The Kariba REDD project reveals a “deep structural flaw” in Verra’s carbon credit programme
Analysis by CarbonPlan shows that Verra’s proposed solution to Kariba’s “excess credits” simply does not work.
At the beginning of October 2025, almost two years after starting an investigation into the project, Verra finally completed what it claims was an “in-depth review of the carbon accounting” of the project.
Verra estimated that it had issued 15.2 million “excess credits” to the project. These are “hot air” credits that do not represent any climate benefits. In fact they worsen the climate crisis because they have been sold to corporations that used them to offset continued fossil fuel emissions. In perhaps the most egregious case, TotalEnergies claimed to have delivered a shipment of “carbon neutral” liquefied natural gas, thanks in part to Kariba REDD credits.
Verra’s figure of 15.2 million is considerably less than carbon trading firm South Pole’s estimate of 27 million fake carbon credits issued by the project. But Verra hasn’t released the calculations that its estimate was based on — even to Carbon Green Investments, the project developer. And we only know about South Pole’s estimate because it was leaked to Follow the Money.
CarbonPlan, a US-based climate science and analysis organisation, released its critique of Verra’s review on 21 October 2025. Written by Grayson Badgley, Chris Allen, and Freya Chay, the critique describes Kariba as “the REDD+ offset project that has been synonymous with carbon market dysfunction”.
A “deep structural flaw”
In its review, Verra claims that the fact that millions of Kariba’s excess credits have been used to offset emissions doesn’t matter, because, as Verra explains in its review, Verra will “request compensation for the 15,220,520 excess credits” from Carbon Green Investments.
Verra states that 10.3 million of the excess credits have been used. A further 4.9 million excess credits from Kariba are unused within various accounts in Verra’s registry. Verra has “invited” the owners of these 4.9 million credits to voluntarily cancel them. South Pole subsequently announced the cancellation of 2.5 million Kariba credits.
According to Verra, Carbon Green Investments should buy the remaining excess credits (10.3 million carbon credits plus whatever is left of the 2.4 million credits after the voluntary cancellations) from other projects on Verra’s registry and cancel them.
Badgley, Allen, and Chay write,
Market participants should treat Verra’s proposed solution to Kariba’s excess credits with skepticism. Rather than putting Kariba to rest, it reveals a deep structural flaw in the largest registry of the global carbon market.
Verra did not set a deadline by which Carbon Green Investments or the owners of Kariba credits should respond. CarbonPlan notes that the Verra standard sets a deadline of 60 days for dealing with excess credits. But since Carbon Green Investments has withdrawn the project from Verra’s registry it’s not clear whether this standard applies.
CarbonPlan sums up Verra’s proposed solution to the issuance of 15.2 million fake carbon credits as follows:
In short, Verra announced that it plans to rely on voluntary corporate action to remedy a major failure of its promise that credits issued under its program are real.
Verra’s system doesn’t work
Verra’s system for dealing with excess crediting is to get the project developer to:
immediately cancel any excess credits it still owns;
replace the excess credits from future issuances; or
buy replacement credits on the carbon market.
If those options fail, Verra can impose sanctions on the project developer’s account.
Since Carbon Green Investments has withdrawn the project from the Verra registry, there will be no future issuances for Verra to recover the excess credits. Verra cannot sanction Carbon Green Investments’ account, because there is no such account. “In effect,” CarbonPlan writes, “Kariba ran up its tab and skipped town.”
CarbonPlan adds that,
The possibility of a project’s withdrawal from the Verra registry exposes a profound structural flaw with Verra’s rules for addressing excess crediting. Any project that generates excess credits faces a choice: remain enrolled in Verra’s program and repay the credits, or leave the program entirely.
Remaining in Verra’s programme would be very expensive — especially for projects, like Kariba, that have already sold large numbers of excess credits. There is a clear incentive to withdraw.
As CarbonPlan notes, buying replacement credits for at least 10.3 million credits would be expensive. Data-analysis firm Allied Offsets reports that the price for the lowest rated REDD credits is about US$3.60 per credit. That comes to a little over US$37 million.
CarbonPlan points out that according to its 2024 Annual Report, Verra has less than US$24 million in net assets.
Verra’s buffer pool is another possibility. It currently contains 72 million credits. Kariba would wipe out 14% of the buffer pool. But, as CarbonPlan writes,
[I]t doesn’t seem that Verra designed its buffer pool to absorb both the scale of excess crediting that might exist and the types of natural risks (e.g., wildfire, drought) that many of its projects face.
Faced with a bill it can’t pay, or a gaping hole in its buffer pool, Verra is, as CarbonPlan puts it, “now betting the environmental integrity of its offsets program on CGI’s willingness to voluntarily pay up”.
But Carbon Green Investments is extremely unlikely to pay up.
First, a bill of US$37 million is a very large amount of money for most companies to pay.
Second, as Carbon Green Investments noted in a letter to Verra dated 25 September 2025, Verra issued the carbon credits, not CGI:
The Kariba REDD+ credits have all been issued after five separate audits over the period, and we are not sure why Verra did not pick up any anomaly on the project then.
And third, Carbon Green Investments is a company incorporated in the tax and secrecy haven of Guernsey. It was founded by Steve Wentzel, who admitted to Heidi Blake, a journalist with The New Yorker, that the way he moved money from Guernsey to Zimbabwe was “illegal”.
Even if Carbon Green Investments did agree to hand over US$37 million to bail out Verra’s crumbling credibility, that would solve only one of Verra’s problems. “What about the next project that receives excess credits?” CarbonPlan asks. “Or the one after that?”
“Global carbon market is awash in excess credits”
CarbonPlan points out that Verra is likely to have to deal with very large numbers of excess credits over the next few years:
A mounting body of evidence suggests that the global carbon market is awash in excess credits. And Verra is likely responsible for a significant share. For example, Verra has over 100 million credits backed by wind energy projects in India that rely on rules developed under the Clean Development Mechanism. A recent detailed analysis suggests excess crediting for these types of projects is likely 50 percent or more. Other high profile studies have found systematic excess crediting under four tropical forest protocols administered by Verra, due to unrealistic assumptions about deforestation rates. Reporting by The Guardian, SourceMaterial, and Die Zeit based in part on this work estimates that roughly 90 percent of credits across these protocols were issued in error.
Verra has issued 450 million credits under these four tropical forest protocols. 290 million have already been used to “offset” fossil fuel emissions.
CarbonPlan concludes with a series of questions:
What will happen if Verra needs to acknowledge even a fraction of the excess credits that academics and journalists have identified?
Will those projects also withdraw?
Will they agree to voluntarily repay excess credits?
And perhaps most importantly, are offset buyers willing to bet on a future in which their credits aren’t backed by a guarantee from Verra, but instead rely on an ad hoc system of voluntary IOUs from developers with a financial incentive to walk away?
To the final question, CarbonPlan answers, “It seems unlikely.”