of projects with the
environmental additionality of individual projects in order to maximise overall environmental
effectiveness of the mechanism.
Figure 5 illustrates the possible effect of baseline complexity and baseline stringency (a proxy
for the environmental additionality of a project) on the number of projects undertaken. One
of the reasons suggested for the relative paucity of AIJ projects was that setting baselines was
too complex and costly. Reducing the complexity and cost of setting baselines will therefore
reduce one of the barriers to implementing project-based mechanisms (IEA 1997), and is
therefore likely to increase the numbers of projects initiated. This is represented in Figure 5
by the left-hand side assuming a greater number of projects than those on the right hand side
at the same level.
Figure 5 also illustrates the hypothesis that project numbers are related to baseline stringency.
A stringent baseline is one using assumptions that result in a lower, rather than higher,
baseline level. A stringent baseline will therefore lead to fewer credits per project, but -
despite the uncertainties surrounding baselines - is more likely to result in a project that is
environmentally additional. If investors in project-based mechanisms are rational actors, they
would prefer having a less stringent (or a lax) baseline as this would result in more credits for
42
The implications of different levels of effectiveness may be different for JI and for CDM
because of the differences already noted between these mechanisms (see section 1.2).
39
the same project. This is shown in Figure 5 by the assumption that project numbers will be
larger with a lax baseline than with a stringent one.
The third hypothesis implicit in Figure 5 is that the baseline complexity is not related to the
baseline stringency. Thus, complex baselines could be either stringent or lax, and therefore
generate credits for projects that may be additional or not. Equally, simple baselines could
lead to greater or lesser stringencies – therefore leading to projects where the additionality is
more or less certain. Clearly, some degree of complexity may be warranted. For example, we
would not choose to set a random level for a baseline in the forest sector – but rather would
need to know something about expected business-as-usual lifetimes of existing trees and new
trees planted; developing such a baseline could be a data intensive, and relatively complex
operation.
Figure 5: Possible effect of baseline stringency and complexity on project
stringent
Lower transaction costs Higher transaction costs
Additionality Additionality
Some projects Few projects
Lower transaction costs Higher transaction costs

Baseline stringency
Additionality Additionality
Many projects Some projects
simple complex
Baseline
Lower transaction costs Higher transaction costs complexity
No additionality No additionality
Many projects Some projects
lax
numbers and project’s environmental additionality
If these three hypotheses are correct, the maximum environmental effectiveness of the
project-based mechanisms (as opposed to individual projects) is unlikely to be achieved either
with highly complex baselines, or with overly stringent baselines. As a general rule,
minimising the degree of complexity for emission baselines is desirable as long as we do not
compromise our ability to determine “what would have happened otherwise”. Similarly,
maximising the environmental stringency of the baseline must be traded off against the
desirability of maximising the overall global environmental benefits from the project-based
mechanisms – a greater number of good projects will be more beneficial for the environment
than a smaller number of individually better projects. However, neither the relationship
between project numbers and baseline stringency or between project numbers and baseline
complexity is known. It is therefore not yet possible to determine at which baseline level(s)
the number of projects are likely to be so low that the mechanisms would no longer be an
effective means of reducing global GHG emissions.
Table 6 summarises the different criteria for, and consequences of, different emission baseline
approaches.
40
Table 6: Summary comparison of different baseline approaches
Baseline approach Lead time for Data needs and Transparency Applicability Possible significance of:
rulemaking1 costs for project to many Free riders Gaming Leakage
preparation and different potential*
approval sectors/
project types
Project-specific Low-
Low High Low High Medium High --
Multi-project
Medium-High Low High Medium Low Low --
- technology level
High Low Medium-High MediumHig Medium3 Low- --
h2 Medium4
- aggregate (e.g. sector,
subsector)
Hybrid Medium Medium Low High Low Low --
* The leakage potential is likely to vary substantially between different baseline approaches. However, since this paper does not analyse this in depth, a qualitative
assessment is not presented in this table.
1
e.g. for some aggregate baselines, either host countries, researchers, organisations or others may be involved in setting up baselines.
2
Once a baseline is established for a project, it could be applied to any project in that sector. The difficulty will be to gather enough data to form a baseline for some sectors.
3
Whether or not this baseline type could systematically lead to free riders is highly dependent on the level of the baseline.
4
Gaming could be important at the macro level if not independently verified.
41
6. Conclusions
There are many technical issues that still need to be resolved for both JI and the CDM43.
These include questions related to baselines as well as to additionality, eligibility, monitoring,
reporting and verification. Many of these issues are interlinked; defining one will influence
others. Determining how to set emission baselines for project-based mechanisms may need to
be advanced in parallel with these other rules, in order to take these other rules into account.
Determining emission baselines is critical for the assessment of a project’s environmental

additionality, but it can be a complex process. Moreover, baselines estimate what would
happen in future, and are therefore hypothetical and subject to uncertainty. If “perfection is
the enemy of the good” a balance needs to be drawn between striving for a perfect baseline
and what is practical from an environmental, institutional and investor’s point of view.
There are different approaches that could be used to establish emission baselines for project-
based mechanisms. “Project-specific” approaches evaluate emission reductions generated
from a particular project. “Multi-project” approaches can be applied to many different
projects of a similar type or types. “Hybrid” baselines could also be designed, and would aim
to maximise the benefits from the other approaches. Distinguishing between different
approaches may be difficult in some cases.
Some cross-cutting issues are relevant to all baseline approaches. These include the length of
time emission credits can accrue and whether or not the baseline is fixed at the start of the
project (static) or revised during the project operation (dynamic). Static baselines are
predictable, less of an administrative, monitoring and reporting burden, and cheaper.
Dynamic baselines may better reflect actual trends, but would result in greater uncertainty as
to how many credits a particular project would generate. A number of analyses suggest that
dynamic baselines may be more appropriate for some project types.
There are many different views on how an emissions timeline could be set, but given the
importance of the timeline on total credits from a project, applying standardised
methodologies to assess a project’s emissions timeline could offer a simple way of limiting
the effects of free riders and gaming. Any such methodologies would need to vary between
projects that reduce emissions, where short timelines help to ensure the environmental
additionality of any credits, and those that increase sinks, where longer timelines would have
the same effect. A distinction may also be needed between the timelines for JI and CDM
projects, given the potential emission leakage associated with the latter.
Baseline approaches differ in their costs, transparency, data requirements and monitoring,
reporting and verification needs. They may also, in some cases, have different environmental
additionality implications. Different baseline approaches may also be most appropriately
applied for different project types. For example, projects with homogeneous output (e.g. in
the electricity sector) may be more suitable for multi-project baselines than highly site-
specific projects.
High baseline development costs will discourage potential investment in JI and CDM
projects. The relatively high cost of establishing project-specific baselines for AIJ pilot phase
projects has stimulated interest in less costly, more standardised approaches. Moreover,
although baselines that could be applied to more than one project may be costly to develop at
the outset, they could lead to economies of scale and facilitate project preparation once initial
guidance is established.
43
There are different views on whether or not the baseline-setting framework should be the
same for JI and CDM.
42
The environmental additionality of different emission projects is based on the difference

between the baseline and actual emissions. Different baseline approaches can influence the
environmental additionality of a project by influencing the level of gaming, free riders and
leakage associated with it. Assumptions used to develop the baseline and what is included in
that baseline can also have significant effects on these factors.
Gaming affects the level of a project baseline, and thus the resulting amount of emission
credits generated by a JI or CDM project. Gaming is likely to be more prevalent in project-
specific baselines where a large variation in underlying assumptions is expected. Free riders
affect the number of projects credited, allowing projects that would have happened without JI
or CDM credit to get credit anyway. The level of free riders could be significant if certain
project types are deemed a priori eligible to obtain credits (e.g. by having per unit output
emissions below a certain level). Leakage is a measure of how accurately the project
boundary reflects the overall environmental effects of a project. The importance of leakage is
likely to vary significantly by project type and may be lower in more aggregated baselines
(e.g. multi-project baselines) where project boundaries are wider.
Increased levels of gaming, free riders and leakage would artificially inflate acquisitions of
ERUs and CERs by investors. If not all CERs accruing from CDM projects were “real” the
environmental effectiveness of the CDM would be lessened because the environmental
benefits of CDM investment is not assured. Higher levels of gaming, free riders and leakage
in JI projects will not affect the environmental effectiveness per se, (because the overall
Annex I assigned amount would remain unaffected by acquisitions and transfers of JI credits).
However, it could increase the difficulties for host countries in meeting their individual Kyoto
commitments. Determining the relative importance of these different factors is difficult in the
absence of information on emission baseline levels and on the number and type of projects
likely to be undertaken under each approach.
This paper presents a preliminary survey of analyses done on the effect that different baseline
approaches and/or assumptions have on the level of credits generated by a particular project.
This paper also illustrates the influence that different national circumstances can have on the
absolute level of the baseline and the resulting emission credits. For example, if both India
and Brazil set a multi-project baseline for electricity projects at the level of current weighted
average emissions, potential gas-fired projects could generate certified emission reductions if
they were undertaken in India, but not in Brazil. Because of variable conditions in the
electricity sector among different countries it may not be appropriate to use a common multi-
project baseline assumption (e.g. average fossil fuel) across different countries
The review of empirical data presented in this paper, and the theoretical case study of
electricity in Brazil and India, indicate that the assumptions and approaches used to develop
baselines are independent variables. Both can influence the level of credits for a particular
project. In the examples examined here, the range in different possible assumptions within
one baseline approach can be as large as the range in assumptions between different baseline
approaches. However, not all approaches may be equally appropriate in all circumstances,
and different approaches may be viewed as more or less suitable for different types of projects
(e.g. forestry, electricity or landfill).
Both baseline assumptions and approaches will have an impact on the environmental
additionality of individual projects. They will also influence the overall environmental
effectiveness of the mechanisms, i.e. the contribution of JI or the CDM as a whole to
achieving the objectives of the Protocol.
The baseline approach will influence the environmental effectiveness of the mechanisms in
part through its impact on the complexity of setting up an emissions baseline. More complex
methods are likely to limit the number of projects initiated because of high transaction costs.
43
Baseline assumptions are also likely to have an impact on the effectiveness of the mechanisms
through their impact on the baseline stringency, which affects the level of credits, and through
this the number of projects initiated. Both stringent and lax baselines are possible under the
different baseline approaches. Overly stringent baselines are likely to limit the number of
projects initiated (and thus the amount of emission reductions achieved through the
mechanisms). Overly lax baselines may encourage greater overall participation in the project-
based mechanisms by raising project eligibility and the number of emission credits, but could
have a negative environmental effect.
In circumstances where different baseline approaches are plausible, the independence of

baseline stringency and approaches suggests that maximum environmental effectiveness
across the project-based mechanisms (as opposed to individual projects) is likely to be
achieved by optimising baseline stringency and reducing baseline complexity. In practice,
this means: (i) seeking to minimise baseline complexity, as long as the ability to determine
"what would have happened otherwise" is not compromised; and (ii) optimising the baseline
stringency so that it maximises the overall global environmental effectiveness from the
project-based mechanisms.
An optimal strategy takes into account that a high volume of projects will be needed to deliver
strong environmental effectiveness from the mechanisms: a higher number of good projects
will be more beneficial for the environment (in terms of total GHG reductions) than a lower
number of individually better projects. However, since the relationship between project
numbers/baseline stringency and between project numbers/baseline complexity are both
unknown, it is not yet possible to determine the levels at which baseline stringency or baseline
complexity will limit the number of projects initiated.
There may be differences in the most desirable baseline approach for different project types in
different sectors. This would need to be examined further, as would how to minimise
complexity within the different approaches. Any future decisions on baseline approaches and
assumptions may therefore aim to reduce the complexity of baseline-setting procedures and to
set the baseline stringency so that it properly balances environmental benefit, cost, and
participation.
A decision on baseline approach(es) would open the door for assessment and eventual
agreement on other credit-related aspects of JI and CDM projects, such as:
• how to determine the value of key assumptions;

• how long any project should be allowed to generate emission credits for (the emissions
timeline),
• whether or not emissions baselines should be static or dynamic;
• maximising the environmental effectiveness of the project-based mechanisms by
minimising the level of leakage, free riders and gaming, and
• assessing how to determine the additionality of a climate-friendly project in a country that
has low emission levels.
44
Annex A: Possible reporting requirements for baselines

The current reporting requirements for emission baselines for AIJ projects are laid out in the
UNFCCC’s Uniform Reporting Format (URF)44. These reporting requirements cover all
aspects of an AIJ project (including e.g. contact points, financial aspects etc.) as well as an
estimation of emissions “with the project activity” (section E2) and “without the activity
(project baseline)” (section E1). However, the guidance in the current URF of what should be
reported in these sections is limited to a request for a “description of the baseline or reference
scenario, including methodologies applied” and a “description of the scenario, including
methodologies applied”. The current URF does not, for example, request that underlying data
(such as fuel use, fuel emission factors, technology efficiencies etc.) be reported, although
some AIJ project reports do in fact include some or all of this information.
Emission baselines may need to be reported to the UNFCCC both for CDM and for JI
projects. It may be decided that the URF for these project reports may need to be more
detailed, and/or that more specific guidance is included, such as examples of what
could/should be reported for different project types. More detailed reporting formats could
help to reduce the time and cost of developing a reporting format for individual project
developers, and could also help to reduce data-related costs by indicating which data are
needed for which project type. This is a matter which will need to be decided in the
negotiating fora. Nevertheless, possible more detailed reports for project-based baseline
reports in selected sectors could include the following outlined overleaf.
Forestry: Re/afforestation projects (from MacDicken, 1997)
General information: Local name of project site

Address, State, Country
Latitude
Longitude
Elevation (m)
Species (before and after project)
Site history since last inventory: (description of significant changes in management, pest and
disease problems, harvesting or other mortality).
Month of inventory determination
Carbon pool Area (ha) Mean carbon Total carbon Confidence

density (Mg/ha) (Mg) interval (Mg)
Reference case NA
Above-ground
Below-ground
Forest floor
Soil to depth of 30cm
Total- reference case
Project case ha
Above-ground
Below-ground
Forest floor
Soil to depth of 30cm
Total- project case
NET CARBON STORED
44
Draft revisions to this URF are outlined in FCCC/SB/1999/5/Add.1.
45
Energy supply: replacement or refurbishment of isolated electricity/heat production facilities

(from Ellis 1999)45
General information
• name of project site;
• address, country
What is the distance of the proposed plant from:
• the nearest gas grid;
• the nearest district heating system?
Is the plant currently connected to the electricity grid?
Old plant Retrofit/converted plant

Fuel(s) used* > Fuel(s) used* >
Emission factor(s) (t C/TJ) > Emission factor(s) (t C/TJ) >
Average annual quantity of > Projected annual fuel use in >
fuel used in old plant retrofitted plant (TJ or toe**)
(TJ or toe**)
Average annual energy Projected annual energy
production at old plant: > (electricity) production at modified plant: > (electricity)
> (heat) > (heat)
Age of plant (years) > Size and/or type of new > (size, MW)
equipment > (type, e.g. boiler)
Expected life of plant at > Is this new equipment added >
construction (years) or is it replacing other eqpt.?
Description of plant - Expected life of retrofit plant >
technology.
* If more than one fuel used, state approximate proportion of each
** It is important that this figure is expressed in terms of energy units in order to be consistent
with the expression of the emission factor, and for comparison (transparency) purposes.
45
Reporting the information in the table would help to determine if emissions with the project
were lower than emissions prior to implementing the project. However, it would not
determine whether, for example, an existing plant would have been retrofitted anyway, or
whether emissions would have been expected to decrease anyway because of decreasing
demand. Such information could be requested either in this part of a new URF, or in the part
of the URF indicating why the particular project fits with national economic development
and socio-economic and environment priorities and strategies. The type of information
reported for a new plant, and for a grid-connected plant may be different (because of its
interaction with other plants in the same system).
46
Energy efficiency (from Vine and Sathaye, 1999a)
ENERGY USE AND CARBON EMISSIONS
Table A1: Estimated Energy Use and Carbon Emissions in Baseline [At Time of
Project Registration]
Estimate annual energy use and carbon emissions (1) for the unadjusted baseline (without free riders), (2) free
riders, and (3) for the baseline (adjusted for free riders). Indicate the level of precision for each value.
Estimated Unadjusted Level of Free Level of Without - Level of

Baseline (1) Precision a Riders (2) Precision a Project Precision a
Baseline
(3=1-2)
On-site fuel use
(Terajoules =
1012 joules/yr.)
Carbon emissions
factor b
Type of fuel:
Carbon emissions -
(tC/yr.)
On-site electricity
use (MWh/yr.)
Carbon emissions
factor b
Type of fuel:
Carbon emissions
(tC/yr.)
Off-site fuel use

(Terajoules =
1012 joules/yr.)
Carbon emissions
factor b
Type of fuel:
Carbon emissions
(tC/yr.) c
Off-site
electricity
use(MWh/yr.)
Carbon emissions
factor b
Type of fuel:
Carbon emissions
(tC/yr.) c
TOTAL
Carbon emissions
(tC/yr.)
a Indicate the level of precision used for project values: use either (1) standard deviation around the mean value,
or (2) general level of precision (e.g., low, medium, high) Ñ if more information is available, additional levels of
precision can be used.
b Specify type of fuel used for calculating carbon emissions factor.
c Indicate carbon reductions from off-site electric utility plant(s).
47
Table A2: Estimated Gross Changes in Energy Use and Carbon Emissions
from Project [At Time of Project Registration]
Estimate annual energy use and carbon emissions (1) for the unadjusted project, (2) from
positive project spillover, (3) from market transformation, and (4) for the “with-project”
scenario. Indicate the level of precision for each value.
Estimated Unadjusted Level of Free Level of Without - Level of

Baseline (1) Precision a Riders (2) Precision a Project Baseline Precision a
(3=1-2)
On-site fuel use
(Terajoules =
1012 joules/yr.)
Carbon
emissions factor b
Type of fuel:
Carbon -
emissions (tC/yr.)
On-site
electricity use
(MWh/yr.)
Carbon
emissions factor b
Type of fuel:
Carbon
emissions (tC/yr.)
Off-site fuel use

(Terajoules =
1012 joules/yr.)
Carbon
emissions factor b
Type of fuel:
Carbon
emissions
(tC/yr.) c
Off-site
electricity
use(MWh/yr.)
Carbon
emissions factor b
Type of fuel:
Carbon
emissions
(tC/yr.) c
TOTAL
Carbon
emissions (tC/yr.)
48
Table A3: Estimated Net Changes in Energy Use and Carbon Emissions from
Project [At Time of Project Registration]
Calculate the net change in annual energy use and carbon emissions by subtracting “with-project” values (taken
from Table A2) from “without-project baseline” values (taken from Table A1). Indicate the level of precision for
each value.
Estimated Unadjusted Level of Free Level of Without - Project Level of

Baseline (1) Precision Riders (2) Precision a Baseline (3=1-2) Precision a
a
On-site fuel use

(Terajoules =
1012 joules/yr.)
Carbon
emissions
factor b
Type of fuel:
Carbon -
emissions
(tC/yr.)
On-site
electricity use
(MWh/yr.)
Carbon
emissions
factor b
Type of fuel:
Carbon
emissions
(tC/yr.)
Off-site fuel use
(Terajoules =
1012 joules/yr.)
Carbon
emissions
factor b
Type of fuel:
Carbon
emissions
(tC/yr.) c
Off-site
electricity
use(MWh/yr.)
Carbon
emissions
factor b
Type of fuel:
Carbon
emissions
(tC/yr.) c
TOTAL
Carbon
emissions
(tC/yr.)
49
Annex B: Data used to assess different possible baseline assumptions in India and Brazil
Table B 1 : CO2 emissions from electricity generation, Brazil
1996 electricity 1996 Source 25% less CO2 diff b/w source planned new tech. diff. b/w source average average best available diff. B/w source
output (GWh) specific average intensive than average and 25% In Brazil and planned tech. technology average and BAT
(t CO2/GWh) average more efficient
(t CO2/GWh) (t CO2/GWh) (t CO2/GWh) (t CO2/GWh) (t CO2/GWh) (t CO2/GWh)
Nuclear(i) 2429 (0.8%) 0,00 0.00 0.00 0 0.00 0.00 0.00
Coal (ii) 4764 (1.6%) 1102 827 275 1042 60.2 781 320
Oil (iii) 8929(3.1%) 776 582 194 n.a. n.a. n.a. n.a.
Gas(iv) 714 (0.2%) 742 557 186 404 338 382 361
Bio & sugar cane** 7208 (2.5%) 0.0 0.0 0.0 n.a. 0.00 n.a. 0.00
Hydro** 265773 (92%) 0.0 0.0 0.0 n.a. 0.00 n.a. 0.00
TOTAL 289817 437 328 -
Average (foss.fuel) 873 654 219
Weighted average (all sources) 43.9 32.9 10.97
Weighted average (thermal only) 755 566 189
Weighted average (fossil only) 882 662 221
50
Table B 2: Cost of electricity generation, Brazil
Investment cost of new Total cost (incl. O&M +fuel) total cost of average BAT approx. incremental cost of BAT (US
plant ('96Uscents/KWh) of new plant ('96US cents/KWh) cents/KWh)
(1996UScents/KWh)
Nuclear (i) 3.58* 5.15* 5.35* 0.2
Coal (ii) 2.41* 4.3 * 5.15* 0.85
Oil (iii) n.a. 5.18* n.a. n.a.
Gas (iv) 1.41* 3.38* 4.77 * 1.39
Wind n.a. 4.87* 6.4 ** 1.53
Biomass/sugarcane residues* 3.27-6.54 n.a. (site-specific) n.a.
Hydro* 33% below 3.27 n.a. (site-specific) n.a.

39% b/w 3.27-5.75
28% above 5.75
Output and CO2 (fossil fuel combustion) data from 1998 IEA Statistics, IEA 1998b, IEA 1998c
* Most cost data based on NEA/IEA (1998)
(i) nuclear assumptions : new plant is pressurised heavy water reactor; BAT includes reactors in several OECD countries, e.g. boiling water reactor, pressurised water reactor, advanced boiling
water reactor.
(ii) Coal assumptions: planned new tech (pulverised coal comb. @ 33% efficiency); BAT (clean coal, i.e. pulverised coal comb. (supercritical) @ 44% efficiency, based on average of several OECD
countries; mix of sub-bituminous(60%) & bituminous(40%)
(iii) Oil: no new oil-fired plants are planned in OECD (except Turkey)
(iv) Nat. Gas assumptions: planned new tech(CCGT@50% efficiency); BAT([email protected]% efficiency, based on average of several OECD countries)
** Renewables : zero value given to emissions, as per IPCC recommendations (however, lifecycle emissions estimates available in IEA 1998b; cost estimates from IEA 1997b
51
Table B 3: CO2 emissions from electricity generation, India
1996 electricity 1996 Source 25% less CO2- diff b/w source planned new diff. b/w source Average best diff.b/w source
output (GWh)* specific average intensive than average and tech. average and available average and BAT
(tC02/GWh) average 25% more In India planned new technology
efficient tech.
(t CO2/GWh) (t CO2/GWh) (t CO2/GWh) (t CO2/GWh) (t CO2/GWh)
(t CO2/GWh)
Nuclear (i) 8400 (1.9%) 0,00 0,00 0,00 0 0.00 0.00 0.00
Coal (ii) 318357 (73%) 1222 917 306 1017 205 786 436
Oil (iii) 12000(2.8%) 670 503 168 n.a. - n.a. -

Gas (iv) 27189(6.2%) 515 386 129 403 111 382 133
Renew.(wind..)** 57 (0.01%) 0.0 0.0 0.0 n.a. - n.a. -
Hydro** 69072(15.9%) 0.0 0.0 0.0 n.a. - n.a. -
TOTAL 435075 401 301 101
Average (fossil fuels) 802 602 201
Weighted average (all sources) 945 709 236
Weighted average (thermal only) 1123 843 281
Weighted average (fossil only) 1150 862 288
52
Table B 4: Cost of electricity generation, India
Investment cost of new plant Total cost (incl. O&M +fuel) of average total cost of BAT Approx. Incremental cost
('96UScents/KWh) new plant (' 96UScents/KWh) (UScents/KWh) of BAT (US cent/KWh)
Nuclear (i) 3.73 5.19 5.35 0.16
Coal (ii) 1.59 4.23 5.15 0.92
Oil (iii) n.a. n.a. n.a. n.a.
Gas (iv) 2.18 4.50 4.77 0.27
Wind n.a. 6.44** -
Hydro (almost all cost is capital inv.) 5.51* n.a. -
- Output and (fossil fuel) CO2 data from the 1998 IEA Statistics, IEA 1998c, IEA 1998d
- Most cost estimates based on NEA/IEA (1998)
* Based on the IEA, World Energy Outlook assumptions
(i) nuclear assumptions : new plant is pressurised heavy water reactor; BAT includes reactors in several OECD countries, e.g. boiling water reactor, pressurised water reactor, advanced boiling
water reactor.
(ii)coal assumptions : planned new technology is pulverised coal combustion @ 35%efficiency; BAT is clean coal, i.e. pulverised coal comb, (supercritical) @ 44% efficiency, based on average of
several OECD countries; mix of lignite (6%), sub-bituminous(70%) & bituminous(24%), (NEA/IEA, 1998)
(iii) oil : no new oil-fired plants planned in OECD (except Turkey)
(iv) gas assumptions: As no India-specific data was available, costs estimated to be similar to those of Korea (in NEA/IEA, 1998) : planned new technology is CCGT @ 50% efficiency; BAT is
[email protected]% efficiency, based on average of several OECD countries (NEA/IEA, 1998)
Renewables (**) : zero value given to emissions, as per IPCC recommendations (however, lifecycle emissions estimates available in IEA 1998b; cost estimates from IEA 1997b
53
References
Begg, Katie, Stuart Parkinson, Tim Jackson, Poul-Erik Morthorst, Peter Bailey, 1999, Overall Issues for
Accounting for the Emissions Reductions of JI Projects, presented at Workshop on Baselines for
CDM, Tokyo, February 25 and 26, 1999, http://www.gispri.or.jp/
Carter, Lisa, 1997, Modalities for the Operationalization of Additionality, prepared for presentation at
UNEP/German Federal Ministry of Environment workshop on AIJ (Leipzig, March 1997)
Centre for Clean Air Policy (CCAP), 1999, Options for Determining Baselines and Additionality under
Article 6 JI, presentation at SBSTA 10
Chomitz, K. M., 1998, Baselines for Greenhouse Gas Reductions: Problems, Precedents and Solutions,
World Bank
Department of Industry, Science and Resources (DISR), 1999, Workbook for Calculating Greenhouse Gas
Reductions from Projects Using Renewable Energy, Canberra,
http://www.dpie.gov.au/resources.energy/environment/greenhouse/igp/index.html
Ellis, Jane, 1999, Experience with Emission Baselines under the AIJ Pilot Phase, OECD/IEA Information
Paper, OECD, http://www.oecd.org/env/cc/freedocs.htm
Emmer, Ignio, J.A. Verweij and J.M. de Ligt, 1999, Reforestation and flexible instruments – from project
implementation to certification presented at Project types for Flexible Instruments – the situation
after Buenos Aires (January 28-29, 1999)
Ernst Basler + Partners, 1999, Swiss-Czech Cooperation Project “Conversion of Heating Centers”:
Activities Implemented Jointly (AIJ) Study (prepared for the Swiss Federal Office for Foreign
Economic Affairs), Zurich
Fearnside, Philip M., 1999, Forests and global warming mitigation in Brazil: opportunities in the Brazilian
forest sector for responses to global warming under the “clean development mechanism” in
Biomass & Bioenergy, 16 (1999) 171-189
Friedman, Shari, 1999, Benchmarking for Project-Level GHG Emission Reductions, presentation at
UNFCCC Technical workshop on Mechanisms, April 9-15 1999
Friedman, Shari, 1999b, The Use of Benchmarks to Determine Emissions Additionality in the Clean
Development Mechanism, presented at Workshop on Baselines for CDM, Tokyo, February 25 and
26, 1999, http://www.gispri.or.jp/
Hassing, Paul and Matthew S. Mendis, 1999, An International Market Framework for CDM Transactions,
paper prepared for the UNFCCC Technical workshop on Mechanisms, April 9-15 1999
IEA, 1997, Activities Implemented Jointly: Partnerships for Climate and Development, Paris
IEA, 1997b, Renewable Energy Policy in IEA Countries- Volume I : Overview, Paris
IEA, 1998, World Energy Outlook, Paris
IEA, 1998b, Benign Energy? The Environmental Implications of Renewables, Paris
54
IEA, 1998c, Energy Statistics & Balances of Non-OECD Countries: 1995-1996, Paris
IEA, 1998d, CO2 Emissions from Fuel Combustion, Paris
Jackson, T and Begg, K, 1999, Accounting and Accreditation of Activities Implemented Jointly: Final
Report, Prepared for DGXII of the European Commission under the Environment and Climate
Programme 1994-1998, March 1999
JIRC (Joint Implementation Registration Centre), 1997, Joint Implementation Registration and
Certification Procedure, The Netherlands
Lile, Ronald, Mark Powell and Michael Toman, November 1998, Implementing the Clean Development
Mechanism: Lessons from US Private-Sector Participation in Activities Implemented Jointly,
Resources for the Future discussion paper, available at http://www.rff.org/disc
papers/abstracts/9908.htm
MacDicken, K. G., 1997, A Guide to Monitoring Carbon Storage in Forestry and Agroforestry Projects,
Winrock International Institute for Agricultural Development, available at
http://www.winrock.org/REEP/forest_carbon_monitoring_program.htm
Michelowa, Axel, 1998, Joint Implementation - the baseline issue: economic and political aspects, Global
Environmental Change, Vol 8, No. 1
Michaelowa, Axel, 1999, Project-specific, benchmark, top-down approaches for baselines and
additionality: a comparison, paper prepared for the UNFCCC Technical workshop on Mechanisms,
April 9-15 1999
Nuclear Energy Agency (NEA) and the International Energy Agency (IEA), Projected Costs of Generating
Electricity – Update 1998, OECD France, 1998
Nordic Environment Finance Corporation (NEFCO): Joint Implementation Study of Nordic Multilateral
Pilot Projects; Helsinki/Stockholm, 1996
Puhl, Ingo, 1998, Status of Research on Project Baselines Under the UNFCCC and the Kyoto Protocol,
OECD/IEA Information Paper, OECD, available at http://www.oecd.org/env/cc/freedocs.htm
Puhl, Ingo, Tim Hargrave and Ned Helme, 1998, Options for Simplifying Baseline Setting for Clean
Development Mechanism Projects, Center for Clean Air Policy, Washington DC
Trexler and Associates, 1999a, Benchmarking the Additionality of Land Use Change and Forestry
Mitigation Projects: Identifying Issues and Potential Priorities
Trexler, Mark and Rebecca Gibbons, 1999b, Advancing the development of forestry and land-use based
project baseline methodologies for the developing CDM Regime, presented at Workshop on
Baselines for CDM, Tokyo, February 25 and 26, 1999, http://www.gispri.or.jp/
UNDP and WRI, 1999, Promoting development while limiting greenhouse gas emissions: Trends &
Baselines, New York
UNFCCC, 1995, Report of the Conference of the Parties on its First Session, Held at Berlin from 28
March to 7 April 1995, FCCC/CP/1995/7/Add.1
55
UNFCCC, 1998a, COP4 Decisions and Resolutions, FCCC/CP/1998/16/Add.1
UNFCCC, 1998b, Review of the Implementation of Commitments and of other Provisions of the
Convention. Activities Implemented Jointly: Review of Progress under the Pilot Phase (Decision
5/CP.1), Second synthesis report on activities implemented jointly, FCCC/CP/1998/2
UNFCCC, 1999, Principles, Modalities, Rules and Guidelines for the Mechanisms under Articles 6, 12 and
17 of the Kyoto Protocol: Submissions from Parties, FCCC/SB/1999/MISC.3 and Add.5
UNFCCC, 1999b, Activities Implemented Jointly under the Pilot Phase: Issues to be addressed in the
review of the pilot phase, including the third synthesis report on activities implemented jointly,
FCCC/SB/1999/5 and Add.1
Venezuela, 1999, Principios que deben guiar el diseño y aplicacion del mecanismo de desarrollo limpio
(MDL) contemplade en el protocolo de Kioto, submission to UNFCCC
Vine, Edward and Jayant Sathaye, March 1999, Guidelines for the Monitoring, Evaluation, Reporting,
Verification, and Certification of Energy-Efficiency Projects for Climate Change Mitigation, Report
number LBNL-41543, Lawrence Berkeley Laboratory, USA, http://eetd.lbl.gov/ea/ccm/ccPubs.htm
Vine, Edward, Jayant Sathaye and Willy Makundi, March 1999, Guidelines for the Monitoring,
Evaluation, Reporting, Verification, and Certification of Forestry Projects for Climate Change
Mitigation, Report number LBNL-41877, Lawrence Berkeley Laboratory, USA
available at http://eetd.lbl.gov/ea/ccm/ccPubs.htm
WBGU (German Advisory Council on Global Change), 1998, The Accounting of Biological Sinks and
Sources under the Kyoto Protocol: a step forwards or backwards for global environmental
protection?, Special Report 1998, at http://www.awi-
bremerhaven.de/WBGU/wbgu_sn1998_engl.html
Yager, Andrew J and Hans J Mydske, 1998, Fuel Switch in Boilers in the Slovak Republic: report to the
Norwegian Pollution Control Authority, Institute for Energy Technology, Norway
Selected AIJ project reports submitted by Parties to the UNFCCC, http://www.unfccc.de/program/aij/
- Dutch micro-Hydel (Bhu/Nld)
- Tyumen (Nld/Rus)
- Cizkovice (Cze/Fra)
- Viljandi boiler conversion (Est/Swe) - 1997, 1998, top-down 1999, bottom-up 1999
- Jochy (Slv/Nor)
56
Glossary
AIJ Activities Implemented Jointly

BAT Best available technology
CDM Clean Development Mechanism (defined in Article 12 of the Kyoto Protocol)
CER Certified Emission Reductions (generated from CDM projects)
CH4 Methane
CO2 Carbon dioxide
COP Conference of the Parties to the United Nations Framework Convention
on Climate Change (UNFCCC)
EIT Countries with Economies in Transition
ERU Emission Reduction Unit (generated from Article 6 JI projects)
FCCC United Nations’ Framework Convention on Climate Change
GHG Greenhouse gases
IPCC Intergovernmental Panel on Climate Change
JI Joint Implementation (outlined in Article 6 of the Kyoto Protocol)
KP Kyoto Protocol
kt thousand tons
LUCF Land Use Change and Forestry
N2O Nitrous oxide
OECD Organisation for Economic Co-operation and Development
URF Uniform Reporting Format (form on which countries submit AIJ
project-specific information to the UNFCCC)
57
Approach The general methodology used to calculate an emissions baseline (e.g. project-
specific,hybrid, or multi-project).
Assumption The underlying component(s) used to develop an emissions baseline. For a project-
specific baseline assumptions could include the technology type that would have
been used and the efficiency that would have been achieved in the absence of the
JI/CDM project. For a multi-project baseline, the potential assumption could be , for
example,. Average performance, marginal performance, best available technology
etc.
Baseline An emissions baseline is a hypothetical emission reference case representing the

estimated level of GHG emissions that would have been emitted in the absence of
the JI or CDM project.
Complexity The complexity of a baseline is a measure of how difficult it is to set up.

Complexity will increase with the number of parameters in a baseline). A more
complex baseline is likely to be more costly to set up than a simple baseline. A less
complex baseline could also facilitate the verification process.
Credits Unit used for the measurement, (e.g. in tonnes of CO2-equivalent), transfer and
acquisition of emission reductions associated with JI and CDM projects.
Dynamic baseline A baseline that is revised during the lifetime of the JI/CDM project, without
specifications at the outset on how the revision(s) will be made.
Emissions timeline Time (number of years) over which emission credits resulting from a JI or CDM
project accrue.
Environmental additionality Difference between baseline emissions and actual emissions for a JI/CDM project.
Environmental additionality of JI and CDM projects is required by the Kyoto
Protocol
Environmental credibility Quality of a baseline with respect to reflecting the emission level that would occur
without the JI or CDM project(s). Given the hypothetical nature of emission
baselines, environmental credibility is difficult to assess (even on an ex post basis).
Environmental effectiveness Extent to which the JI and CDM regimes, under different baseline approaches and
assumptions, result in maximum emissions reductions and maximum participation,
thereby contributing to achieving the objectives of the Kyoto Protocol.
Free riding A situation whereby a project generates emission credits from JI or CDM projects
even if the project would have gone ahead in the absence of JI or CDM, i.e. is not
“additional”. Free riding therefore affects the numbers of projects obtaining credits.
Gaming Actions or assumptions taken by the project developer and/or project host that would
artificially inflate the baseline (and therefore the credits from the project). Gaming
generally leads to lax baselines.
Greenfield projects New projects, e.g. those taking place at a new site and increasing the host country’s
capacity for heat output, electricity generation etc. (as opposed to refurbishment
projects that build on existing projects)
Hybrid baseline A baseline made up of different components that could be applied to a number of
projects. A hybrid baseline could combine some aspects of project-specific
58
baselines (e.g. incorporating some site-specific analysis) and some aspects of multi-
project baselines (e.g. to include some standardised emission values for different
technologies).
Leakage Leakage occurs if actual emission reductions (or increase in sinks) from a project
results in emissions increasing (or sinks decreasing) elsewhere. Leakage occurs if
the system boundaries do not capture all emission effects of a project.
Multi-project baselines Emission baselines (also defined as “benchmarks”, “activity standards”, “intensity
indicators” in the literature) that can be applied to a number of similar projects, for
example within the same sector of a country or region. These emission baselines
may be drawn up using an engineering approach to set standard baseline values for a
particular technology or industrial sector.
Project-specific baseline Project-specific emission baselines are those that have been drawn up for individual
projects by examining them on a case-by-case basis. Each project-specific baseline
is used only for the project for which it was developed. The project-specific
approach is based on some combination of engineering judgement and site-specific
analysis.
Refurbishment projects Projects in which existing equipment/processes is upgraded or replaced.
Standardised assumptions Assumptions, e.g. on emission performance of a technology, that are standardised
(generalised). These standardised assumptions may be used in the emission
baselines for a number of projects.
Static baseline A baseline that is fixed at the start of the JI/CDM project and
remains fixed for the duration of the project. It is possible to have a static baseline
for which the level changes over time, as long as this is specified at the outset of the
project.
Static baseline A baseline that is fixed at the start of the JI/CDM project and remains fixed for the
duration of the project. It is possible to have a static baseline for which the level
changes over time, as long as this is specified at the outset of the project.
Stringency (of a baseline) The stringency of a baseline is a measure of how difficult it is for projects to
generate emissions below the baseline level. Given the uncertainty and
measurement errors inherent in establishing a baseline, there is a range of plausible
emission levels for any particular baseline. A stringent baseline will be at the low
end of that range, and a lax baseline will be at the high end of that range. The
determination of whether a particular baseline is stringent or not depends on the
project types and national/regional circumstances.
Transaction costs The costs associated with the process of obtaining JI or CDM recognition for a
project and obtaining the resulting emissions credits. Transaction costs would
include, for example, costs of developing a baseline and assessing the
“additionality” of a project, costs of obtaining host country approval, monitoring and
reporting, etc. Transaction costs would not include the direct investment,
maintenance and operational costs of the project.
Transparency A transparent baseline is one where the construction is clearly explained, where the
parameters and methods used in the baseline are referenced and traceable, and where
the baseline can be reconstructed by a third party. Transparency is also a quality
that can reduce investors’ uncertainty associated with JI and CDM projects and the
credits that they can generate.
59
Value of Credits The monetary value resulting from the supply and demand of emission credits
resulting from JI and CDM projects.
60

Options For Project Emission Baselines

Uploaded by

Copyright:

Available Formats

Options For Project Emission Baselines

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Options For Project Emission Baselines

Uploaded by

Copyright:

Available Formats

COM/ENV/EPOC/IEA/SLT(2000)8

OECD ENVIRONMENT DIRECTORATE

OPTIONS FOR PROJECT EMISSION BASELINES

Organisation for Economic Co-operation and Development 2000

Questions and comments should be sent to:

EXECUTIVE SUMMARY .......................................................................................................................... 5

ANNEX A: POSSIBLE REPORTING REQUIREMENTS FOR BASELINES .................................. 45

ANNEX B: DATA USED TO ASSESS DIFFERENT POSSIBLE BASELINE ASSUMPTIONS IN

Table 1: TERMINOLOGY OF DIFFERENT BASELINE APPROACHES..................................................... 14

Table B1: C02 Emissions from Electricity Generation, Brazil ............................................................................. 50

Figure 1: REPRESENTATION OF DIFFERENT BASELINE APPROACHES ............................................... 13

The environmental additionality of individual projects is clearly correlated to the level or

• how to determine the value of key assumptions;

1.1 Objectives and approach

2. Baseline approaches described

• project-specific: baseline determined on a case-by-case basis, with project-specific

Figure 1: Representation of different baseline approaches

Project-specific Hybrid Multi-project

Terminology and baseline approaches

Table 1: Terminology of different baseline approaches

Baseline approach Description Calculation of annual Credits

Multi-project (various designs):

2.1 Project-specific baselines

Project-specific emission baselines are elaborated on a case-by-case basis and could, by

2.2 Multi-project baselines

Multi-project baselines could be set up at a fairly disaggregate level, such as by technology

2.3 Hybrid project baselines

“Hybrid” project baselines could therefore be designed to be applicable to a small number of

Table 2: Potential for standardising different components of emission

Baseline Project types for Potential for Comment

Development of standardised components for hybrid or multi-project emission baselines

Baseline Emission Project lifetime Case in which baseline applies

3.1 Dynamic vs Static baselines

continued proportion of fossil fuels in electricity supply. Such a baseline would be

Technological improvements or process changes can also be relatively sudden and

3.2 Emissions timeline

3.3 Data sources

4. Quantitative comparison of baseline approaches

4.1 Survey of experience with different baseline approaches

4.1.1 Conversion of heat plants in the Czech Republic

Table 4: Variation in calculated carbon offsets under different baselines:

4.1.2 Boiler conversions in Estonia

4.1.3 Other projects

Report Baseline Timeline Annual emissions (t CO2) Cumulative

4.2 Survey of experience with different baseline assumptions

Figure 2: Comparison of Decin Project Baselines

Comparison of Decin Project Baselines

25,000 Hist Bnchmk

Source: CCAP 1999

4.3 Case study simulations with multi-project baselines

Assumptions and Caveats

4.3.1 Electricity Generation in Brazil

Given that projections indicate a significant growth in electricity generation, it is interesting to

Figure 3: Effect of different baseline assumptions on baseline levels, Brazil

Of course, the actual

usual investment plans) that

4.3.2 Electricity Generation in India

Figure 4: Effect of different assumptions on baseline levels, India

However, the lower