How is the carbon removal ecosystem best supported? — Learnings from Microsoft’s 2022 CDR purchases

Robert Höglund
5 min readMar 13, 2022


Microsoft is one of the largest purchasers of carbon removal credits globally, with a target to remove as much carbon as it emits in the company’s lifetime. They are very transparent about their endeavors, publishing learnings and detailed data on their emissions and purchases of carbon removal. Like Stripe, they also publish all the proposals they received in their carbon removal RFP. This allows us to draw learnings on the wider carbon removal space.

In 2022 Microsoft purchased 1.5 million tons of carbon removal for an average of $19.40 per ton totaling $29 million. 99% came from land-based removals with low durability, including reforestation, improved forest management, and soil carbon sequestration. 1% of the tons (and probably 5-6% of the budget) went to medium-high durability removal projects, which this blog post focuses on.

Specifically, Microsoft bought 10600 tons of biochar from six companies, 200 tons of bio-oil from Charm, and 100 tons of CO2 from biogas stored in concrete from Neustark. (Now added to the database I maintain on CDR purchases.) Last year Microsoft purchased 1900 tonnes of biochar, 1400 tons of Direct air capture with storage, and 2000 tonnes of bio-oil.

Microsoft FY22 project submissions dashboard.

You can see signs that supply was a significant constraint for Microsoft’s purchases, and this is one of the main learnings they highlight in their report. In many cases, Microsoft bought all of the supply offered by their chosen providers. It seems like Carbofex, for example, got a smaller order than Freres biochar (ACT Commodities Inc) because they offered much fewer tons.

How should the CDR ecosystem best be supported?

Microsoft is prioritizing near-term delivery over projects offering volumes post-2023. No project offering zero tons 2022 was chosen. Unlike Stripe and Shopify, Microsoft is not taking the lead in kick-starting new technologies with their carbon removal purchases (more on other forms of support further down). New methods most often do not have tons to sell delivered today.

Pre-purchasing carbon removal tons offers a structured way of supporting new CDR methods and companies, but it’s not a panacea either. Carbon removal providers can sell the tons from their first facility before it’s built as a way to finance it. But those tons are often very few, often just a couple of hundred. And it would be precarious to start selling tons from facility number two before the first one is built. Companies that don’t already have their methods completely worked out can’t presell tons either and would benefit from R&D grants instead.

Microsoft knows this and has also made a $100 million donation to Breakthrough Energy Catalyst that will be used for grants on direct air capture, green hydrogen, long-duration energy storage, and sustainable aviation fuel. Given that the funding is non-dilutive grants, I think it’s likely this donation is more effective in terms of climate impact than Microsoft’s carbon removal purchases, and not only because it’s larger in monetary terms. A few other corporate supporters are taking a similar route. Chan Zuckerberg Initiative (CZI) gave a $21 million grant to research institutions working on CDR as well as making carbon removal purchases. Stripe also has mixed purchases with pure R&D grants.

Wider learnings about the CDR market

When looking at Microsoft’s complete list of received proposals, it is interesting to note that they got fewer applicants this year than the first one. The same is the case for Stripe, whose fall 2021 call got less than half the submissions than in the two rounds before. I counted to 26 companies applying with durable carbon removal solutions to Microsoft this year, just over half (14) of them being biochar companies. At the same time, we see more and more CDR startups emerge, so I would interpret the low number of applications as a sign that few startups are ready to start selling tons rather than that the starting field is shrinking.

Using a request for proposals have considerable benefits, but not all companies will apply. Many promising CDR companies have not sent in applications to Microsoft and could not be evaluated for their support. (With the Milkywire Climate Transformation Fund that I work with, we have done outreach to carbon removal companies that we identified to be interesting instead of doing an RFP. Both approaches have pros and cons, and a mixed approach could be interesting.)

Moving forward

In their request for proposals, Microsoft also included questions on future supply from the carbon removal providers, which gives insights into how fast the market could grow. The 26 companies with durable CDR solutions estimated having around 1 million tons on the market in 2025 (one biochar project stood for 25% of that volume). For what it’s worth, 15 million tons in supply for 2030 is what the answers sums up to, but all companies did not give estimates so far out. Even if this is just a subset of CDR providers, it is still sobering and in line with the growth estimates I previously made.

One of the critical issues Microsoft raise in their paper is the need for clear definition and standards. I agree common definitions are very much needed today, and standards are an absolute requirement for scale when we move from innovator-buyers to the early adopters and early majority of buyers. However, I think we need to be realistic about what standards can achieve at this point. It can open some “safe pathways” for support, for example, by determining what type of biochar projects are effective and sustainable to scale. But for most CDR pathways, we do not have all the information needed to judge if a particular method or company using that method is effective from a system perspective and has sufficiently low social or ecological risk. Any standard created today will need to change. We must be careful not to develop standards that permanently exclude support to high-impact options due to insufficient information being available today or locking in methods that later turn out to be less effective than we thought.

Carbon removal methods should be responsibly scaled up in a way that leads to more knowledge about their effectiveness and risk. Corporate support is crucial to making this happen, but early backers need to be transparent about what their support entails and carefully consider and understand the role they are choosing in supporting the CDR ecosystem.

We are still piecing together what large-scale CDR solutions should look like. Photo by Kevin Jarrett on Unsplash


  • I am glad Microsoft also highlighted the need for responsible reforestation and is following the Kew golden rules for forestation.
  • One of the biochar companies Microsoft purchased from is Climate Robotics. They have a super exciting solution for biochar with the potential of lowering costs. They will collect the biomass on the field, convert it to biochar and then incorporate it into the soil, all without leaving the field!
  • Interestingly, Microsoft did not renew its purchase from NCX / SilviaTerra of 1-year ton removals. As previously communicated, I am very critical of equating short-term removals with permanent storage.



Robert Höglund

Advising companies and organizations on carbon removal and climate.