The first two projects, both related to cookstoves, approved for use under the UN’s carbon market massively overestimate their climate impact, despite efforts to rein in overcrediting.

In April of last year, Carbon Market Watch published an analysis demonstrating that the first Clean Development Mechanism project (known as Programme of Activities, or PoA, 10415) in line to transition to the new UN carbon market established under Article 6.4, known as the Paris Agreement Crediting Mechanism (PACM), was on course to issue 26 times more credits than than justified by the likely climate impact of the project. 

Following our alarming findings, the Article 6.4 Supervisory Body attempted to stem the bleeding. 

But do the improvements meaningfully limit over-crediting? We checked the numbers.

In March 2026, another Clean Development Mechanism (CDM) project, PoA 10471, received the first official approval to issue carbon credits. Joined quickly by PoA 10415, two projects have now received a green light for issuance. PoA 10471 is also a clean cooking project in Myanmar, implemented by a Korean NGO. The credits can be used in the Republic of Korea’s emissions trading system (ETS) and count toward its nationally determined contribution (NDC).

In official communications from the United Nations Framework Convention on Climate Change (UNFCCC), Executive Secretary Simon Stiell praised the project and the chair of the Article 6.4 Supervisory Body, Mkhuthazi Steleki, stressed that “by applying updated values and more conservative calculations, the credited reductions are about 40% lower than what older systems would have issued. The result is consistent with environmental integrity requirements and ensures that each credited tonne genuinely represents a tonne reduced and contributes to the goals of the Paris Agreement.”

Tightening the screws

These updated values and more conservative calculations are not a voluntary gesture of goodwill by the applicants. To assuage fears about low environmental integrity following our analysis of PoA 10415, the Supervisory Body upgraded what had previously been just an “encouragement provision” for transitioning CDM activities into a strict requirement. Transitioning projects must now re-evaluate their fraction of non-renewable biomass (fNRB) values and leakage factors.

PoA 10471 is the first project to have applied these new requirements. Specifically, the project’s fNRB value has been downsized by… you guessed it, roughly 40%. However, such a reduction is probably a mere drop in the ocean of over-crediting, as this newly approved batch of credits relies on the same shaky foundations as the transitioning project we analysed last year. While a 40% haircut for PoA 10471 sounds significant on paper, applying it to a baseline that appears to have been overestimated by as much as 1,100% still leaves the environmental integrity of these newly approved credits highly questionable.

A 2024 peer-reviewed study concluded that, among all cookstove methodologies, the AMS-II.G approach used by PoA 10471 is associated with the highest levels of over-crediting. This problematic methodology has also been rejected by the Integrity Council for the Voluntary Carbon Market (ICVCM). 

Overcooked… again

Our analysis of PoA 10471, released today, shows that this project will likely be over-credited by a factor of seven across the monitoring period authorised for transition. This shows that the updated fNRB and leakage requirements, while a step in the right direction, still fail to bridge the gap with peer-reviewed scientific literature. We calculate that, while the original 1.1 million credits received a 40% haircut through the fNRB adjustment, resulting in a little under 650,000 total credits becoming eligible for issuance, a 92% haircut would have been needed to land at the 90,000 or so credits that reflect the actual likely climate impact of the project.

The biggest remaining driver of over-crediting in the project is the omission of what is called “stacking” in the cookstoves world: the simultaneous use of other, potentially more polluting stoves by the beneficiaries that are not monitored by the project. 

Failing to account for the possibility that households operate multiple stoves creates the illusion of significantly lower emissions. To address this, scientific literature considers stacking rates of 68% to be conservative. However, neither the flawed AMS-II.G methodology nor the project documents we analysed reflect this. In fact, stacking is not accounted for at all.

While fNRB values and stacking rates are important for calculating emission reductions from cookstove projects, the entire set of variables identified as relevant in the scientific literature  – such as baseline fuel use, adoption and usage rates, conversion factors for charcoal and firewood, and the Hawthorne effect – must be included and reflect the best available science in a methodology to truly curb over-crediting from this project type. This is especially true for non-metered methodologies, such as AMS-II.G, where stove use is not directly monitored by a device attached to the project stove.

Dire consequences

The stakes for getting this right could not be higher. It’s already problematic that activities dating as far back as 2021 – the monitoring period of PoA 10471 spans from 1 January 2021 to 31 May 2022 – can generate new PACM credits, raising serious additionality concerns. Allowing massively overestimated transition projects to generate the first Article 6.4 credits adds insult to injury and does not just undermine the credibility of the PACM; it also actively harms the global fight against climate change, as these credits will be used as a substitute for real, domestic emission reductions. 

On top of this, it will negatively affect host countries, as credits they authorise under Article 6 will be deducted from the mitigation they can count towards their own climate targets. A mandatory requirement for “corresponding adjustments” is important to prevent double counting (one carbon credit used towards two climate targets); but this will make it harder for host countries to reach their NDCs if they approve hot air credits, since they will have to make up for the difference, probably through more costly measures in the future. 

This is not to mention the ongoing crimes against humanity being perpetrated against civilians by the military junta in Myanmar, which raises serious concerns about the human rights situation in the project areas, as well as questions about how well projects can be implemented and validated, including their ability to guarantee permanence over time due to people’s displacement and the environmental degradation that accompanies war. Global Forest Coalition will soon publish a more comprehensive analysis of PoA 10471 in relation to the situation in Myanmar.

Stopgaps and systemic fixes

Carbon Market Watch reiterates its call on all countries to hold off on approving or purchasing credits from CDM transition projects until they pass an independent quality assessment based on the latest science, and a full scrutiny of their human rights and social dimensions is carried out. Governments, such as South Korea’s, must not use these credits to meet their climate targets or allow them into compliance systems, such as emissions trading systems. 

Moreover, the Supervisory Body must dedicate special attention to the work of leading scientists in the field that suggest proactive improvements to methodologies, reflecting insights from years of field research on cookstove projects, and act upon science-based adjustments to the CDM transition standard.

Allowing hot air remnants from the rickety Clean Development Mechanism to enjoy the legitimacy of the supposedly better Paris Agreement Crediting Mechanism will ultimately undermine the credibility of the latter: overcrediting, at any magnitude, cannot be compatible with the climate ambition of a world striving to limit global warming to 1.5ºC.