The U.S. is preparing to spend hundreds of billions of dollars to deploy technologies that could help us reach climate goals — both through grants under 2021’s Infrastructure Investment and Jobs Act, also known as the Bipartisan Infrastructure Law, and subsidies under 2022’s Inflation Reduction Act. But whether that spending will actually reduce climate pollution is unclear in some cases, notably for carbon capture and storage, or CCS, 45Q tax credits and for hydrogen 45V tax credits.

A recent environmental assessment, or EA, issued by the U.S. Department of Energy for a CCS project at a coal-fired power plant in North Dakota, Project Tundra at the Milton R. Young station, includes numerous egregious errors that call into question the department’s ability to figure this out — which could have major implications for the success of much bigger programs.

The core issue is this: if we emit GHGs while doing something to reduce GHGs, we need to make sure that the reductions outweigh the new emissions — preferably by a lot. We’d also typically like to know if we’re creating new problems with our new activity — think slave labor in mineral supply chains, or water contamination from chemical use.

One common way to evaluate these issues is to use life cycle assessment, or LCA, and related methods, which seek to look at all the different kinds of impacts that an activity might cause, over the entire value chain, or life cycle – sometimes called “cradle to grave.” (When we only look at one issue, like GHGs, we sometimes call this “life cycle analysis.”)

In the case of GHGs, these evaluations are exceptionally important for processes that use a lot of energy, especially fossil energy. And the inputs aren’t always straightforward to identify, or estimate. For example, if you make hydrogen from electricity and increase overall demand, which is partly served by fossil fuels, how do you decide which electricity emissions belong to the hydrogen? If a carbon capture unit is added to a power plant and the investment allows the plant to keep operating for another 20 years, how should you think about emissions from the plant that would not have happened if the plant had closed instead?

As a life cycle assessment scholar and the former deputy assistant secretary of carbon management at DOE, where I dealt with hydrogen and carbon capture, I know how nuanced these questions are. They’re difficult even for LCA experts, particularly when they involve questions about what would have happened otherwise, which is fundamentally speculative. And they matter — one of my recent analyses suggests that doing carbon capture poorly could lead to $4 trillion in taxpayer expenditures on subsidies that increase GHG emissions.

Analysis of the hydrogen tax credit suggest subsidies could promote hydrogen production that is dirtier than the unabated, fossil-based systems produce now. Plus, we’ve already been down this road with corn ethanol, which set off huge debates about GHG and energy benefits that were never fully resolved in policy. This is why I am so concerned that the recent draft EA for Project Tundra includes a GHG life cycle analysis that doesn’t even get the basics right, let alone the complex parts.

Project Tundra, which has already received substantial funding from the DOE — and which the current assistant secretary for fossil energy and carbon management advocated for prior to joining the federal government — is a proposal to put carbon capture on a coal plant that has reached its end of life, which would likely extend operations into the 2040s.

My analysis suggests that adding CCS could be worth $5-6 billion for this power plant, while increasing emissions by 6 to 8 million tonnes of CO2-equivalent relative to closing the plant at its end of life. Demonstrating that the project would not lead to this outcome — and in fact, would reduce emissions — is nonnegotiable for obtaining federal funding, in my view. But the analysis published, and sent out for public comment by Sept. 19, does not even meet the basic standards for an LCA, and falls far short of what the DOE requires of applicants for carbon capture demonstration projects on coal plants, which Tundra likely applied for — a massive waste of time for those interested in the project, and waiting to participate. Not only does the draft GHG analysis contain numerous errors so obvious that even a non-expert should have noticed them immediately, not to mention the LCA preparers themselves and the numerous reviewers who sign off on a draft EA before it is published, but the analysis suggests that the project would emit more than three tonnes of CO2 equivalent per tonne of CO2 stored, and still recommends moving forward — a deeply alarming sign for a project whose entire purpose is to reduce GHG emissions.

The errors in the draft EA’s life cycle analysis of GHGs are numerous and often structural, but for illustration, the analysis overestimates one value by a factor of 23,500 without so much as a comment that the value seems unusually high. It also claims that the capture unit would require about three times as much coal-fired power as the power plant is capable of producing, but that this would not lead to any additional GHG emissions. These are basic competency errors that suggest either total inattention to this crucial analysis, or a startling lack of capability.

Evaluating whether technologies like hydrogen and carbon capture provide more benefit than harm is a critical activity for the U.S. on our decarbonization journey. The risks are large, and serious. The analyses we need are nuanced and require careful attention: this cannot be a “check the box exercise.” Not taking this seriously risks potentially trillions of dollars and billions of tonnes of GHG emissions, not to mention the trust and goodwill of the American public, which is reasonably skeptical of these potentially critically important technologies. The Project Tundra EA is shocking, with scary implications for doing this right. We must do better, and we must demand better.