Technical ReportPDF Available

Has Joint Implementation reduced GHG emissions? Lessons learned for the design of carbon market mechanisms

Abstract and Figures

This study systematically evaluates the environmental integrity of Joint Implementation (JI) in the first commitment period of the Kyoto Protocol. Our analysis indicates that about three-quarters of JI offsets are unlikely to represent additional emissions reductions. This suggests that the use of JI offsets may have enabled global GHG emissions to be about 600 million tonnes of carbon dioxide equivalent higher than they would have been if countries had met their emissions domestically. Of the six largest project types assessed in more detail, we find only one – N2O abatement from nitric acid production – had overall high environmental integrity. Our evaluation clearly shows that oversight of an international market mechanism by the host country alone is insufficient to ensure environmental integrity. The paper makes recommendations for the ongoing review of the JI Guidelines, for carbon markets generally, and for a new climate agreement.
Content may be subject to copyright.
Anja Kollmuss, Lambert Schneider and Vladyslav Zhezherin
Has Joint Implementation reduced GHG emissions?
Lessons learned for the design of carbon market mechanisms
Stockholm Environment Institute, Working Paper 2015-07
Stockholm Environment Institute
Linnégatan 87D
104 51 Stockholm
Sweden
Tel: +46 8 674 7070
Fax: +46 8 674 7020
Web: www.sei-international.org
Author contact:
Anja Kollmuss,
anja.kollmuss@sei-international.org
Director of Communications: Robert Watt
Editors: Elaine Beebe and Marion Davis
Cover photo: A steel plant in Ukraine with a coal waste heap in the background.
Photo © Mykola Ivashchenko.
This publication may be reproduced in whole or in part and in any form for educa-
tional or non-profit purposes, without special permission from the copyright holder(s)
provided acknowledgement of the source is made. No use of this publication may be
made for resale or other commercial purpose, without the written permission of the
copyright holder(s).
About SEI Working Papers:
The SEI working paper series aims to expand and accelerate the availability of our
research, stimulate discussion, and elicit feedback. SEI working papers are work in
progress and typically contain preliminary research, analysis, findings, and recom-
mendations. Many SEI working papers are drafts that will be subsequently revised
for a refereed journal or book. Other papers share timely and innovative knowledge
that we consider valuable and policy-relevant, but which may not be intended for
later publication.
Copyright © August 2015 by Stockholm Environment Institute
STOCKHOLM ENVIRONMENT INSTITUTE
WORKING PAPER NO. 2015-07
Has Joint Implementation reduced GHG emissions?
Lessons learned for the design of carbon market mechanisms
Anja Kollmuss, Lambert Schneider
Stockholm Environment Institute U.S. Centre
Vladyslav Zhezherin
Independent consultant
ABSTRACT
This study systematically evaluates the environmental integrity of Joint Implementation (JI) in the
first commitment period of the Kyoto Protocol. Our analysis indicates that about three-quarters
of JI offsets are unlikely to represent additional emissions reductions. This suggests that the use
of JI offsets may have enabled global GHG emissions to be about 600 million tonnes of carbon
dioxide equivalent higher than they would have been if countries had met their emissions
domestically. Of the six largest project types assessed in more detail, we find only one N2O
abatement from nitric acid production had overall high environmental integrity. Our
evaluation clearly shows that oversight of an international market mechanism by the host
country alone is insufficient to ensure environmental integrity. The paper makes
recommendations for the ongoing review of the JI Guidelines, for carbon markets generally,
and for a new climate agreement.
ACKNOWLEDGEMENTS
We thank Oleksandr Baskov for his research assistance and Boris Orlowsky for creating the
random sample for us. We thank (in alphabetical order) Karoliina Anttonen, Christopher Brandt,
Lennard de Klerk, Piotr Dombrowicki, Angela Friedrich, Roland Geres, Thomas Kleiser, Trine
Kopperud, Harri Laurikka, Michael Lazarus, Laurence Mortier, Konrad Raeschke-Kessler, Ingo
Ramming, Igor Shishlov, Kevin Tempest, and those who preferred to remain anonymous for
helpful input and comments. This report has been commissioned by the Austrian Federal
Ministry of Agriculture, Forestry, Environment and Water Management, the Ministry of the
Environment of Finland, and the Federal Office for the Environment of Switzerland. The views
expressed are those of the authors and do not necessarily reflect the official views of the
Austrian, Finnish and Swiss governments.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
2
CONTENTS
Executive Summary ............................................................................................................ 5
Introduction ...................................................................................................................... 5
Impact of JI on global emissions ........................................................................................ 5
Plausibility of additionality claims ...................................................................................... 5
Environmental integrity of key project types ........................................................................ 6
Environmental integrity of Track 1 and Track 2 .................................................................. 8
Accredited Independent Entities ......................................................................................... 8
Differences among host countries ...................................................................................... 9
Lessons learned for the design of crediting mechanisms..................................................... 9
Implications for the reform and future of JI ...................................................................... 10
Implications for market mechanisms under a new climate agreement .............................. 11
1. Introduction ................................................................................................................ 13
1.1 Purpose of this study ................................................................................................. 13
1.2 Research approach ................................................................................................... 13
2. Joint Implementation: an overview ................................................................................ 17
2.1 JI governance structures ............................................................................................ 17
2.2 JI rules and procedures on environmental integrity .................................................... 19
3. When does the environmental integrity of JI projects matter? .......................................... 22
3.1 Is the JI project additional, and are emission reductions correctly credited? ................ 22
3.2 Does the host country have a surplus of AAUs? ......................................................... 23
3.3 Are the JI emission reductions reflected in the host country’s GHG inventory? ............. 24
3.4 Impact of environmental integrity of JI projects on global emissions ........................... 25
4. Assessment of approaches for demonstrating additionality ............................................. 28
4.1 Identification of alternative scenarios ......................................................................... 28
4.2 Investment analysis ................................................................................................... 29
4.3 Barrier analysis ......................................................................................................... 31
4.4 Common practice analysis ........................................................................................ 32
4.5 Prior consideration .................................................................................................... 33
4.6 Retroactive crediting of emissions reductions .............................................................. 35
4.7 Demonstration of additionality by reference to a comparable project ......................... 36
4.8 Overall assessment of the likelihood of additionality of JI projects .............................. 37
5. Assessment of specific JI project types ........................................................................... 40
5.1 Spontaneous ignition of coal waste piles.................................................................... 41
5.2 Energy efficiency in industry and power production and distribution ........................... 49
5.3 Associated petroleum gas utilization .......................................................................... 53
5.4 Natural gas transportation and distribution ............................................................... 58
5.5 HFC-23 and SF6 abatement ...................................................................................... 61
5.6 N2O abatement at nitric acid plants .......................................................................... 65
5.7 Summary of findings by project type .......................................................................... 72
6. Assessment by country ................................................................................................. 73
6.1 Ukraine .................................................................................................................... 75
6.2 Russia ....................................................................................................................... 80
6.3 EU Member States as JI host countries ....................................................................... 84
6.4 Poland ...................................................................................................................... 85
6.5 Germany .................................................................................................................. 89
6.6 Comparison of environmental integrity of the project portfolio by country .................. 92
7. Assessment of differences between Track 1 and Track 2 ................................................. 93
7.1 Size difference between Track 1 and Track 2 projects ................................................ 94
7.2 Environmental integrity of Track 1 and Track 2 projects ............................................. 96
8. Assessment of accredited independent entities ............................................................... 97
9. Conclusions and recommendations ............................................................................ 101
9.1 Lessons learned for the design of crediting mechanisms .......................................... 102
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
3
9.2 Implications for the reform and future of JI .............................................................. 103
9.3 Implications for market mechanisms under the new climate agreement .................... 104
References .................................................................................................................... 106
Annex 1: Recommendations for the review of the JI guidelines .......................................... 112
A1.1 State of the review of the JI Guidelines .................................................................. 112
A1.2 Project cycle overview ........................................................................................... 112
A1.3 International oversight on host country implementation ......................................... 114
A1.4 Assessment of project and issuance review ............................................................ 115
A1.5 Accreditation of AIEs ............................................................................................. 116
A1.6 Registration of existing projects under the new JI rules ........................................... 116
A1.7 Crediting period ................................................................................................... 117
A1.8 Additionality requirements .................................................................................... 118
A1.9 Baseline requirements ........................................................................................... 119
A1.10 Features supporting environmental integrity: transparency, stakeholder consultations
and appeals procedures................................................................................................ 120
A1.11 Atmospheric benefit ............................................................................................ 121
A1.12 Issuance of ERUs during the interim period .......................................................... 122
Annex 2: List of random sample of 60 JI Projects ............................................................. 123
TABLES
Table 1: Categorization of project types used to establish a random sample ......................... 15
Table 2: ‘Other: CO2’ and ‘Other: non-CO2’ project type categories .................................... 16
Table 3: Requirements for countries' participation in JI and ERU transfer ............................... 18
Table 4: Bodies involved in JI Track 2 project procedure and their functions .......................... 19
Table 5: ERUs and registered projects by JI track .................................................................. 19
Table 6: Impact of JI projects on global emissions if the host country does not have significant
expected AAU surplus+ ......................................................................................................... 26
Table 7: Impact of JI projects on global emissions if the host country has a significant expected
surplus+ .............................................................................................................. 27
Table 8: Type of investment analysis used ............................................................................. 29
Table 9: Types of barriers cited in the sampled projects ........................................................ 32
Table 10: Project types evaluated, number of projects and their ERU shares .......................... 40
Table 11: JI project types not evaluated and rated ................................................................ 41
Table 12: Energy efficiency project types in industry and power production/distribution ......... 50
Table 13: Energy efficiency projects in industry and power production/distribution, by country ...
.............................................................................................................. 50
Table 14: Countries with nitric acid projects .......................................................................... 66
Table 15: Determination of baseline emission factor for nitric acid projects ........................... 70
Table 16: Assessment of overall environmental integrity by project types ............................... 72
Table 17: JI host countries: number of registered projects and their ERU issuance ................. 73
Table 18: Ukrainian JI projects by type ................................................................................. 78
Table 19: Russian JI projects by type ..................................................................................... 83
Table 20: Polish JI projects by type ....................................................................................... 88
Table 21: German JI projects by type ................................................................................... 91
Table 22: Ten largest JI projects ........................................................................................... 95
Table 23: Track 1 project types and their ERU issuance ......................................................... 96
Table 24: Track 2 project types and their ERU issuance ......................................................... 97
Table 25: Bodies involved in project procedure and functions per JI draft rules .................... 113
Table 26: Bodies involved in JI Track 2 project procedure and their functions ...................... 113
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
4
FIGURES
Figure 1: Accumulated issuance and ERU prices ................................................................... 17
Figure 2: Level of information provided on input parameters to the investment analysis ......... 30
Figure 3: Time period between project starting date and issuance of LoE ............................... 34
Figure 4: Issuance of LoE by year ......................................................................................... 35
Figure 5: Project cycle duration in sampled projects (time between LoE and ERU issuance) ....... 36
Figure 6: Plausibility of the additionality claims of the sample projects ................................... 39
Figure 7: Plausibility of additionality claims of projects by year of approval ........................... 39
Figure 8: Plausibility of additionality claims of projects by size ............................................... 40
Figure 9: Emission reduction claims by coal waste pile projects by emission source................ 42
Figure 10: LoA issuance dates of coal waste pile projects and emission reductions estimated in
PDDs .............................................................................................................. 45
Figure 11: Methane emissions from underground coal mining and implicit baseline emission
levels in CP1 without waste coal extraction JI projects............................................................ 49
Figure 12: Size distribution of 125 energy efficiency projects with ERU issuance ..................... 51
Figure 13: Time gap between project start date and earliest date of LoE or determination to
seek JI status .............................................................................................................. 51
Figure 14: Retroactive crediting in energy efficiency projects ................................................. 52
Figure 15: Time lag for APG JI projects between project operation start date and date of
determination report by ERU volume .................................................................................... 55
Figure 16: APG flaring in Russia: satellite estimates and national data in BCM ...................... 57
Figure 17: CO2e emissions from Russian APG flaring per tonne of oil produced and implicit
emission levels in CP1 without JI projects .............................................................................. 57
Figure 18: Project start years and years of LoE issuance in natural gas transportation and
distribution projects .............................................................................................................. 59
Figure 19: Specific emissions from natural gas distribution in Ukraine per volume of distributed
gas and implicit emission levels in CP1 without JI projects ..................................................... 61
Figure 20: Overall environmental integrity of project types by ERUs issued ............................ 72
Figure 21: JI host countries’ average 2003–2007 emissions as % of their CP1 target ............. 74
Figure 22: Estimated AAU surplus or shortage compared with the host country’s total initial
assigned amount for CP1 ..................................................................................................... 74
Figure 23: Ukraine: GHG emissions in base year and 19902012 ....................................... 75
Figure 24: Ukrainian JI project approval by year .................................................................. 78
Figure 25: Ukrainian ERUs issuance by year ......................................................................... 78
Figure 26: Russia: GHG emissions in base year and 1990-2012 .......................................... 80
Figure 27: Project approval and ERU issuance in Russia by year ............................................ 82
Figure 28: Retroactive crediting in Russia: number of projects and ERU volumes .................... 83
Figure 29: Poland: GHG emissions in base year and 1990-2012 ......................................... 86
Figure 30: JI project approval and ERU issuance in Poland by year ....................................... 87
Figure 31: Germany: GHG emissions in base year and 19902012 ..................................... 89
Figure 32: Environmental integrity of ERUs issued in each country by project type .................. 93
Figure 33: Number of projects approved by year, by track .................................................... 94
Figure 34: Average annual issuance (in kERUs) by year when projects received LoA .............. 95
Figure 35: Environmental integrity comparison of Tracks 1 and by ERUs issued to project types .
.............................................................................................................. 97
Figure 36: Number of projects determined and verified by AIEs ............................................ 99
Figure 37: AIE verifications by number of ERUs issued .......................................................... 99
Figure 38: Plausibility of additionality claims of the sampled projects by the AIE conducting
determination, by number of projects ................................................................................. 100
Figure 39: Plausibility of additionality claims of the sampled projects by AIE conducting
determination, by ERUs issuance ........................................................................................ 100
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
5
EXECUTIVE SUMMARY
Introduction
Joint Implementation is one of two offsetting mechanisms under the Kyoto Protocol, along
with the Clean Development Mechanism (CDM). It enables countries with emission reduction
commitments under the Kyoto Protocol to generate Emission Reduction Units (ERUs) from
greenhouse gas (GHG) reduction projects and transfer them to other countries. As of March
2015, almost 872 million ERUs had been issued under JI, about one-third of all Kyoto offset
credits. This study systematically evaluates the environmental integrity of JI in the first Kyoto
commitment period. Our analysis can directly inform the ongoing review of the JI Guidelines,
and is also deeply relevant to discussions about carbon markets more generally.
Impact of JI on global emissions
Our findings indicate that use of JI may have enabled global GHG emissions to be about
600 million tonnes of carbon dioxide equivalent (tCO2e) higher than they would have been
if countries had met their emissions targets domestically.
In principle, offsets are a zero-sum game for the atmosphere. Buyers of offsets can increase
their emissions by a corresponding amount above the target level, while emissions are
reduced by that amount in the host country, keeping global emissions the same. If offsets
come from non-additional or overcredited projects, however, using them will lead to an
increase in global emissions relative to a scenario without the use of offsets.
The design of JI should, in theory, avoid that outcome. Under the Kyoto Protocol, each
country with an emissions target receives allowances (called Assigned Amount Units, AAUs)
equivalent to its total emissions budget for the commitment period. For every ERU it issues, a
host country must cancel one AAU. Thus, if a JI project is overcredited or not additional, the
host country would have to make up the difference and engage in more mitigation action.
However, in the first commitment period, several countries had emissions targets well above
their BAU emissions, resulting in large AAU surpluses. In such cases, host countries can use
surplus AAUs to cover their ERUs, and will not have to engage in additional mitigation
action. Thus, non-additional or overcredited JI projects in those countries will lead to higher
global emissions.
Our analysis indicates that at about three quarters of ERUs are unlikely to represent additional
emissions reductions, and about 95% of the total ERUs were from countries with a significant
AAU surplus. This suggests that the use of JI may have enabled global GHG emissions to be
about 600 million tCO2e higher than they would have otherwise been. The implications for
the European Union’s Emissions Trading System (EU ETS) are particularly serious. As of
April 2015, more than 560 million ERUs had been used in the EU ETS. JI may therefore have
undermined the EU ETS emission reduction target by about 400 million tCO2.
Plausibility of additionality claims
In a random sample of 60 projects, the additionality claims do not seem plausible for 73%
of the ERUs issued and are questionable for another 12%.
We assessed the plausibility of additionality claims of JI projects through an in-depth review
of the information available for a sample of 60 projects, drawn in a representative manner
taking into account the host countries, project types and project scale. While this approach has
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
6
clear limitations it is a subjective judgment of the authors based on the limited information
publicly available it is based on a careful analysis applied in a consistent manner across
projects, assessing the plausibility of the timeline of project implementation and registration
under JI as well as the information on the main additionality tests used to determine
additionality (investment analysis, barrier analysis, common practice analysis, reference to a
comparable project). We used three broad categories to classify each project:
“Plausible” means that based on the available information, the claims for
demonstrating additionality seem plausible.
“Questionable” means that the available information raises questions or doubts about
the additionality.
“Not plausible” means that the available information suggests that the projects are
unlikely to be additional.
Figure ES-1 shows that for 43% of the projects and 73% of the ERUs the additionality claims
were not plausible based on the available information.
Figure ES-1: Plausibility of the additionality claims of the sample projects
Data source: Random sample of 60 registered projects
Environmental integrity of key project types
Of the six largest project types assessed in more detail, we find only one N2O abatement
from nitric acid production had overall high environmental integrity. For many JI
projects, either additionality seems unlikely, or unrealistic assumptions are used that
grossly overestimate the actual emission reductions. We find 80% of all ERUs come from
projects types with questionable or low environmental integrity.
The environmental integrity of the six project types with the highest ERU issuance was also
examined in more detail. These project types represent 84% of the ERUs issued and 53% of
registered projects in the first commitment period. Figure ES-2 and Table ES-1 provide an
overview of the results of our analysis.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
7
Table ES-1: Environmental integrity of the six largest JI project types
Project types
Registered
projects
Main countries
Overall envir.
integrity
Spontaneous ignition of coal piles
78
26%
all in Ukraine
Low
This project type avoids GHG emissions from uncontrolled fires from coal waste piles. Most JI projects extract
coal from the piles, leaving bare rock which does not ignite; others extinguish the fires.
Additionality not plausible: The timeline of project implementation shows that almost all projects were
registered in 2012 but were implemented at least four years earlier. Additionally is usually demonstrated by
long chains of reference to a similar project.
• Overcrediting likely to be very significant: Baseline emissions are overstated due to unrealistic
assumptions. All coal waste pile JI projects together implicitly claim that they have produced around 30% of all
coal in Ukraine. This is a highly unrealistic scenario.
Energy efficiency in industry and
power production and distribution
164
23%
mainly Ukraine
and Russia
questionable
This project type includes a large variety of energy efficiency improvement measures in diverse sectors, such as
large industrial facilities, and power and heat plants.
• Additionality questionable: Projects of this type are in many cases financially attractive without JI and JI
may only have sped up implementation. The additionality claims do not seem plausible for the majority of
projects, questionable for some, and plausible for a few.
• Overcrediting not assessed: Because of the wide variety of technologies and sectors that make up this
project type, we were unable to assess the overall validity of emission reduction claims.
Associated petroleum gas (APG)
utilization
22
14%
all in Russia
low
This project type utilizes associated petroleum gas that would otherwise be flared at oil field operations.
• Additionality not plausible: The timeline of project implementation shows that most projects, accounting for
almost 80% of ERUs, were implemented 69 years before their auditing and were registered even later.
• Overcrediting likely to be significant: The claimed reductions do not match Russia’s GHG inventory data,
suggesting overcrediting or inaccuracies in Russia’s inventory: The JI projects implicitly claim that in their
absence, Russian emissions from APG flaring in oil production would have increased well above any historical
values observed since 1990.
Natural gas transportation/distrib.
32
10%
mostly Ukraine
low
This project type involves reducing methane leaks from natural gas transportation and distribution or expanding
natural gas networks in order to replace coal or oil.
• Additionality not plausible: The project starting dates of the 30 projects located in Ukraine were between
2003 and 2006, while most projects received their Letter of Endorsement only in 2012.
• Some overcrediting likely: The network expansion projects assume that they solely replace fossil fuels such
as coal and heavy oil. But in rural areas newly available gas would also substitute biomass. The exclusion of the
use of biomass may inflate the baseline emissions. For projects addressing methane leaks, the implied leakage
rates in the absence of JI exceed historical emission rates reported in Russia's GHG inventory, which suggests
that either in the absence of the JI projects Ukraine’s emissions from this activity would have risen, or emission
reductions claimed by the projects are overestimated.
Abatement of HFC-23 and SF6
4
7%
mainly Russia
questionable
These projects incinerate HFC-23 and SF6 waste gas streams in industrial facilities.
• Additionality plausible: In the absence of regulations or other policies, this project type can be regarded as
likely to be additional because plant operators do not save costs or generate revenues from the installation of
abatement technology.
• Overcrediting likely to be very significant: Two of the four projects initially implemented a conservative
approach to calculate emission reductions. In 2011, safeguards to prevent perverse incentives were removed,
leading to significant over-crediting. One project assumed a baseline emission rate by far exceeding common
levels.
N2O abatement from nitric acid
43
5%
EU
high
These projects abate unwanted N2O that is generated as a by-product in nitric acid plants.
• Additionality plausible: In the absence of regulations or other policies such as the EU ETS, this project type
can be regarded as likely to be additional because plant operators do not save costs or generate revenues from
the installation of abatement technology.
• Overcrediting unlikely: Ambitious emission benchmarks based on European regulations (1.42.5 kg N2O/t
nitric acid) were used in Western Europe except Sweden. Higher values (4.313.5 kg N2O/t nitric acid) were
used in Eastern Europe and Sweden.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
8
Figure ES-2: Overall environmental integrity of project types by ERUs issued
Data source: Evaluation of the largest six project types, applied to the portfolio of 642 projects registered as of March
2014.
Environmental integrity of Track 1 and Track 2
Under the current rules, JI projects can be implemented under two tracks. Under Track 1, host
countries can largely establish their own rules for approving projects and issuing ERUs,
without international oversight. The host country can determine whether it deems emission
reductions as additional. Under Track 2, a UN body the Joint Implementation Supervisory
Committee (JISC) reviews projects and requests for ERU issuance and accredits JI auditors.
To date, 97% of ERUs have been issued under Track 1. Figure ES-3 compares the
environmental integrity of projects under the two tracks (the fraction of ERUs from projects
types we did not evaluate is shown in grey). The share of ERUs issued from project types with
plausible environmental integrity was considerably larger under Track 2 than under Track 1.
Figure ES-3: Environmental integrity of project types each track, by ERUs issued
Source: Evaluation of the largest six projects applied to the portfolio of projects registered under Track 1 and Track 2 as of
March 2014.
Accredited Independent Entities
Accredited Independent Entities (AIEs) have the key role of ensuring the compliance of the
projects with JI requirements, including those related to environmental integrity. In many
cases, they did not perform their auditing functions appropriately. Under Track 1, they had
no incentives to do so, as appropriate oversight was not provided, and any non-performance
had no consequences.
AIEs have the key role of ensuring the compliance of the projects with JI requirements,
including those related to environmental integrity. The findings of the study call into question
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
9
the ability of and incentives for the involved AIEs to perform their auditing functions
appropriately. AIEs often failed to identify obvious mistakes, inconsistencies, questionable
assumptions or claims, or changes to the project activity or monitoring plan. In many
instances, validation and verification expert statements are very brief and do not specify how
key requirements are assessed.
Most host countries rely on AIEs accredited under Track 2 to perform determination and
verification functions under Track 1, but their performance is only monitored and assessed by
the JISC for functions performed under Track 2. For this reason, AIEs did not have to fear
sanctions if they did not perform their functions appropriately under Track 1. Moreover, the
fact that JI project participants select and pay their AIE may create an inherent conflict of
interest.
Bureau Veritas Certification Holding SAS audited by far the most JI projects. It performed
project determinations and/or emission reduction verifications of 56% of all projects, which
generated 78% of total ERUs. The popularity of Bureau Veritas increased towards the end of
the commitment period, while the market share of other AIEs that were popular in the
beginning went down. In our random sample, 77% of the projects determined by Bureau
Veritas made additionality claims that were not plausible, and 17% had questionable claims,
while only 12% of projects determined by other AIEs made implausible additionality claims,
and 46% made questionable claims.
Differences among host countries
The four countries with the highest ERU issuances Ukraine, Russia, Poland and Germany
registered 439 projects and issued more than 800 million ERUs, accounting for 94% of ERU
issuance. An assessment of the project portfolio in each country indicates significant
environmental integrity concerns for more than 80% of ERUs from Russia and Ukraine,
whereas the environmental integrity was rated as high for 70% of ERUs in Poland and 97% in
Germany.
Lessons learned for the design of crediting mechanisms
A key finding of our analysis is that crediting mechanisms need to be very carefully designed
to ensure environmental integrity. In particular, our evaluation of the environmental integrity
of JI offers the following insights:
Crediting mechanisms should adopt project cycle procedures which ensure full
transparency and make all documentation publicly available. Lack of transparency is
an important concern in some JI host countries, where key project documentation, such as
project design documents (PDDs), monitoring reports, and determination and verification
reports are not available or incomplete for a number of projects. To avoid this problem,
crediting mechanisms need rules and enforcement to ensure timely and complete
reporting. However, it is important to note that transparency, though crucial for ensuring
environmental integrity, is not enough by itself. One host country, Ukraine, ensured a
high degree of transparency but nevertheless issued mostly ERUs of very questionable
environmental integrity.
Only internationally accepted methodologies should be eligible for use: Many projects
applied their own, JI-specific approaches for additionality demonstration and the
calculation of emission reductions. In many cases, these projects used inappropriate
approaches, made unrealistic assumptions, or applied questionable values for key
parameters, often leading to overcrediting and significantly higher emission reductions
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
10
estimates than if, for example, Clean Development Mechanism (CDM) methodologies had
been applied. We therefore recommend that only internationally accepted methodologies
that have undergone thorough review by experts and which were developed for specific
and defined project types be used, and that any deviations from such methodologies,
before or after registration, be assessed using appropriate regulatory oversight.
Auditors should be fully accountable for all their activities to the authority
regulating the mechanism: We recommend that crediting mechanisms adopt
accreditation systems which continuously monitor the performance of auditors and which
apply sanctions in the case of non-performance, including the suspension or withdrawal
of accreditation. Merging the JI and CDM accreditation systems could further improve
the oversight of the operations of AIEs.
Retroactive crediting should not be allowed: Retroactive crediting of emission
reductions has seriously undermined the integrity of JI. We recommend that current and
future crediting mechanisms avoid any retroactive crediting and provide for procedures
which ensure that projects must be approved or pre-approved (e.g. through a letter of
endorsement) prior to the decision to proceeding with their implementation.
Investors should have reasonable certainty: In several JI host countries, project
developers faced considerable uncertainty as to whether their projects would ultimately
be approved and ERUs issued. This uncertain environment may have favoured projects
that did not rely on ERU revenues, thereby also negatively affecting the overall
environmental integrity of the project portfolio. We recommend establishing a stable and
predictable regulatory environment for crediting mechanisms.
Implications for the reform and future of JI
The ongoing review of the JI Guidelines offers an important opportunity to address the
shortcomings identified with regard to the environmental integrity of JI. The planned merger
of the two tracks could potentially strengthen the environmental integrity through more
international oversight.
However, key issues which could continue to strongly undermine the integrity of JI have yet
to be addressed. For example, the current draft JI Guidelines allow existing projects to
continue issuing ERUs under the new guidelines without having their additionality
reassessed. The threat that existing projects with low environmental integrity would continue
undermining mitigation outcomes is real and significant and should therefore be weighed
strongly. JI projects only had certainty about ERUs up to the end of the first commitment
period. We therefore recommend that only existing projects which would stop operating
without the incentives from JI should be able to continue receiving ERUs in the second
commitment period.
The draft JI Guidelines furthermore propose different options for the length of crediting
periods and their renewal. The issuance of ERUs is tied to the availability of AAUs from the
same commitment period when the emission reductions occurred. It is unlikely that the Kyoto
Protocol will be extended by a third commitment period, and JI may therefore not continue as
a mechanism beyond 2020. Allowing for crediting periods that last beyond the end of 2020
could leave investors facing considerable uncertainty as to whether and how any reductions
beyond 2020 could be credited. We therefore recommend that the revised JI Guidelines
explicitly state that the crediting period ends at the end of the second commitment period.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
11
Implications for market mechanisms under a new climate agreement
Countries are negotiating a new agreement under the United Nations Framework Convention
on Climate Change (UNFCCC), to be approved at the Paris Climate Change Conference in
December 2015 and become effective in 2021. As part of this process, each country is to
submit its “intended nationally determined contribution” (INDC) under the agreement.
Countries have also started to discuss how carbon markets could be incorporated into this new
climate agreement.
Our evaluation clearly shows that oversight of an international market mechanism by the host
country alone is insufficient to ensure environmental integrity, in particular for countries with
a significant AAU surplus which had no incentives to ensure environmental integrity. A new
regime could create situations similar to those experienced with JI:
Unclear ambition of INDCs: If INDCs are set above business-as-usual (BAU)
emissions, host countries would not have incentives to ensure environmental integrity of
units transferred internationally. A key challenge is that future (emission) developments
are rather uncertain. An INDC which appears ambitious from today’s perspective could
turn out to be easily achievable due to unforeseen developments, such as an economic
recession, new low carbon fuels becoming available, or technological developments.
Absence of international accounting rules: If countries with mitigation commitments
do not account for units transferred to other jurisdictions, they could sell credits without
having to engage in additional mitigation action if these credits lack environmental
integrity. An internationally agreed accounting approach is crucial to ensuring that
international transfer of units does not lead to double counting of emission reductions,
and that host countries have incentives to ensure environmental integrity of units.
INDCs not converted into multi-year emission targets: If countries with single-year
targets (e.g. a 20% reduction by 2025) did not convert their target into a multi-year
emissions target and were allowed to transfer units issued for years up to the target year,
they would have no incentive to ensure the units’ environmental integrity. Accounting of
international transfer of units towards commitments should indeed only be possible if
countries take on quantifiable, multi-year emission reduction targets.
Ensuring that INDCs are ambitious and converted into multi-year emission budgets, and
agreeing on international accounting rules, is therefore critical to avoid that the experience
with JI is repeated under a new climate regime.
While international oversight is important, we also note its limitations. Information
asymmetry between proponents and auditors or regulators remains a major challenge that is
difficult to address, even with international oversight. Furthermore, international oversight
can only be effective if countries do not hamper the strengthening of a crediting mechanism’s
integrity. CDM and JI reform efforts under the UNFCCC have shown that often no consensus
can be reached to address and rectify environmental integrity shortcomings.
A broader question that remains is what the scope and role of crediting mechanisms can be in
the longer term. The experience with JI shows that in countries with ambitious caps, the
potential of a crediting mechanism may be quite limited. The EU, for example, had to limit
the eligible project types considerably in order to avoid double counting and overlap with the
EU ETS and other climate policies. Given the reduced potential of crediting mechanisms in a
world where most emissions are covered under other mitigation policies, and given the
general challenges of ensuring environmental integrity for crediting mechanisms, the role of
crediting mechanisms beyond 2020 may be rather limited.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
12
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
13
1. INTRODUCTION
1.1 Purpose of this study
Joint Implementation (JI) is one of the two offsetting mechanisms under the Kyoto Protocol,
together with the Clean Development Mechanism (CDM). It enables countries with
mitigation commitments under the Kyoto Protocol, called Annex B countries, to generate and
transfer Emission Reduction Units (ERUs) from greenhouse gas (GHG) abatement projects.
The CDM has been widely studied, but few analyses have been done for JI (Shishlov et al.
2012; Ukraine 2006a; Armenteros and Michaelowa 2003; Gaast 2002; Jackson 1995; Schmitz
and Michaelowa 2005; Sterk 2008). In particular, an in-depth evaluation of its environmental
integrity has not been done to date. This is a significant knowledge gap, given that the almost
872 million issued ERUs account for one-third of all Kyoto offset credits (from both CDM
and JI) that had been issued as of March 2015.
1
This study systematically evaluates the environmental integrity of JI in the first commitment
period of the Kyoto Protocol (KP). It aims to answer the question: To what extent would
emissions have been higher or lower without JI, keeping everything else constant? We assess
how the current rules and governance arrangements have affected the environmental integrity
of JI and what lessons can be learned for both the reform of JI and new market mechanisms.
Our analysis is directly relevant to ongoing carbon market discussions. First of all, JI rules are
being revised; the current JI Guidelines are to be replaced by Modalities and Procedures for
JI”, which are under consideration by the Subsidiary Body for Implementation (UNFCCC
2006b; UNFCCC 2014c). The study makes recommendations on how JI could be reformed.
Second, lessons learned from JI can provide vital insights for the design of future market
mechanisms and accounting rules. Many countries have expressed support for continuing to
use market mechanisms under the United Nations Framework Convention on Climate Change
(UNFCCC), including the existing mechanisms CDM and JI, as well as a “new market
mechanism” and a “framework for various approaches (UNFCCC 2013b; UNFCCC 2012).
It is unclear whether or how carbon markets will be part of a post-2020 climate deal that
countries aim to agree on at the Paris Climate Change Conference this December. As JI was
implemented in countries with economy-wide mitigation targets, it can provide important
lessons for market mechanisms under a new climate agreement, as well as for domestic
mechanisms in countries with mitigation targets.
1.2 Research approach
The environmental integrity of an offsetting mechanism depends on both the environmental
integrity of the projects for which units are issued and the way in which units are accounted
i.e. how they are issued, transferred and used towards meeting mitigation commitments (Prag,
Hood, et al. 2011; Prag, Aasrud, et al. 2011; Prag et al. 2012; Schneider et al. 2015). Because
JI projects are hosted by countries with economy-wide emission reduction targets, they differ
from CDM projects in an important way. Using units from CDM projects that lack
environmental integrity to help meet a mitigation pledge will always lead to an increase in
global emissions, but this is not necessarily the case under JI. As we discuss further in Section
3, the global emissions impact of using ERUs from JI projects that lack environmental
1
See: http://ji.unfccc.int/statistics/2015/ERU_Issuance_2015_03_31_1200.pdf [accessed 14 May 2015].
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
14
integrity depends to a large extent on whether the host country has a significant surplus of
Kyoto allowances.
The environmental integrity of an offset project depends on whether the project is additional
and the emission reductions are not overestimated.
Additionality refers to the question whether a project is implemented due to the incentives
from JI. If the project would also have been implemented in the absence of JI, it is not
considered additional. All the ERUs issued from such a project would be based on emissions
reductions that would also have occurred in the absence of JI.
To ensure that emission reductions are not overestimated and ERUs are not overcredited,
several factors must be considered:
1. Baseline emissions: a baseline represents the assumed emissions level that would occur if
the JI project had not been implemented. Because baseline emissions are based on a
counterfactual scenario and assumptions, they are associated with considerable
uncertainty. In order to ensure environmental integrity, baselines need to be based on
robust data and a credible business-as-usual scenario, and then be quantified
conservatively in order to avoid overcrediting.
2. Whether project emissions are quantified in an accurate or conservative way.
3. The way how potential leakage effects, i.e. indirect emission effects outside the project
boundary, are addressed.
4. Whether the length of the crediting period is appropriate.
5. Whether the emission reductions are measurable and verifiable i.e. whether the
achieved emissions reductions can be measured, quantified, and independently verified
with reasonable accuracy.
6. Whether the emission reductions are attributable to the mitigation activity. For some
project types, this is easy to determine, e.g. when the GHGs in a tail gas stream can be
directly measured before they enter a GHG destruction facility. However, for some
project types it can be difficult or impossible to distinguish the emissions impact of the
project from other changing parameters that affect emissions. This can be the case when
emissions are influenced by several factors (such as fuel prices and availability) that make
it difficult to attribute reductions to the incentives created by the offset programme.
7. Whether the reductions are permanent, or whether the potential non-permanence of
emissions reductions is addressed in an appropriate manner.
The study primarily focuses on the assessment of additionality and the determination of
baseline emissions, which particularly affect environmental integrity, but we also discuss
other issues as relevant to specific project types or regions. Given the uncertainties noted
above, and the limited availability of data, we cannot precisely determine whether projects are
additional or the degree of any overcrediting or undercrediting. Instead, we discuss the
likelihood of additionality and the range of overcrediting or undercrediting for different types
of JI projects and countries.
Our analysis is based on an array of evidence: we evaluated relevant literature; interviewed
stakeholders involved in JI; analysed the circumstances of particular project types through
economic, sectoral, and policy evaluations, assessing differences between regions; and
evaluated publicly available project documents. The latter was a key element of the study; we
evaluated a random sample of 60 JI projects, drawn from all projects registered as of 1 April
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
15
2014 based on information provided by UNEP Risoe (2014). The UNEP Risoe database
includes 604 registered projects (both under Track 1 and Track 2) and 38 registered
Programmes of Activities (PoAs).
2
Rejected or withdrawn projects and projects that did not
proceed to registration (i.e. projects at earlier stages of the JI project cycle) were not included.
The sample was chosen in a representative manner based on the following criteria:
a. Region/Host country: To identify potential differences between host countries,
the projects were clustered into four categories: Russia, Ukraine, other Annex I
countries with economies in transition, and Annex II countries.
3
b. Track type: To identify differences between tracks we used three categories:
Track 1 projects, Track 2 projects, and projects that were initiated under Track 2
but later switched to Track 1. (See Section 2.2 for a definition of the tracks.)
c. Project types: To identify differences in technologies and project types, we
clustered the projects into 12 categories (see Table 1).
We drew the sample in a way that the three categories above are representative with regard to
the projects’ estimated emission reductions in the first commitment period (CP1), until the
end of 2012, as stated in the project design documents (PDDs). We used the estimated
emission reductions and not the number of projects because we aim to assess the overall
average integrity of ERUs which have or could be issued for CP1. We used the number of
expected, not issued, ERUs in order to include projects which were registered but had not yet
issued ERUs. A list of the projects in the sample is provided in Annex 2.
Table 1: Categorization of project types used to establish a random sample
4
Project types
Registered
projects
% of ERUs
issued in CP1
% of ERUs expected in
CP1 (according to PDD)
Source: UNEP Risoe (2014)
2
The UNEP Risoe JI database does not include all registered projects. The JI section of the UNFCCC website
(http://ji.unfccc.int) shows, as of 31 August 2014, 597 registered Track 1 projects and 51 projects with final
determination under Track 2, a total of 648 projects. We used the UNEP Risoe database for sampling because it
provides more data in tabular format that can be used for analysis. There are also JI projects with no ITL number,
because they did not pay the registration fee, that are still listed on the UNFCCC website and marked as “registered
in the UNEP Risoe database. These projects were also included in the database from which we drew the sample.
3
Annex II Parties under the UNFCCC consist of the Organisation for Economic Co-operation and Development
(OECD) members of Annex I, but not the Economies in Transition (EIT) Parties.
4
We used project types as defined in the UNEP Risoe database. Though the type definitions are not always precise,
we used this database with its categorization because it enables the analysis of the whole array of JI projects.
Spontaneous ignition of coal waste piles
78
26%
26%
Energy efficiency supply side
117
16%
19%
Associated petroleum gas utilization
22
14%
11%
Natural gas transportation and distribution
32
10%
9%
Other: non-CO2
167
8%
8%
Energy distribution
47
7%
7%
HFC-23 abatement from HCFC-22
3
5%
4%
N2O abatement from nitric acid
41
5%
5%
PFC and SF6 reduction
7
4%
3%
Fossil fuel switch
17
3%
3%
Other: CO2
81
2%
3%
Coal mine methane
28
1%
4%
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
16
The two project type categories Other: non-CO2 and Other: CO2 include a number of
project types with relatively small emission reductions that were combined into these two
categories (see Table 2).
Table 2: Other: CO2 and Other: non-CO2project type categories
Project type
Number of registered
projects
% of ERUs issued
in CP1
% of ERUs
expected in CP1
Other CO2
Total
167
7.9%
7.8%
Other non-CO2
Total
81
2.3%
2.5%
Source: UNEP Risoe (2014)
The evaluation of the sample projects focuses on an analysis of the PDDs; in addition, other
project documentation, such as monitoring reports, determination and verification reports,
letters of approval, and ERU approval and issuance data were reviewed where available. The
analysis is used to evaluate the environmental integrity of the projects and to assess JI rules
that affect environmental integrity. Where possible, quantitative results are identified;
otherwise qualitative approaches are used to estimate impacts.
We also examine six project types in more detail, in order to compare their environmental
integrity. The project types were selected by their share of ERUs; together they cover 82% of
ERUs issued and 53% of the projects registered. The project type assessment includes
economic aspects and sector-specific information, drawing upon the sample of JI projects,
relevant literature, and structured interviews with market participants.
Agriculture: no tillage
7
2.5%
1.9%
Biomass energy
46
0.8%
1.2%
Hydro
20
0.7%
1.0%
Energy efficiency in households
12
0.3%
0.9%
Energy efficiency in service sector
23
1.7%
0.8%
Wind
43
0.7%
0.8%
Cement
4
0.5%
0.4%
Transport
4
0.0%
0.4%
Afforestation
2
0.5%
0.2%
Geothermal
5
0.1%
0.2%
Avoided deforestation
1
0.1%
0.0%
N2O adipic acid
3
1.8%
1.4%
Landfill gas
67
0.3%
0.7%
Methane avoidance
7
0.2%
0.3%
Agriculture
4
0.0%
0.1%
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
17
2. JOINT IMPLEMENTATION: AN OVERVIEW
Joint Implementation is one of three flexibility mechanisms established under the Kyoto
Protocol to enable countries with binding GHG emissions targets to meet their obligations
more cost effectively. JI is described in Article 6 of the KP (UNFCCC 1997). Under JI, one
Annex I country can invest in an emission reduction activity in another Annex I country as an
alternative to reducing emissions domestically.
The first host country approval for a JI project was given in 2000, to a Polish project. The first
ERUs were issued in 2008. As shown in Figure 1, issuance of credits grew exponentially in
the first few years before plateauing in 2013. Prices dropped from more than EUR 12 in 2010
to less than EUR 0.10 in early 2013, and have remained at well below EUR 0.50. This
precipitous drop occurred because the supply of ERUs from JI and Certified Emission
Reductions (CERs) from the CDM exceeded demand. The demand was limited due to various
factors, including a cap on the use of ERUs and CERs in the EU Emissions Trading System
(EU ETS), and less demand from EU governments for mitigation commitments in non-ETS
sectors due to the economic slowdown. On the other hand, the supply of ERUs was much
larger than expected, in particular when ERU issuance accelerated strongly in 2012.
Figure 1: Accumulated issuance and ERU prices
Data sources: UNEP Risoe (2014) and http://www.quandl.com/futures/ice-ecx-eru-emission-futures.
As of August 2014, there were 648 registered JI projects,
5
of which 496 have generated ERUs
almost 857 million in total. Ukraine and Russia account for more than 90% of ERUs issued.
They also have the largest share of projects. Poland, Germany, France and Romania account
for another 7% of ERUs (see Section 6). There are also significant differences by project
type: some types, such as projects related to the ignition of coal waste piles, have generated
large numbers of ERUs, while others types, such as wind power, have many registered
projects but have issued far fewer ERUs (see Tables 1 and 2 in the previous section).
2.1 JI governance structures
Under the current JI Guidelines, projects can be implemented under two different tracks
(UNFCCC 2006b). Under Track 1, host countries are responsible for all aspects of the project
cycle, including the approaches for determination of additionality, quantification and
verification of emission reductions, registration of projects, and issuance of ERUs. Hence,
5
See: http://ji.unfccc.int/JI_Projects/ProjectInfo.html [accessed 17 October 2014].
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
18
under Track 1 each Party establishes its own procedures and requirements for projects and
verification and has the responsibility to ensure the environmental integrity of the resulting
credits. While some requirements differ between Parties, there are also many common
elements. Under Track 2, some of these responsibilities are assigned to a UN body, the Joint
Implementation Supervisory Committee (JISC), which has requirements and procedures that
resemble, to a large extent, those of the CDM.
To qualify as a Track 1 host country, a Party has to meet the eligibility requirements listed in
paragraph 21 of the JI Guidelines, which are related to the Party’s ability to account for its
emissions and carbon units while ensuring transparency (see Table 3). If a host country does
not meet any of these requirements, JI projects have to be implemented under Track 2. Track
2 was intended to provide flexibility to Parties and project participants in case a Party is
unable to meet Track 1 requirements. It was feared that compliance with the requirements
related to the national GHG inventory and inventory system could be a big challenge for some
countries. However, in practice the large majority of Parties have been continuously eligible
to participate in both tracks. Parties that meet the Track 1 eligibility criteria may still elect to
apply the Track 2 verification process.
Table 3: Requirements for countries' participation in JI and ERU transfer
Minimum requirements
(common to Track 1 and Track 2)
Additional requirements for Track 1
Source: UNFCCC (2006b)
JI Track 2 became operational in 2006, after the JISC was established. Track 1 took longer to
start, since host countries had to be in compliance with Track 1 requirements and needed time
to establish their national JI procedures and infrastructure. Many host countries used Track 2
as a model when they designed their national Track 1 rules.
Under both tracks a JI project has to be approved by the Parties involved, i.e. by the host
country and at least one investor country (UNFCCC 1997). Many host countries implemented
a two-step approval procedure: An initial project endorsement for which a basic description of
the project idea is usually sufficient and which results in a Letter of Endorsement (LoE). This
is followed by final project approval after the project design document (PDD) has been
audited by a third-party auditor. Final project approval is expressed in a Letter of Approval
(LoA) issued by both the host and the investor Parties. The host country LoA has to be issued
before project registration, while an investor Party LoA can be obtained prior to the first ERU
transfer at the latest.
The governance arrangements of Track 2 are defined in the JI guidelines (UNFCCC 2006b).
Under Track 2 the JISC sets rules and procedures, oversees the project cycle and performs the
accreditation and supervision of the third-party auditors, called Accredited Independent
Entities (AIE). The latter perform the determination of PDDs and the verification of the
emission reductions claimed by the project. The JISC may request a review both before a
project is registered and each time before ERUs are issued. Three members of the JISC need
to request such a review, otherwise the project determination is automatically deemed final
(i.e. the project gets registered) 45 days after the AIE has submitted the determination report.
1. Being a Party to the Kyoto Protocol
2. Having its assigned amount established for
the relevant commitment period
3. Having in place a national GHG registry
(carbon units registry)
4. Having in place a national GHG inventory
system
5. Having submitted the most recent annual GHG
inventory
6. Having submitted required supplementary
information on its Assigned Amount
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
19
Likewise, verification of emissions reductions becomes final 15 days after its submission by
an AIE, unless the JISC requests a review. The JISC reviewed nine projects out of 52 Track 2
determinations and rejected one of them. All 129 verifications were accepted by the JISC, and
only one of them was reviewed.
6
Under Track 1 the host country is solely responsible for project registration and verification
of emission reductions. The main elements of national procedures of several selected host
countries are further discussed in Section 6. Many host countries use AIEs accredited by the
JISC in their national procedures to perform PDD determination and emission reduction
verification under Track 1. Once a JI Track 1 project is registered by a host country, the
project information is communicated to the UNFCCC Secretariat and the project is assigned a
unique number by the International Transaction Log. Under both tracks, the issuance of ERUs
is performed by the host country, since ERUs are obtained by the way of conversion of the
party’s AAUs or RMUs in its national registry.
Table 4: Bodies involved in JI Track 2 project procedure and their functions
Function / step in the project cycle
Responsible body
Despite the fact that Track 2 was operational before Track 1, almost 90% of JI projects have
been registered and 97% of ERUs have been issued under Track 1 (see Table 5). Many
projects that were initiated under Track 2 switched to Track 1 once it became operational.
Table 5: ERUs and registered projects by JI track
Source: http://ji.unfccc.int.
2.2 JI rules and procedures on environmental integrity
This section gives a brief overview of the most pertinent JI rules on environmental integrity.
As noted in the introduction, the JI rules are being revised, and the current JI Guidelines are to
be replaced by Modalities and Procedures for JI, which are under consideration by the
Subsidiary Body for Implementation (SBI). Here we focus on the current rules, but the
6
See http://ji.unfccc.int/JI_Projects/DeterAndVerif/index.html [accessed 22 October 2014].
Project endorsement (usually required)
Host Party’s Designated Focal Point (DFP)
PDD determination
Accredited Independent Entity
Project approval
DFPs of host and investor Parties
Final determination (registration)
JI Supervisory Committee
Verification of emission reductions
Accredited Independent Entity
Final verification
JI Supervisory Committee
Issuance of ERUs
Host Party
Track
Number of
registered
projects
Share of
registered
projects
Total million
ERUs issued
Share of
ERUs
issued
ERUs issued per track
Track 1
597
92%
832
97%
Track 2
51
8%
25
3%
Total
648
857
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
20
December 2014 draft of the new Modalities and Procedures for JI is discussed in more
detail in Annex 1.
We start with the rules as stated in the KP, followed by the JI Guidelines, which apply to both
Tracks 1 and 2, and specific rules on Track 2 issued by the JISC.
Article 6, paragraph 1b, of the Kyoto Protocol requires that all JI projects have to be
additional: Any such project provides a reduction in emissions by sources, or an
enhancement of removals by sinks, that is additional to any that would otherwise occur
(UNFCCC 1997).
The JI Guidelines reiterate the requirement of additionality for each track. Under Track 1 it is
the sole responsibility of the host country to ensure additionality: “… a host Party may verify
reductions in anthropogenic emissions by sources or enhancements of anthropogenic
removals by sinks from an Article 6 project as being additional to any that would otherwise
occur …” (UNFCCC 2006b; Annex, paragraph 23).
The JI Guidelines do not provide any further specific requirements to Track 1 projects except
that a Track 1 host party is obliged to communicate to the UNFCCC Secretariat its national
guidelines and procedures for approving Article 6 projects, including the consideration of
stakeholders’ comments, as well as monitoring and verification (UNFCCC 2006b; Annex,
paragraph 20b). Track 1 procedures relating to environmental integrity of several host
countries are considered in more detail in Section 6.
Under Track 2 an AIE has to check whether the project would result in a reduction of
anthropogenic emissions by sources or an enhancement of anthropogenic removals by sinks
that is additional to any that would otherwise occur (UNFCCC 2006b; Annex, paragraph
33b) and whether the project design document (PDD) includes an appropriate baseline and
monitoring plan (UNFCCC 2006b; Annex, paragraphs 31c and 33c).
Proper baseline-setting is important because it affects the number of emission reductions
being credited. The baseline scenario is defined in the JI Guidelines as the scenario that
reasonably represents the anthropogenic emissions by sources or anthropogenic removals by
sinks of greenhouse gases that would occur in the absence of the proposed project
(UNFCCC 2006b, Annex, Appendix B, paragraph 1). The JI Guidelines further define that the
baseline shall be established on a project-specific basis and/or using a multi-project emission
factor (UNFCCC 2006b, Annex, Appendix B, paragraph 1). The JI Guidelines further
require the baselines to be established in a transparent manner based on conservative
assumptions, taking into account uncertainties, relevant national and/or sectoral policies,
circumstances and developments (UNFCCC 2006b, Annex, Appendix B, paragraph 1).
The JISC further developed more specific Track 2 rules on additionality, baseline setting,
monitoring, and accreditation of auditors. The most important documents are the Guidance
on Criteria for Baseline Setting and Monitoring (the initial version was adopted in 2006
(UNFCCC 2006a), Version 03 was adopted by JISC 26 in 2011(UNFCCC 2011b); the Joint
Implementation determination and verification manual (UNFCCC JISC 2009); and the Joint
Implementation Accreditation Standard (UNFCCC 2010).
According to the Guidance on Criteria for Baseline Setting and Monitoring, project
participants can select one of the following approaches for baseline setting and monitoring
(UNFCCC 2011b):
a) Develop their own methodology in accordance with the JI Guidelines: A JI specific
approach can build on selected elements of CDM methodologies or elements of
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
21
approaches already taken in comparable JI projects. Under this approach a baseline
shall be identified by listing and describing plausible scenarios on the basis of
conservative assumptions and selecting the most plausible one.
b) Use a methodology approved by the CDM Executive Board: When a CDM
methodology is used, its most recent version has to be applied in its totality and
strictly followed.
c) Use an approach already taken in a comparable registered JI Track 2 project.
7
The additionality of the project can be demonstrated by using one of the following approaches
(other approaches can be proposed too):
a) Use a project-specific approach: Demonstrate that the project scenario is not part of
the conservatively identified baseline scenario, and that the project will reduce
emissions below the baseline by providing relevant traceable and transparent
information.
b) Refer to a comparable project: Demonstrate the same approach for additionality proof
which already has been taken in a comparable registered JI Track 2 project. In this
case the project participants do not need to provide project-specific additionality
demonstration (UNFCCC 2008).
c) Apply CDM Additionality Tool
8
in its most recent version.
The JISC rules also specify how project boundaries have to be set and how leakage should be
addressed. The project boundary has to be clearly defined in the PDD and encompass all
significant GHG emissions that are reasonably attributable to the project and under control of
the project participants. Leakage has to be taken into account when it is projected to exceed
1% of the project emission reductions or 2,000 tonnes of CO2 equivalent (CO2e), whichever is
lower (UNFCCC JISC 2009; paragraph 32).
Under JI rules, projects may have begun construction or implementation before the start of
CP1 as early as 2000 but ERUs are only granted for reductions occurring during the
commitment period (UNFCCC 2006b). The crediting period can start as soon as the project
begins generating emission reductions, but not earlier than the beginning of 2008, since ERUs
are sourced from AAUs that are tied to the commitment period of the Kyoto Protocol.
9
Unlike
under the CDM, ERUs can be issued retroactively for the period from 2008 before project
registration (so-called retroactive crediting).
The JISC rules allow the crediting period to extend beyond 2012, subject to host country
approval. However, it is noted: The status of emission reductions or enhancements of net
removals generated by JI projects after the end of the first commitment period may be
determined by any relevant agreement under the UNFCCC (UNFCCC 2009, paragraph 19).
The issuance of ERUs for emissions reductions that occurred after 2012 will only be possible
once countries have AAUs for the second commitment period; see Annex 1.12.
7
A project can be considered comparable if it uses the same emission reduction technology in the same host
country, the time span between the starting dates of the projects is less than five years and the relevant regulatory
framework has not changed over this time, and the difference in scale of activities does not exceed 50%
8
See: “Tool for the demonstration and assessment of additionalityapproved by the CDM Executive Board
https://cdm.unfccc.int/methodologies/PAmethodologies/tools/am-tool-01-v5.2.pdf/history_view [accessed 17
September 2014].
9
Some countries – Bulgaria, Czech Republic, Poland, Romania and Ukraine – established schemes which granted
AAUs to project developers for reductions prior to 1 January 2008, with the view to incentivizing early action and
enhancing the attractiveness of JI (Shishlov et al. 2012).
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
22
3. WHEN DOES THE ENVIRONMENTAL INTEGRITY OF JI PROJECTS MATTER?
In principle, offsets are a zero-sum game for the atmosphere. The purchase of offset credits
allows the buyer country to increase its own emissions by a corresponding amount above its
target level, while emissions are reduced by that amount in the host country. The net result is
the same global emissions as would occur in the absence of offsets. But this only holds true
under a number of conditions, which we discuss below.
JI differs from an offsetting mechanism such as the CDM, as JI projects are located in
countries with economy-wide emission reduction targets under the KP (Annex B countries). If
ERUs from JI projects that lack environmental integrity are used for compliance they can but
do not necessarily have to lead to an increase in global GHG emissions. To avoid double
counting, a host country has to give up a Kyoto emissions permit called Assigned Amount
Units (AAUs) or Removal Units (RMUs) obtained for carbon sequestration activities for
each ERU it issues.
10
In this way the overall emissions budget of Annex B countries remains
unaffected by the issuance of ERUs and the emissions reductions achieved through a JI
project are only counted once. If the JI host country did not convert a Kyoto emission permit
an AAU or RMU to issue ERUs, the same emission reduction would be reflected in both
an ERU issued and the host country’s emissions inventory (which is used as the basis for
accounting to fulfil its own obligations under the KP).
However, whether overall emissions from these countries are affected or not, depends on
several circumstances and assumptions. Below we explore the circumstances and assumptions
that determine the answer to the question: Would global GHG emissions be higher, lower, or
the same in the absence of JI, keeping everything else constant?
3.1 Is the JI project additional, and are emission reductions correctly credited?
Whether a JI project is additional, and whether an additional project is undercredited,
overcredited or correctly credited can have different impacts on global GHG emissions.
In principle, the environmental integrity of JI projects does not influence total global GHG
emissions. Consider a hypothetical example: A country has an emissions budget of 10,000
AAUs and would emit 12,000 tonnes without a mitigation target (business-as-usual
emissions). It therefore has to either reduce domestic emissions by 2,000 tonnes, or purchase
units from another country. JI projects are implemented in that country. The country converts
1,000 of its AAUs and issue 1,000 ERUs to those projects. It now has 9,000 AAUs remaining.
The 1,000 ERUs are used for compliance by another KP country with a reduction
commitment. We consider two cases:
a) The JI projects are additional: The projects have actually reduced emissions by
1,000 from BAU, to 11,000 tonnes. The country has 9,000 remaining AAUs. To meet
its target, the country has to engage in mitigation actions (or buy units) to reduce its
emissions by 2,000 tonnes just as if the projects had not occurred.
b) The JI projects are not additional: In this case, the countrys BAU emissions
would remain at 12,000 tonnes. The country has 9,000 remaining AAUs. To meet its
target, the country has to engage in mitigation actions (or buy units) to reduce its
emissions by 3,000 tonnes 1,000 more than if the JI projects had not occurred.
10
Under the Kyoto Protocol, each country with a reduction obligation receives an emissions budget (initial
assigned amount) that is calculated the following way: baseline emissions x years in the commitment period x
reduction target expressed as fraction of baseline emissions (e.g. 80% of 1990 emissions by 2020).
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
23
Thus, although the JI projects were not additional, the issuance of the 1,000 ERUs
(and their use by another country to meet KP obligations) will be compensated for by
the host country’s further reductions.
In both cases above, global GHG emissions are the same with or without JI. If a country
issues ERUs from non-additional or overcredited projects, it will lose AAUs without
achieving emissions reductions and it will have to engage in more mitigation action to
compensate for the lost AAUs.
3.2 Does the host country have a surplus of AAUs?
The above considerations do not hold true if a country has a (significant) surplus of AAUs.
11
Such an oversupply can build up if the target is unambitious i.e. higher than the projected
business-as-usual (BAU) emissions. Oversupply can also be created if the country engages in
more mitigation than what is required by its target. If a target is ambitious, however, such
overachievement will be transient and will not lead to a build-up of surplus. In a cap-and-
trade system that has a long-term oversupply problem, on the other hand, cancelling of
allowances would not lead to immediate additional mitigation by covered sources, because it
would simply remove some of the built-up surplus.
If the country has sufficient surplus, it can issue ERUs to non-additional or overcredited JI
projects without having to engage in more mitigation action, because it will still have enough
AAUs to cover its emissions. If ERUs from non-additional or overcredited JI projects in a
country with an oversupply are used for compliance by the buyer country, global emissions
would therefore increase. Section 3.4 explains the impacts under different circumstances and
assumptions, Section 9 estimates the impact on overall global emissions from JI.
For the purpose of our analysis, we consider a country to have a significant expected surplus if
its average emissions in the five years before the start of CP1 (20032007) were more than
20% below its CP1 reduction target. By this definition, more than 95% of ERUs issued up to
March 2014 came from countries with a significant expected surplus. Among the four JI host
countries with the largest shares of ERUs, only Germany, accounting for 1.6% of total ERUs
issued, did not have significant expected surplus (see Section 6 for country-specific analysis).
Another important consideration for our analysis is whether surplus AAUs from CP1 could
be used or sold at a later date. Removing allowances from an oversupplied market may lead
to additional emission reductions later in time, assuming the surplus was temporary and that
allowances will be in short supply at a later date. If surplus AAUs could be used or sold at a
later date, a country may want to hold on to them instead of issuing ERUs from non-
additional JI projects now. Second, if all the surplus is used at a later date, i.e. in subsequent
commitment periods, then non-additional or overcredited JI projects would not lead to higher
global GHG emissions than in a scenario without JI where all of the AAU surplus is used
over time. But as we discuss below, we do not see this as a realistic assumption.
The Kyoto Protocol allows for full carry-over of AAUs from one commitment period to the
next (UNFCCC 1997; Article 3, paragraph 13). This carry-over provision is meant to
encourage and reward early action mitigation activity that goes beyond the committed
target. But this carry-over provision also enables surplus AAUs to be carried over into the
11
Countries with an emission reduction target inscribed in Annex B of the KP issue AAUs for each commitment
period corresponding to their emissions budget over that period. In addition, countries may issue RMUs – or may
have to cancel AAUs – to account for land use, land-use change, and forestry activities under Articles 3.3 and 3.4
of the KP.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
24
next commitment period. As we show in Section 6, most of the 13 billion AAU surplus that
accumulated in CP1 is due to weak reduction targets. Policy-makers are well aware of this
problem and have tried for many years to amend the Kyoto rules in a way that would
minimize the use of surplus AAUs in CP2. At the UN Climate Change Conference in Doha in
2012, the Parties adopted a compromise relating to surpluses from CP1 and CP2 (Kollmuss
2013). The decision does not limit the carry-over of surplus AAUs from CP1 but limits their
use in CP2. It also makes it impossible for countries without a reduction target in CP2 to sell
their surplus to countries with a reduction target. Russia, which generated 30% of all ERUs in
CP1, did not join CP2 and cannot sell AAUs to CP2 countries.
The Parties further decided in Doha to restrict the number of CP2 AAUs that a country will
have available for the CP2 by requiring the cancellation of AAUs that exceed the country’s
average emissions in 20082010 in order to avoid the build-up of new surplus. The Doha
decisions would mean that Ukraine, which has generated almost 60% of all ERUs, will not be
able to use its CP1 surplus to meet its CP2 target if it does not strengthen its CP2 target from
76% to 42% of 1990 emissions (Morel and Shishlov 2014). It is therefore unclear if Ukraine
will ratify CP2 and, if it does, if it will change its target (Storchylo 2014).
Over 91% of ERUs were generated in countries that likely cannot use their CP1 surplus in
future commitment periods. It is even less likely that AAU surplus can be used in the post-
2020 agreement that is currently being negotiated. Since AAUs are the currency of the Kyoto
Protocol the view of the majority of Parties seems to be that they should not be used in the
post-2020 agreement. For these reasons, we assume in our analysis that countries with a
significant expected surplus will not be able to use their AAUs in the future.
A last important consideration for the incentives to issue ERUs from non-additional JI
projects is whether the demand and price is higher for ERUs or for AAUs. If countries with
significant AAU surplus could make a larger profit from selling AAUs than selling ERUs
they would not have incentives to issue ERUs from non-additional projects. Of the 13 billion
AAU surplus that accumulated in the first commitment period only about 450 million, less
than 4%, have been sold (UNEP Risoe 2014). Almost twice as many ERUs were sold
almost 857 million ERUs have been transferred as of August 2014. The demand for ERUs
could be larger due to the fact that the European Emission Trading Scheme (EU ETS), the
largest market of emissions units, only allows ERUs to be used for compliance but not AAUs.
Anecdotal evidence also indicates that some countries were hesitant to buy AAUs from
countries which were seen as having a large AAU surplus. This seems to indicate that
overall the demand for ERUs is higher than for AAUs.
3.3 Are the JI emission reductions reflected in the host country’s GHG inventory?
Emission reductions achieved by JI projects are in most cases reflected in the host country’s
GHG inventory. In other words, the reported host country’s GHG emissions will be lowered
by the number of tonnes of emissions reductions achieved through JI. However, this is not
necessarily the case with all project types. Some emissions reductions may not be reflected in
the country-wide GHG inventory, for example, because the country uses simple Tier 1
methods to estimate an emissions source which do not account for the emission reductions
achieved through JI projects or because the reductions occur in a sector that is not covered by
the host country's GHG inventory. For example, a project-based mechanism may issue units
for N2O abatement from nitric acid production. If the country uses a simple Tier 1 method
with default values to estimate N2O emissions from nitric acid production, the emission
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
25
reductions achieved through the project-based mechanism may not be reflected in the GHG
inventory of the country.
12
This issue has been referred to as inventory visibility (Prag et al.
2013). In our analysis we distinguish between both cases. We explore in Section 5 which
project types may not be reflected in the host country’s emissions inventory.
Inventory visibility is important with regard to the impacts of JI projects on global GHG
emissions. Inventory inconsistencies with JI projects may point to potential overcrediting or a
lack of inventory accuracy, as we discuss further in Section 5, where we analyse the most
important project types.
3.4 Impact of environmental integrity of JI projects on global emissions
Tables 6 and 7 below show how the three key assumptions and circumstances of JI projects
discussed above whether JI projects are additional and whether they are undercredited,
correctly credited, or overcredited; whether the JI host country has a significant AAU surplus;
and whether the emission reductions from the JI project are reflected in the GHG inventory
affect global GHG emissions. The tables describe the emissions impact compared with a
situation where JI would not be used, assuming that the ERUs will be used for compliance
under the Kyoto Protocol.
It is important to note that in Table 6, we assume that the country exactly meets its target to
clearly distinguish from the case illustrated in Table 7, where the country has significant
surplus. In reality a country may fall slightly short or have some surplus at the end of its
commitment period.
12
GHG inventories for Annex 1 aim to consistently estimate anthropogenic emission by all sources and removals
by all sinks of all GHGs, as covered by the Revised 1996 IPCC Guidelines for National Greenhouse Gas
Inventories and IPCC good practice guidance, in accordance with relevant decisions of the COP and/or COP/MOP
FCCC/CP/2001/13/Add13, Decision 20.CP.7 Annex 1 Art 9. Inventories use different types of information Tier 1
methods use default values, Tier 2 are similar but with country-specific emission factors and other data, Tier 3 are
more complex approaches, possibly models.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
26
Table 6: Impact of JI projects on global emissions if the host country does not have
significant expected AAU surplus+
Non-
additional***
Emission reductions (ER) correctly
reflected in host country inventory
Emission reductions (ER) not correctly
reflected in host country inventory
Zero: emissions decrease in host country
and increase in buyer country by number
of ERUs sold.
Zero: emissions decrease in host country and
increase in buyer country by number of ERUs
sold.
+ We classify a JI host country as having a “significant expected surplus” if the country’s average 2003-2007 emissions
were more than 20% below its CP1 reduction target, which we see as a strong indication that a country could have
expected to have significant surplus well before 2008, when JI crediting started.
* Zero means global emissions are not changed compared to a scenario without JI.
** Decrease means a net atmospheric benefit is achieved.
*** We did not distinguish between over-, under- and correctly credited non-additional projects since although the
number of ERUs would vary in each case, the outcome has no impact on global emissions and the host country has to
engage in more mitigation activity equivalent to the number of ERUs issued.
Project and ERU
characteristics
Impact on global emissions
Additional and…
Emission reductions (ER) correctly
reflected in host country inventory
Emission reductions (ER) not correctly
reflected in host country inventory
… correctly
credited
Zero*: emissions decrease in host country
and increase in buyer country by number
of ERUs sold.
Decrease**: equivalent to ERs from the JI
project. Host country loses AAUs but
equivalent ERs are not reflected in its
inventory.
Example:
JI project reduces
100 tonnes and
receives 100 ERUs
Host country emissions decrease 100t:
100 AAUs converted to ERUs, 100t
reduced
Buyer country emissions increase 100t
Total GHG emissions change: Zero
Host country emissions decrease 200t:
100 AAUs converted to ERUs, 100t reduced
but not reflected in emissions inventory
host country has to decrease emissions by
another 100t to compensate for 100 lost
AAUs
Buyer country emissions increase 100t
Total GHG emissions change:
100t 100t +100t = Decrease of 100t
… overcredited
Zero: emissions decrease in host country
and increase in buyer country by number
of ERUs sold.
Decrease**: equivalent to ERs from the JI
project. Host country loses AAUs but
equivalent ERs are not reflected in its
inventory.
Example:
JI project reduces
100 tonnes and
receives 120 ERUs
Host country emissions decrease 120t:
120 AAUs converted to ERUs. 100t
reduced. host country has to decrease
its emissions by 20t more to compensate
for 20 overcredited ERUs = 20 lost AAUs
Buyer country emissions increase 120t
Total GHG emissions change is zero:
100t 20t +120t = 0t
Host country emissions decrease 220t:
120 AAUs converted to ERUs. 100t reduced
but not reflected in emissions inventory
has to decrease emissions by 120t more to
compensate for 120 lost AAUs
Buyer country emissions increase 120t
Total GHG emissions change:
100t 120t +120t = Decrease of 100t
…undercredited
Zero: emissions decrease in host country
and increase in buyer country by number
of ERUs sold.
Decrease**: equivalent to ERs from the JI
project. Host country loses AAUs but the
equivalent ERs are not reflected in its
inventory.
Example:
JI project reduces
100 tonnes and
receives 80 ERUs
Host country emissions decrease 100t:
80 AAUs converted to ERUs. 100t
reduced. host country has 20 extra
AAUs available due to 20 none-credited
ER. Host country has to decrease
emissions by 20t less
Buyer country emissions increase 80t
Zero-sum: 100t +20t +80t = 0t
Host country emissions decrease 180t:
80 AAUs converted to ERUs. 100t reduced
but not reflected in emissions inventory
has to decrease emissions by another 80t to
compensate for 80 lost AAUs
Buyer country emissions increase 80t
Total GHG change:
100t 80t + 80t = Decrease of 100t
Example:
JI project receives
100 ERUs, but
does not reduce
emissions
Host country emissions decrease 100t:
100 AAUs converted to ERUs but no ERs
has to decrease emissions by 100t
more to compensate for 100 lost AAUs.
Buyer country emissions increase 100t
Total GHG emissions change is zero:
100t +100t = 0t
Host country emissions decrease 100t:
100 AAUs converted to ERUs but no ERs
has to decrease emissions by 100t more to
compensate for 100 lost AAUs.
Buyer country emissions increase 100t
Total GHG emissions change is zero:
100t +100t = 0t
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
27
Table 7: Impact of JI projects on global emissions if the host country has a significant
expected surplus+
Project and ERU
characteristics
Impact on global emissions
Additional and
Emissions reductions (ER) correctly
reflected in host country inventory
Emissions reductions (ER) not correctly
reflected in host country inventory
Non-
additional***
Emissions reductions (ER) correctly
reflected in host country inventory
Emissions reductions (ER) not correctly
reflected in host country inventory
+ We classify a JI host country as having a “significant expected surplus” if the country’s average 2003-2007 emissions
were more than 20% below its CP1 reduction target, which we see as a strong indication that a country could have
expected to have significant surplus well before 2008, when JI crediting started.
* Zero means global emissions are not changed compared to a scenario without JI.
** Decrease means a net atmospheric benefit is achieved.
*** We did not distinguish between over, under and correctly credited non-additional projects since although the number
of ERUs would vary in each case, the outcome is the same: an increase in global emissions equivalent to the number of
ERUs issued.
… correctly
credited
Zero*: emissions decrease in host country
and increase in buyer country by number
of ERUs sold.
Zero: emissions decrease in host country and
increase in buyer country by number of ERUs
sold.
Example:
JI project reduces
100 tonnes and
receives 100
ERUs
Host country emissions decrease 100t:
100 surplus AAUs converted to ERUs,
100t reduced.
Buyer country emissions increase 100t
Total GHG emissions change: Zero
100t +100t = 0t
Host country emissions decrease 100t:
100 surplus AAUs converted to ERUs, 100t
reduced but not reflected in emissions
inventory.
Buyer country emissions increase 100t
Total GHG emissions change: Zero
100t +100t = 0t
…overcredited
Increase equivalent to number of
overcredited ERUs
Increase equivalent to number of overcredited
ERUs
Example:
JI project reduces
100 tonnes and
receives 120
ERUs
Host country emissions decrease 100t:
120 surplus AAUs converted to ERUs,
100t reduced, 20 surplus AAUs are lost
Buyer country emissions increase 120t
Total GHG emissions change:
100t +120t = Increase of 20t
Host country emissions decrease 100t:
120 surplus AAUs converted to ERU, 100t
reduced but not reflected in emissions
inventory, thus 120 surplus AAUs are lost
Buyer country emissions increase 120t
Total GHG emissions change:
100t +120t = Increase of 20t
…undercredited
Decrease** equivalent to the ERs not
credited
Decrease equivalent to the ERs not credited.
Example:
JI project reduces
100 tonnes and
receives 80 ERUs
Host country emissions decrease 100t:
80 surplus AAUs converted to ERUs, 100t
reduced, host country has 20 extra AAUs
available.
Buyer country emissions increase 80t
Total GHG emissions change:
100t +80t = Decrease of 20t
Host country emissions decrease 100t:
80 surplus AAUs converted to ERU, 100t
reduced but not reflected in emissions
inventory. Thus 80 surplus AAUs are lost.
Buyer country emissions increase 80t
Total GHG emissions change:
100t +80t = Decrease of 20t
Increase equivalent to the number of
ERUs issued
Increase equivalent to the number of ERUs
issued
Example:
JI project receives
100 ERUs, but
does not reduce
emissions
Host country emissions stay the same:
100 surplus AAUs converted to ERUs.
Buyer country emissions increase 100t
Total GHG emissions change:
increase of 100t
Host country emissions stay the same:
100 surplus AAUs converted to ERUs.
Buyer country emissions increase 100t
Total GHG emissions change: increase of
100t
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
28
Table 6 shows that no matter what the environmental integrity of the JI project global
emissions do not increase if the JI host country does not have a significant AAU surplus.
In this case, the overall emissions are determined by the emissions budget assigned under the
KP. If projects are non-additional or overcredited, the cost will be borne by the host country as
it will have to engage in more mitigation activity in order to meet its own reduction target. If JI
projects are undercredited, the opposite holds true: the host country has to engage in less
mitigation compared to not hosting such JI projects.
13
If JI emissions reductions are not
reflected in the host country’s emissions inventory, they may lead to a net atmospheric benefit
a decrease of global emissions beyond that which could have been expected by the reduction
targets. The cost for such net atmospheric benefit will be borne by the host country: it will
have to engage in more mitigation activity equivalent to the number of emission reductions not
reflected in its inventory in order to meet its mitigation target.
Table 7 shows that if the JI host country has significant AAU surplus, non-additionality
as well as overcrediting leads to higher global emissions, whereas undercrediting leads
to some net atmospheric benefit. Regardless of the project’s environmental integrity and
inclusion in the national GHG inventory, there is no need for the host country to engage in
more mitigation action because it can issue the ERUs using its AAU surplus. Therefore, JI
host countries with significant AAU surplus have little economic incentive to ensure the
environmental integrity of ERUs.
4. ASSESSMENT OF APPROACHES FOR DEMONSTRATING ADDITIONALITY
This section provides an assessment of the main approaches used for additionality
demonstration in JI projects. We use the random sample of 60 projects to assess how JI
projects demonstrated additionality and discuss the likelihood of projects being additional.
The rules for Track 2 projects are set by the JISC. Track 1 project additionality requirements
are determined by the host countries but often Track 2 rules are followed, see Section 2 for
details. Many JI projects use the CDM additionality tool or elements of it to demonstrate
additionality (UNFCCC 2008). For this reason we assess the following core elements of the
CDM additionality tool:
Identification of alternative scenarios
Investment analysis
Barrier analysis
Common practice analysis
We also look at other aspects of demonstrating additionality, such as prior consideration of JI,
the use of retroactive crediting, and the overall timing of steps in the project cycle. We also
assess the demonstration of additionality through a reference to other registered projects.
Differences between project types, host countries, the track of JI projects are assessed in
Sections 5, 6 and 7, respectively.
4.1 Identification of alternative scenarios
As a first step, many PDDs identify realistic and credible alternatives that provide outputs or
services comparable with the proposed JI project. This is also the required first step in the
CDM additionality tool. In our sample, 85% of the 54 projects for which PDDs are available
identify alternative scenarios. Most projects identify several alternative scenarios, although
13
If undercrediting is too severe, it may lead to some JI projects not being implemented.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
29
24% of the projects for which PDDs are available list only the current status quo or no
implementation of the JI project as an alternative. For example, partial implementation of a
project without JI is often not considered as an alternative, in particular for projects that
bundle several activities. The justifications why other alternatives are not further considered
are in some cases rather weak.
4.2 Investment analysis
The investment analysis is used to demonstrate that a proposed JI project would have been
economically unattractive or less attractive than another plausible course of action (and thus
would not have proceeded) without the incentives from JI. Of the 54 sampled projects for
which PDDs are available, 54% use the investment analysis to demonstrate additionality.
These 29 JI projects account for 43% of the ERUs issued to the sample projects.
14
The projects use the following approaches for the investment analyses (see Table 8):
The simple cost analysis is used to demonstrate that a project has no revenues other
than ERUs but involves costs (such as for example secondary N2O abatement from
nitric acid production);
The investment comparison analysis is used to demonstrate that the proposed
project activity is economically and financially less attractive than another alternative
that provides similar outputs or services;
The benchmark analysis is used to demonstrate that a proposed project is, without
revenues from ERUs, economically not attractive (i.e. it does not meet a financial
benchmark).
Table 8: Type of investment analysis used
Type of investment analysis
Number of
projects
% of projects
% of ERUs issued to projects that
used the investment analysis
Data source: Random sample of 60 projects
Demonstrating additionality is straightforward for projects that do not save costs or generate
revenue other than from JI because there is little incentive to implement the project as long as
it is not mandated and enforced by the government. This applies for example to most
industrial gas destruction projects. In our sample, five of the eight projects that use the simple
cost analysis are projects abating N2O from nitric acid production. However, the three other
projects apply the simple cost analysis despite the fact that they reduce costs or generate other
revenues: one agricultural no-till project saves costs from tillage operations, one project
avoiding PFC emissions in the aluminium industry reduces electricity consumption, and one
project for N2O abatement from adipic acid production saves costs from steam generation.
14
Some projects provide information on their costs or profitability but do not use this information to justify
additionality and some projects claim financial or cost barriers without conducting a full investment analysis. We
do not classify these projects as conducting investment analysis. We exclude one programme of activities for
which no CPA-DD is available on any specific activities (DE1000469). We include projects which conduct an
investment analysis as part of claiming financial or cost-related barriers. One project (UA1000416) conducted both
a simple cost analysis and a benchmark analysis for two sub-components; as the benchmark analysis covers the
main investment of the project, we classified this project as conducting benchmark analysis.
Simple cost
8
28%
52%
Benchmark
19
66%
48%
Investment comparison
2
7%
1%
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
30
Using the simple cost analysis for these project types is methodologically inappropriate and
can lead to a wrong conclusion on the assessment of additionality.
The additionality proof is inherently more difficult for projects that save costs or generate
revenue, such as power or energy efficiency projects. Such projects have to apply a
benchmark or investment comparison analysis. One of the criticisms to the application of the
investment analysis under the CDM was a lack of transparency (Schneider 2009). Project
developers may not reveal all information necessary to allow replication of their calculations
and may not provide appropriate justifications for their input values. However, transparency
of the investment analysis is crucial to assess its appropriateness.
In our random sample, only two of the 21 projects that used the investment comparison or
benchmark analysis provided transparent information on costs and revenues in the PDD
which would allow a third party to reproduce the calculation. The level and type of
information is often insufficient or questionable. Many projects do not provide sufficient
information on assumptions and input parameters used in the analysis, such as investment
costs, operation and maintenance costs, or revenues, neither for the baseline nor the project
scenario. One third of the projects provide only information on the result of the investment
analysis, one third provide information on investment costs and the result, and one third also
specify some revenues from project operation. Figure 2 shows that in most projects input
parameters were not provided transparently or appropriately justified. Out of the 21 projects
that used the investment comparison or benchmark analysis, we identified only one project
with a transparent investment analysis. Several PDDs refer to confidential information that is
not made publicly available.
Figure 2: Level of information provided on input parameters to the investment analysis
Data source: Random sample of 60 projects (21 using investment comparison or benchmark analysis).
The financial benchmark is a key parameter for the investment comparison and benchmark
analysis. Among the 21 sampled projects which use investment comparison or benchmark
analysis, four did not include information on which value was used in the calculation, and
seven of the 17 projects which provided a value did not provide clear information how the
value was derived.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
31
The values used for the financial benchmark varied significantly, from 8% to 25%, with an
average of 15.8%. Most projects used values in the range of 15%25%. The average value of
15.8% used in the sampled JI projects is significantly higher than the average value of 9.3%
used in CDM projects.
15
Using high financial benchmarks enables JI projects that are rather
attractive economically to pass the additionality test. Indeed, more than half the sampled JI
projects which provided the relevant information would not pass the additionality test if the
average financial benchmark under the CDM were used.
16
One project (RU1000431) even
reported an IRR without ERU revenues of 22.7% and used a benchmark of 25% to
demonstrate additionality. The use of significantly higher financial benchmarks under JI than
under the CDM does not appear appropriate, since the risks in the most important JI countries
(Ukraine, Russia) were not significantly higher than those of major CDM countries (China,
India, Brazil) in the period up to 2012. The use of significantly higher values under JI than
under the CDM also raises questions whether AIEs assessed financial benchmarks with the
same scrutiny as under the CDM.
Another indicator for the likelihood of additionality is the impact of ERU revenues on the
economic attractiveness of the project (Sutter and Perreno 2007; Schneider 2009). The larger
the impact of ERU revenues, the more likely it is that ERU revenues played a role in the
decision to proceed with the investment. Of the 21 projects which applied investment
comparison or benchmark analysis, 12 projects provided information on the project IRR with
and without ERU revenues. The difference in IRR due to ERU revenues ranges from 0.4% to
23.3%. This large diversity indicates that for some projects ERU revenues could have played
a major role, while for others it seems unlikely that the incentives from JI played a role. For
example for one project (PL1000534), which claimed additionality by comparison to a similar
project (and not based on investment analysis), the IRR changes only by 0.08% due to ERU
revenues, from 5.87% to 5.94%. The ERU prices assumed in PDDs vary considerably,
between 1 EUR and 25 EUR.
A sensitivity analysis is an integral part of the investment comparison and benchmark analysis
under the CDM. Among the 21 sampled projects that used the investment comparison or
benchmark analysis, five did not conduct a sensitivity analysis. Among the projects which
conducted the analysis, the type of parameters varied, including within similar project types.
The range of the variation was mostly 10%, as under the CDM.
Finally, we observed that several projects were funded or subsidised through other sources of
finance, in particular loans provided by the European Bank for Reconstruction and
Development. Only in some of the PDDs these loans are mentioned. The provision of loans
through ODA usually many years before the projects sought determination or registration
under JI further questions the additionality claims of these projects, since the banks seemed
confident that the loan could be repaid.
4.3 Barrier analysis
The barrier analysis is used to demonstrate that a project would not be implemented in the
absence of JI because it would face significant barriers. Under the CDM, the barrier analysis
can be used as an alternative to investment analysis or to supplement it. An important
15
The value for CDM projects was derived from more than 4000 registered CDM projects included in the IGES
CDM investment analysis database as of August 2014 (IGES 2014).
16
Eighteen sampled projects provided a value for the internal rate of return (IRR) of the project without ERU
revenues; for 10 of these projects the project IRR without ERU revenues was higher than the average benchmark
value used under the CDM (9.3%).
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
32
criticism of the barrier analysis is that it is very difficult to assess objectively whether barriers
are prohibitive for implementing a project and whether JI alleviates such barriers.
In our random sample 38 of the 54 projects for which PDDs are available use the barrier
analysis. The barriers most often mentioned are costs or financial risks, lack of capacity,
prevailing practice, and technology risks (see Table 9).
Table 9: Types of barriers cited in the sampled projects
Barrier
Projects
ERUs
Data source: Random sample of 60 projects (of which 38 use barrier analysis).
Overall, the barriers are often not substantiated by evidence. Of 38 projects that used the
barrier analysis, 32 include either no explanation or only a short explanation on how JI will
help overcome the barriers. In many cases the barrier analysis by itself seems insufficient to
differentiate between additional and non-additional projects because its application is highly
subjective and difficult to verify. Nevertheless, nine of the 38 projects used the barrier
analysis as the only additionality test and did not apply an investment analysis or a common
practice test.
4.4 Common practice analysis
The common practice analysis is used to demonstrate that a proposed project activity is not
frequently implemented in the sector. The key challenges of the common practice analysis are
1) defining what is regarded a similar technology used to compare the proposed activity to, 2)
what geographical scale should be considered, and 3) what threshold should be used to assess
whether a project is deemed common practice.
The definition of similar technologies is important because the project technology could be
defined very narrowly (e.g. small scale hydro power), while the technology it is compared to
could be defined broadly (e.g. the country’s power production), in order to show that the
project technology has a low market penetration. Choosing the project technology and the
peer group carefully is therefore essential to ensure the common practice test can identify
projects that are already business-as-usual. The geographical scale is important because
market penetration could vary, e.g. depending on geographical or economic circumstances.
Two thirds of the sampled projects for which PDDs are available use the common practice
analysis. Of these 36 projects, 30 use the host country as geographical scale, four projects the
whole of Europe, and two project a region within the host country.
Under the CDM a quantitative analysis needs to be conducted and common practice is
defined as a market penetration rate of more than 20%. Among the 36 sampled projects which
conduct a common practice analysis, only four projects provide a quantitative assessment.
These projects specify the observed market penetration rate for the project technology, which
ranges from 0 to 23%, but only one project explicitly uses the CDM threshold of 20%; the
other three do not provide explanations which market penetration rate is deemed as common
Costs or financial risks
100%
100%
Lack of capacity
55%
60%
Prevailing practice
47%
29%
Technology risks
47%
28%
Political
24%
11%
First of its kind
13%
12%
Not preference of management
5%
15%
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
33
practice. Thirty-two of the 36 projects, accounting for more than 97% of ERU issuance to the
sampled projects, give only qualitative descriptions. These can be harder to verify, especially
when insufficient information is provided. For example, several projects claim that a
particular technology is not used anywhere else in the host country or only in other JI
projects, without providing references, supporting documentation, or statistics. For six
projects we identify clear flaws, such as the common practice test being performed for a
different technology than the one used by the project.
Overall, in most cases the information provided in PDDs is not sufficient to conclude whether
or not a project is common practice. Under the CDM, the common practice test cannot be
used as a standalone test. However, under JI, three of the sampled projects that use the
common practice analysis applied it as a standalone test.
4.5 Prior consideration
In assessing the additionality of JI projects, one important aspect is whether JI was considered
in the decision-making process to proceed with the implementation of a project. If JI was
considered, the project is not necessarily additional; however, if JI was not considered at all,
the project is clearly not additional.
Under the CDM, project developers have to notify the UNFCCC of their intent to register a
CDM project within six months of the decision to proceed with the project (or before these
current rules they had to provide written evidence that the CDM was known to them). Under
JI, there are no rules that require the proof of prior consideration.
Many PDDs include information on the history of the project and when the decision was
made to proceed with the project. They often provide some kind of evidence to support
information when the decision was made, but given the absence of requirements for prior
consideration only some state that JI was considered in proceeding with the project. In most
cases, the evidence only supports when the project was started (e.g. contracts for equipment
purchase) but not whether JI was considered in proceeding with the investment decision.
However, even in the absence of rules for prior consideration, whether and when JI was
considered by the project developers can be assessed based on the project implementation
timeline, specifically by comparing the starting date of the project
17
with the date of
issuance of the project’s Letter of Endorsement (LoE). If JI was seriously considered and
important for project implementation, it is reasonable to assume that the project participants
would have tried to secure the status of JI for their activity as early as possible by applying for
an LoE. Obtaining an LoE is the first step towards JI registration in most host countries, did
not require elaborate project documents, and was usually not difficult or time-consuming
(unlike getting a Letter of Approval).
18
Figure 3 illustrates the time period between the project starting date and the issuance of the
LoE for those 36 projects in our sample of 60 for which the LoE issuance date is available.
17
The starting date of a JI project is the date on which the implementation or construction or real action of the
project begins (ref. Guidelines For Users Of The Joint Implementation Project Design Document Form, JISC)
18
In Russia the applicable JI procedures do not require an LoE. For these countries LoEs are not available. We
therefore limited our analysis to those countries and projects for which LoEs are available.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
34
Figure 3: Time period between project starting date and issuance of LoE
Data source: Random sample of 60 projects (of which the date of the LoE is available for 36 projects)
One quarter of the projects applied for and received the LoEs before their project start. For
these projects it is clear that they considered JI when proceeding with the project. Four
received their LoEs within a year of their starting date. Considering that LoE issuance
requires some time after submitting the application for an LoE, these projects may have
considered JI as well when proceeding with the project. Another four projects received their
LoE within three years of their starting date. For this group it is not clear whether the issuance
of the LoE was delayed, e.g. due to bureaucratic processes of the host country, or whether the
project participants did not consider JI earlier and therefore did not apply for the LoE.
More than 50% of the projects received their LoEs three or more years after the project start.
The majority of these projects obtained their LoEs in 2012 when project endorsements
sharply increased (see Figure 4). In fact, all projects that were endorsed in the last year of CP1
received their LoEs three or more years after the project start. For more than a quarter of the
analysed projects (10) the discrepancy is particularly large: the time gap between project start
and LoE exceeds seven years; some projects received their LoEs at the end of 2012, while
their starting dates go back to 2002.
19
In other words, these projects were initiated 10 years
before they received an LoE. It is unlikely that projects in this group considered JI when
proceeding with their implementation; these projects are very unlikely to be implemented due
to the incentives of JI and are thus very unlikely to be additional.
19
E.g. projects UA1000422, “Implementation of Energy Saving Measures at PJSC Khartsyzsk Pipe Plant,” and
“Lvivoblenergo PJSC Power Distribution System Modernization”
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
35
Figure 4: Issuance of LoE by year
Data source: Random sample of 60 registered projects.
Overall, the lack of any rules with regard to prior consideration is a major shortcoming of JI,
which severely impacted the integrity of the mechanism.
4.6 Retroactive crediting of emissions reductions
Under current JI rules, projects may be implemented before the start of CP1 but ERUs are
only granted for reductions occurring during the commitment period. However, projects can
be registered after their start and be issued ERUs for reductions that occurred before
registration. In other words, under the current JI rules ERUs can be issued retroactively for
the period from 2008 (so-called retroactive crediting), provided that the project was
operational and delivered emission reductions (see also Section 2.2).
Retroactive crediting was meant to enable projects to go ahead before JI was operational and
avoid possible delays due to the approval and registration processes. Thus, the projects did
not lose ERUs if they started operating before JI registration was completed.
However, this rule also enables activities that were not originally intended as JI projects to
receive ERUs retroactively, even if they did not make attempts to register as JI projects for
many years after they were implemented.
To examine the extent of retroactive crediting, we analysed the date of the host country
approval, the crediting period start date, and the number of ERUs issued for all registered JI
projects included in the UNEP Risoe database. As the registration dates are not available for
most JI Track 1 projects, we use the date of host country approval (LoA issuance) instead of
the registration date and deem that projects which received their LoA after the start of their
crediting period benefited from retroactive crediting.
Almost half of all projects were approved in 2012 and these projects generated 71% of all
ERUs issued (more than 590 million ERUs). All of these projects retroactively claimed ERUs
for previous years. Two thirds have their crediting period starting in 2008. This means that
these projects started operation in 2008 or earlier, but did not get approved until 2012 and
then claimed all or most of their ERUs for the past years retroactively. It should be noted that
in the last year of CP1 the prospects for starting new JI projects were highly insecure given
the uncertainty of continuation of JI and demand for ERUs in CP2. It is therefore likely that
the main motivation for registering these projects was not to generate ERUs in CP2 but to
claim CP1 ERUs retroactively. All projects approved in 2011 and 98% of the projects
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
36
approved in 2010 also claim ERUs retroactively. This means the majority of all ERUs were
issued using retroactive crediting.
20
In contrast, for many of the projects the whole JI cycle, from the issuance of the LoE to the
first ERU issuance, was completed in a very short time. We analysed the time period from the
initial project endorsement (LoE) to the first issuance of ERUs for the 25 projects in our
sample for which the dates of both the LoE issuance and first ERU issuance are available. As
shown in Figure 5, the projects fall into two separate groups: for 14 projects, the project cycle
took more than one year (usually more than two years), which seems reasonable considering
that PDD determination, host country approval, verification and ERU issuance typically take
time. However, 11 projects, all approved in 2012, completed all project cycle steps in less
than six months. One project received the host country LoE and LoA, prepared the PDD and
monitoring report, performed PDD determination, got registered, verified emissions
reductions and received ERUs in less than 40 days.
21
Figure 5: Project cycle duration in sampled projects (time between LoE and ERU
issuance)
Data source: Random sample of 60 projects (for only 25 of which both the LoE and first ERU issuance dates are available).
A significant time gap between project start and JI approval, combined with a very short
approval process, raises serious questions about whether projects considered JI when deciding
to implement activities and thus about the additionality of these projects.
4.7 Demonstration of additionality by reference to a comparable project
One of the approaches to demonstrate additionality permitted under Track 2 is to provide
evidence that the proposed activity is comparable to another project that has already been
registered under Track 2; see Section 2.2. When using this approach, a project does not have
to perform its own additionality test, but can refer to a similar project, demonstrating that it 1)
uses the same emission reduction technology; 2) is located in the same host country and the
starting dates of the proposed and the similar project are not more than five years apart; 3) the
20
We did not calculate the exact share of ERUs that were issued retroactively. This would require checking the
monitoring reports of each project. Not all monitoring reports for all projects are available.
21
Project UA1000380.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
37
difference in scale of activities does not exceed 50%; and 4) the regulatory framework has not
changed in a manner that would affect the baseline of these projects (UNFCCC 2011b).
Only seven projects in the sample use this approach to demonstrate additionality, but these
projects are large: they account for 25% of the ERUs issued to the projects in the random
sample. In applying this approach the projects sometimes make long chains of references.
22
We traced from the sampled projects up to five sequential references from one project to
another. Six of the seven projects involve coal extraction from waste piles, which is the
largest project type in terms of ERU issuance (see Section 5.2). We show there that such
chains of references can reduce the comparability of the projects: the last project in the chain
may not be comparable with the first project for which additionality was initially determined.
The demonstration of additionality by making a reference to a comparable project is also
prone to possible flaws in determining additionality in the original project. If additionality is
not appropriately determined in the original project, the wrong conclusion on additionality
could also apply to all following projects that use it as a reference.
4.8 Overall assessment of the likelihood of additionality of JI projects
In this section we assess the overall plausibility of the additionality claims by JI projects
based on the information on additionality determination available from the 60 sampled
projects. We evaluate the plausibility of the additionality claims of each project by assessing
the plausibility of the timeline of project implementation and registration under JI as well as
the information on the main additionality tests used to determine additionality (investment
analysis, barrier analysis, common practice analysis, reference to a comparable project).
We use three broad categories to classify each project:
Plausible means that, based on the available information, the claims for
demonstrating additionality seem plausible;
Questionable means that the available information raises questions or doubts about
the additionality;
Not plausiblemeans that the available information suggests that the projects are
unlikely to be additional.
For six of the 60 projects a PDD is not available; we exclude these projects from our
assessment. For the remaining 54 projects we use a consistent approach to classify each
project in one of the categories. With regard to the timeline of project implementation and
registration, we assess when the project was initiated, i.e. when the decision was taken to
proceed with the project, and when the project made apparent efforts to seek JI status, i.e.
when the project received a LoE or when the determination report was prepared.
23
As pointed out in Section 4.5, projects may require some time after their initiation to seek JI
status, e.g. for preparing a PDD and contracting a DOE or submitting a request for a LoE.
However, if projects do not make any efforts to seek their JI status for several years after their
initiation, it seems questionable that JI was decisive for the implementation of the project. We
use the timelines to classify the projects:
22
The PDDs of these projects also contain elements of other additionality tests, but the demonstration of similarity
to another registered project is their main additionality argument.
23
For the three projects we did not have LoE or determination dates, we used the LoA or PDD date, whichever
was earlier.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
38
Up to three years: If a project made apparent efforts to seek JI status either before
project initiation or within three years thereafter, we deem the timeline as plausible.
Three to five years: We deem the additionality claims as questionable for projects
that have not made apparent efforts to seek their JI status within three to five years of
their initiation.
More than five years: We deem the additionality claims as not plausible for
projects that have not made apparent efforts to seek their JI status within five years of
their initiation.
With regard to the tests used to demonstrate additionality, we consider the tests plausible if
the tests are applied correctly, e.g. a simple costs analysis is applied to a project type that does
not generate revenues other than ERUs, and if key information on the test is provided and
justified, e.g. if key input parameters for the investment analysis are provided and justified, if
the financial benchmark is derived in a transparent manner, or evidence is indicated for the
main barrier. We also deem the additionality claims as plausible for projects that do not lead
to cost savings or generate revenues other than ERUs. Where projects use long chains of
reference to a similar project as a means of additionality demonstration, with the first project
being substantially different from the last, we consider the approach as questionable.
Where a project has both a questionable timeline and questionable application of the
additionality test, we consider the overall additionality claim of the project as not plausible.
It is important to note that this approach has clear limitations. First, the classification can only
consider information that is publicly available, which is in some cases limited. Very early
projects may have provided rather limited information to support additionality claims and are
therefore classified as questionable but which may nevertheless be additional. Similarly, the
plausibility of the information presented does not ensure that a project classified as
plausible is actually additional. And second, the assessment of the completeness,
appropriateness and plausibility of the additionality tests is to some extent subjective, in
particular for more subjective tests, such as the barrier test. We nevertheless argue that this
assessment is useful to provide an overall impression of the frequency, seriousness and
impact of the issues identified in the previous sections.
Figure 6 shows the results of the classification by the number of projects and the ERUs issued
to the projects. The 12 projects with plausible additionality claims account for 14% of the
ERUs issued to the sampled projects; the 16 projects with questionable additionality claims
account for 12% and the 26 projects with additionality claims that were not plausible account
for 73% of the ERUs issued.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
39
Figure 6: Plausibility of the additionality claims of the sample projects
Data source: Random sample of 60 JI projects.
Figure 7 indicates that the share of projects with additionality claims that were not plausible
grew significantly in 2012, correlating with our findings that projects approved earlier were
crowded out by projects with more questionable quality that were approved in 2012 and
retroactively received ERUs.
Figure 7: Plausibility of additionality claims of projects by year of approval
Data source: Random sample of 60 registered projects, excluding the 6 projects without PDDs.
Figure 8 shows additionality rating of the sampled projects by total ERU issuance to each
project. The figure indicates that additionality for larger projects seemed to be more
questionable than for smaller projects. The category of projects with more than 1 million
ERUs issued is dominated by projects with additionality claims that are not plausible. The 14
large projects with additionality claims that are not plausible generated 68% of ERUs issued
to the sampled projects. It is notable that the vast majority of ERUs issued to all projects were
generated by very large projects. The 10 JI projects with more than 10 million ERUs
generated more than 23% of ERUs issued. The 46 projects that generated more than 5 million
ERUs generated 52% of all ERUs and the 179 projects with more than one million ERUs
generated more than 90% of ERUs.
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
40
Figure 8: Plausibility of additionality claims of projects by size
Data source: Random sample of 60 registered projects, excluding the 6 projects without PDDs.
5. ASSESSMENT OF SPECIFIC JI PROJECT TYPES
In this section we examine the six project types with the highest ERU issuance in greater
detail, in order to assess the degree of environmental integrity of different project types. The
project types represent 84% of the ERUs issued and 53% of registered projects in CP1.
Table 10: Project types evaluated, number of projects and their ERU shares
Project types evaluated
Number of
registered projects
ERUs issued
(millions)
% of ERUs
issued in CP1
Total
341
705
83.8%
Source: UNEP Risoe (2014)
We did not evaluate or rate the project types listed in Table 11. These account for 16% of
ERUs issued in CP1.
Spontaneous ignition of coal waste piles
78
219
26.1%
Energy efficiency in industry and power
production and distribution
164
195
23.1%
Associated petroleum gas utilization
22
117
13.9%
Natural gas transportation and
distribution
32
83
9.8%
HFC-23 abatement from HCFC-22 and
SF6 abatement
4
54
6.4%
N2O abatement from nitric acid
41
36
4.5%
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
41
Table 11: JI project types not evaluated and rated
Project types
Registered projects
% of ERUs issued in CP1
Total
299
16%
Source: UNEP Risoe (2014)
5.1 Spontaneous ignition of coal waste piles
Overview
This project type primarily involves avoidance of GHG emissions from uncontrolled fires due
to spontaneous self-ignition of coal waste piles. Coal waste piles contain a certain share of coal
which was not extracted from the bedrock. Large coal waste piles with high coal content can
self-ignite and burn for years. Apart from CO2 emissions, coal waste fires cause severe local air
pollution, including particulate matter, carbon monoxide, polycyclic aromatic hydrocarbons,
nitrogen oxides and other noxious gases (Ewall 2007; Stracher and Taylor 2004).
24
JI projects of this type can be divided into two sub-types according to the approaches applied
to deal with self-ignition of coal waste:
1. Coal extraction and combustion: 65 projects extract coal from coal waste piles, leaving
bare rock which does not ignite, and combust the extracted coal, mostly in power plants.
Emission reductions are claimed for the avoidance of waste pile fires, while emissions
from combustion of the extracted coal are not counted because it is assumed to substitute
coal which would be otherwise obtained from coal mines. For the amount of coal that
would otherwise be obtained from coal mines, projects also claim emissions reductions
for avoiding upstream emissions from coal mining, including methane emissions
24
See: http://www.energyjustice.net/coal/wastecoal.
Fossil fuel switch
17
3%
Agriculture: no tillage
7
3%
PFC reduction
6
2%
Energy efficiency in service sector
23
2%
N2O abatement from adipic acid
3
2%
Coal mine methane
28
1%
Biomass energy
46
<1%
Hydro
20
<1%
Wind
43
<1%
Cement
4
<1%
Afforestation
2
<1%
Energy efficiency in households
12
0%
Landfill gas
67
0%
Methane avoidance
7
0%
Geothermal
5
0%
Avoided deforestation
1
0%
Transport
4
0%
Agriculture
4
0%
HAS JI REDUCED GHG EMISSIONS? LESSONS FOR THE DESIGN OF CARBON MARKET MECHANISMS SEI WP 2015-07
42
associated with deep coal mining and CO2 emissions from electricity consumption by
coal mines (see Figure 9).
25
2. Fire extinguishing: 13 projects implement measures to extinguish and prevent fires in