Recent writing on verification, methodology risk, and 45Q.
Written by whoever on the team had something worth saying, not on an editorial calendar.
On the difference between a baseline and a guess.
Every verification argument eventually comes down to a counterfactual: what would this hectare, this reef, this facility site have done without the intervention. Get the baseline wrong and everything built on top of it — the measurement, the audit, the credit itself — is a precisely calculated wrong answer. We've walked away from more mandates over baseline disputes than over any other disagreement, including price. A client who wants to negotiate the counterfactual, rather than model it honestly and accept whatever it says, isn't a client we can make defensible in front of an accredited validator, and increasingly, validators are the ones asking.
Why the buffer pool exists, and why it's not just overhead.
New clients ask, almost every time, why a share of every issued credit gets set aside in a shared buffer pool instead of being sold immediately. The honest answer is that the alternative was tried, informally, in the early years of voluntary markets, and it produced exactly the failure mode you'd predict: projects that looked sequestered on paper for exactly as long as it took to sell the credits, then quietly reversed. The buffer pool isn't bureaucratic caution. It's the mechanism that makes the difference between paying for carbon storage and paying for a convincing-looking claim about carbon storage, and those two things only look identical until the first wildfire, flood, or failed well.
What happens when a methodology changes mid-project.
A methodology revision is not a footnote. We had a client midway through a four-year mangrove restoration when a soil-carbon measurement methodology changed enough to alter their projected issuance by nearly a fifth — not because anything about the mangroves changed, but because the counting rules did. Methodology risk is real, underpriced by clients who've never lived through a revision, and one of the few things in this business we genuinely cannot make disappear, only plan around. Anyone advising you otherwise is selling certainty the registries themselves don't have yet.
The plants financed on bad verification rarely fail where you'd expect.
When a direct air capture facility built on shaky verification eventually comes apart, it's almost never the capture technology that fails first. The chemistry mostly works. What fails is the paperwork chain connecting a specific tonne captured on a specific date to a specific credit issued against it — a chain that turns out to have a gap nobody checked closely while the money was still coming in. We spend more time auditing the boring middle of that chain than the exciting parts at either end, because the boring middle is where almost everything that eventually goes wrong actually goes wrong.
Have a project where one of these has already happened to you? Get in touch — that's usually how the first conversation starts.