A tonne of fossil CO2 warms the climate for centuries. A tonne stored in a young forest can return to the atmosphere in one dry summer. Permanence is the question of how long the carbon behind a credit stays stored, and what happens if it does not.
Permanence is how long the carbon stored or protected by a project stays out of the atmosphere. Forest and soil carbon lasts decades to centuries and can be reversed; biochar, rock weathering and geological storage hold carbon for centuries to millennia. This article explains how permanence differs per project type, how standards insure against reversal with buffer pools, where those buffers have fallen short, and what to check before you buy.
Why permanence matters
A carbon credit claims that a tonne was kept out of the atmosphere because of a project. If the tonne comes back later, through a forest fire, a drained peatland or a leaking storage site, the claim stops being true.
That matters more as claims get stricter. SBTi’s Corporate Net-Zero Standard V2.0 requires larger companies to support carbon removals from 2035 and asks for a growing share of long-lived removals, defined as “capable of retaining carbon for centuries to millennia”. Short-lived removals, storing carbon for “decades to centuries”, count too, but not for the long-lived share (SBTi carbon removal requirements). The standard deliberately does not attach a number of years to either category.
Permanence per project type
Project type | Where the carbon is | Typical storage horizon | Main reversal risks |
|---|---|---|---|
Forest protection (REDD+) | standing forest | as long as the forest stands | fire, illegal logging, policy change |
Reforestation and agroforestry | trees and soil | decades to centuries | fire, drought, pests, harvesting |
Soil carbon | agricultural soils | years to decades, practice-dependent | change of farming practice, drought |
Mangroves and wetlands | coastal sediment and vegetation | decades to centuries | cyclones, erosion, sea-level rise, clearing |
Biochar | stable carbon in soil or materials | centuries, depending on production and use | low for well-made biochar |
Enhanced rock weathering | carbonates and bicarbonates | millennia | very low once reacted |
BECCS and direct air capture with storage | CO2 in geological formations | millennia | very low, site integrity and monitoring |
Removal registries make durability explicit. Puro.earth has required a minimum of 100 years of storage since 2023, up from 50, and labels credits as 100+, 200+ or 1,000+ years (Puro.earth). The EU’s CRCF uses durations too: permanent removals, carbon farming with a minimum of five years, and carbon stored in products for at least 35 years (CRCF certified carbon credits).
Permanence matters for avoidance credits as well. A forest protection credit only holds if the forest is still standing years later.
How standards handle reversal risk: buffer pools
Most nature-based credits are protected by a buffer pool, a shared reserve that works like self-insurance.
Each project assesses its risk of reversal. At Verra this is done with the AFOLU Non-Permanence Risk Tool, which scores internal, external and natural risks.
Based on that score, the project deposits a share of its credits in a pooled buffer account. Higher risk means a larger share.
Buffer credits are not regular credits: they cannot be sold or retired by buyers.
When carbon is lost, the registry cancels buffer credits equal to the loss, so the total number of credits in circulation still matches the carbon that is stored.
The pool is shared across all projects of the standard. A fire in one project is covered by the contributions of all of them, which is why the health of the pool, not only the contribution of your project, matters.
Since December 2025, Verra has also been piloting alternatives to the pooled buffer: an insurance policy covering at least the credits issued while it runs, with carriers such as CarbonPool, Kita and Artio, or a fund financed from part of the sale proceeds (Verra).
When buffers fall short
Buffer pools only work if contributions match real risk. The clearest test so far comes from California’s compliance forest offset programme. A peer-reviewed study found that wildfires had used up nearly one-fifth of the programme’s total buffer pool in less than a decade, equivalent to at least 95% of the contribution meant to cover fire risk for 100 years (Badgley et al., Frontiers in Forests and Global Change, 2022).
For voluntary buyers, the study shows two things: risk models built on historical fire and drought data underestimate a warming climate, and a buffer is a pooled promise that depends on every other project in it. Forest credits remain useful, provided you know what stands behind them. Standards are revising their risk tools for that reason.
Tip from Bernard de Wit, Founder of Regreener: “For any forest or soil project, ask two numbers: the project’s buffer contribution and how long the project commits to monitoring after its last issuance. A low contribution in a fire-prone region, or monitoring that stops when the credits stop, tells you the reversal risk has been passed on to you.”
Permanence and your claims
Permanence decides what you can say. A reforestation credit can back a contribution claim today, and counts as a short-lived removal under SBTi. Calling it permanent, or using it to neutralise long-lived emissions at net zero, overstates it.
Three rules of thumb we use with clients:
Match storage to purpose. Use long-lived removals for the long-lived share of SBTi’s removal requirement and for neutralising residual CO2. Use nature-based credits for contribution claims and co-benefits.
Name the storage type in the claim. “We financed 500 tonnes of reforestation in Indonesia” is accurate. “We permanently removed 500 tonnes” is not.
Diversify. Spreading credits over project types, regions and standards limits the impact of a single reversal. SparkOptimus combined a reforestation project in Indonesia with biochar in Ghana for that reason.
What to check before you buy
Storage type and horizon: where the carbon sits and for how long, in the project documents.
Buffer contribution and risk score: the percentage the project deposits and why.
Monitoring commitment: how long the project monitors after the crediting period ends.
Reversal history: any past losses and how they were compensated.
Contract terms: whether your supplier replaces credits if a reversal affects them.
Durability label: for removals, the stated storage horizon (for example 100+ or 1,000+ years) and how it is demonstrated.
Our free credit quality checker covers permanence alongside additionality and baseline, and how to assess carbon credit quality puts it in the full checklist. For the full picture of removal types, see carbon dioxide removal explained.
“Permanence is where nature-based and engineered removals differ most, and where buyers most often mix them up. Both have a place. The mistake is paying for a forest and describing it as a vault.”
Bernard de Wit, Founder of Regreener
Check the permanence of a project
Our free credit quality checker scores a project on permanence, additionality and baseline, and gives you the questions to ask your supplier.
