Additionality, Permanence and Leakage in Plain Language

The three quality tests behind every carbon credit, explained with examples from forestry, renewables, cookstoves, waste and industry, plus the questions to ask about a listing.

A carbon credit is a claim that one tonne of CO₂ equivalent was kept out of the atmosphere, or taken out of it, because of a project. Three tests decide whether that claim holds up: additionality, permanence and leakage. Every standard has rules for them, and every verification report addresses them. But the rules leave room for judgement, so a buyer who understands the ideas will spot weak lots faster.

Additionality: would it have happened anyway?

A project is additional if the reductions would not have happened without the money from carbon credits. If a factory was always going to install efficient boilers because they paid back in two years, selling credits for that change rewards something that was coming regardless. The buyer's tonne then offsets nothing.

Standards test this in a few ways:

Financial tests. Does the project earn an acceptable return without carbon revenue? If not, credits plausibly made the difference.
Barrier tests. Were there technical, cultural or access barriers that carbon finance helped overcome?
Common practice. Is the activity already widespread in the region? If most similar sites already do it, the case is weak.
Regulatory surplus. Is the activity required by law? If so, it is not additional.
Standardised lists. Some methodologies treat certain small or unusual activities as automatically additional.

By project type:

Project type	Typical additionality question
Grid solar or wind in India	With falling costs and policy support, would the plant have been built on its own economics? Newer large plants face hard questions.
Household cookstoves	Usually strong, because poor households cannot afford efficient stoves without a subsidy.
Landfill or effluent methane capture	Strong where no rule requires capture and the gas has little commercial value.
Afforestation on degraded land	Strong where the land had no realistic alternative use that involved planting.
Industrial waste-heat recovery	Depends on payback period and whether the plant would have done it for cost savings.

What to ask: which additionality tool the methodology uses, when the project started relative to when carbon finance was first considered, and whether any part of it was mandated.

Permanence: will the carbon stay put?

Permanence matters most where carbon is stored rather than avoided. A tree that absorbs carbon and burns ten years later gives it back. A tonne of methane that a biogas digester stopped from forming does not come back.

So permanence is a big issue for:

forestry and agroforestry
mangroves, peatland and other blue carbon
soil carbon in farming

And a small issue for:

renewable energy, cookstoves, methane capture and efficiency projects, where the reduction has already happened once it is verified

Standards manage reversal risk mainly through a buffer pool. A share of each issuance from a land project is held back in a common reserve. If a fire, flood or illegal clearing destroys carbon, credits from the buffer are cancelled to cover the loss. The share depends on a risk assessment of the project, covering things like fire history, land tenure and the developer's finances.

Projects also commit to monitoring for a set period, often decades. Removals from engineered methods, like biochar or mineralisation, are usually argued to last much longer than trees.

What to ask: the buffer percentage and how it was calculated, how long monitoring continues, who holds legal rights to the land, and whether there has been any past reversal. Our article on forestry and agroforestry credits goes further for tree projects.

Leakage: did the emissions just move?

Leakage happens when a project cuts emissions in one place but causes them to rise somewhere else. The classic case is forest protection: if loggers are kept out of one forest and simply move to the next valley, the atmosphere gains little.

Other examples:

A cookstove project where families keep using their old stove alongside the new one, so firewood use drops less than claimed. (Strictly this is a monitoring issue, but methodologies treat related effects as leakage.)
Planting trees on cropland, pushing farmers to clear forest elsewhere to replace lost fields.
A biogas project that diverts manure previously used as fertiliser, leading to more chemical fertiliser use.

Methodologies deal with leakage by estimating it and deducting it from the credited amount, or by restricting where projects can be sited, such as only on land that was not used for food.

What to ask: what leakage deduction was applied, and whether the project area is designed to avoid displacing other activity.

How we use these tests on the desk

When a buyer asks us to shortlist lots, these three questions are where our review begins. The PDD should state the additionality argument and the leakage approach. The verification report should confirm that both were applied correctly in the period being issued. For land projects, the registry should show the buffer contribution next to the issuance. Our page on project documents buyers receive explains where to find each item.

No project scores perfectly on all three. A practical buyer weighs the residual risk against price, purpose and co-benefits, and documents the reasoning so an auditor can follow it later. If your policy rules out certain risks, for example avoided-deforestation credits or older renewable projects, list those exclusions when you post a requirement.

To compare lots across all three tests, browse by project type on the marketplace and send us the ones you want reviewed in detail.