❗Clearing the air on DACCS costs in the EU ETS ❗

👀 Last week, I shed some light on the pretty shaky assumptions behind the BioCCS cost forecast in the EU ETS proposal. Today, let’s shift focus to the other CDR method eligible for integration into the ETS: DACCS.

💶 Unlike with BioCCS, I didn’t find the DACCS numbers intuitively off. In fact, I had a bunch of DACCS project developers tell me they were quite happy about those numbers, which might be even higher than what they could get on the voluntary carbon market (VCM).

🤷‍♂️ Still, I wanted to understand the logic behind them and dove deep. What I found is a black box with one source and no way to check it.

😮 The punchline is very simple: the entire cost projection is based on a single source, namely a McKinsey study from 2023. Unfortunately, McKinsey is not transparent about their data, besides stating that it drew on data from expert interviews and a cost reduction model. Remember, it was 2023.

🤬 In fact, there is absolutely zero information on what exact DAC technologies were included or what the major quantified cost drivers and assumptions are. We are basically just taking McKinsey’s word for it.

🌍 Further, as so often, these are global, not European numbers. For DACCS, where costs are really driven by energy, transport, and storage, Europe fares much worse than other geographies. Simply put: the cost of DACCS in Iceland or the United States are unlikely to be representative of the EU.

📈 There is another interesting dynamic: the BioCCS costs rely heavily (almost 50%) on high energy costs. DACCS relies on the exact opposite: falling energy costs. So the EU is assuming energy gets more / stays expensive for one CDR method and cheaper for the other. Both projections can't be right at the same time.

📉 In general, McKinsey assumes costs will fall 5.9-6.5% a year to 2035, then 1.3% a year to 2050. There is almost no explanation for it, although it does mention cleaner energy inputs could drive lower emissions and thus cheaper removals.

🔦 Finally, the McKinsey study had data for 2023, 2035, and 2050. The 2030 and 2040 numbers are extrapolated from this initial data by Ramboll and the EU Commission respectively. So the numbers you see for 2030 and 2040 aren't even McKinsey's - they are someone else's guess about McKinsey's guess.

📦 Summing up: I get it, we have to use what is out there. And there are many open questions. At least with BioCCS, we can openly discuss each variable that went into the modelling. In this case, it’s a black box.

🗣️ To all the DACCS experts out there: surely we can do better than this. What are the best sources of up to date industry data?

⁉️ Or are we facing a similar gap to BioCCS which the DACCS industry should fill asap?

Follow the conversation