In September, the Environmental Protection Agency proposed a rule that would require all newly constructed coal power plants in the United States to install carbon capture and storage (CCS) technology. CCS is a new technology that separates the primary greenhouse gas (carbon dioxide) from the power plant’s exhaust and pipes it to underground reservoirs for storage. This proposal represents the first major step in President Obama’s plan to regulate greenhouse gas emissions from power plants, which are this country’s largest emissions source.

At first blush, this policy sounds like a meaningful effort to reduce carbon dioxide emissions from coal power plants, which alone represent about a third of all U.S. emissions. From a legal standpoint, however, the EPA’s choice of CCS represents a gamble that could end up significantly delaying emissions reductions.

To understand why, it is necessary to consider three points. First, the future of coal is already bleak. New coal plant construction will likely wane in the coming years regardless of what the EPA does with its rule. Second, the policy is legally risky because CCS is a young and costly technology that the EPA probably does not have the authority to require. And third, if a court overturns the EPA’s CCS mandate, the decision will significantly delay the next, more meaningful step in the president’s plan: the regulation of existing power plants.

Requiring CCS will not lower emissions / The EPA itself projects that requiring CCS “will result in negligible [carbon dioxide] emissions changes.” That is because, even without a CCS mandate, it is unlikely that new coal power plants will be built in the United States for at least the next decade, given market conditions. If anything is built, it will be plants that burn natural gas (which are much cleaner) because natural gas prices are at historic lows and forecasted to stay that way (thanks to, among other things, the revolutionary extraction technique known as fracking). Natural gas prices would need to double or even triple for new coal plants to become viable again, with or without CCS. As such, what the EPA requires for new coal plants won’t matter much from an emissions perspective.

Legally risky / Under the Clean Air Act, the EPA can only require technology that is both adequately demonstrated and can be installed at a reasonable cost. Yet a court could easily find that CCS meets neither of those criteria. In its proposal, the EPA points to a few small-scale pilot projects as proof that CCS is adequately demonstrated. However, there are no large-scale coal power plants using CCS anywhere in the world today. About two dozen large-scale CCS projects are in the planning or very early construction phase worldwide, but none are operational. In fact, Norway just recently abandoned its plan to install CCS at a large oil refinery, citing doubts over its commercial viability.

The EPA’s position on cost is also tenuous. The agency has admitted that adding CCS to a new coal power plant would increase the plant’s cost by about 80 percent. To put that figure in context, the most expensive environmental control technology the EPA has ever required for a new coal power plant is a sulfur dioxide scrubber, which increases the plant’s cost by about 20 percent.

Indeed, the EPA’s estimate of an 80 percent cost increase from CCS might be too low. One of the largest U.S. utilities, Southern Company, is working with the Department of Energy to build the first full-scale coal power plant with CCS. The demonstration project, located in Mississippi, is only about 75 percent complete and the project costs have already ballooned to about $5 billion—twice the amount Southern Company originally planned to spend.

Legal challenges to the EPA’s proposal must be filed in the U.S. Court of Appeals for the District of Columbia Circuit, a court that has overturned numerous EPA rules in the last few years. The court will give the EPA some deference in determining which technologies are “adequately demonstrated” and cost effective, but it has never before found that such costs are reasonable or that technology in the construction phase is adequately demonstrated. The seminal D.C. Circuit decision on this issue, Essex Chemical Corporation v. Ruckelshaus, provides that to be “adequately demonstrated” a technology must “have been shown to be reasonably reliable, reasonably efficient, and … [not] exorbitantly costly.” Given the current conservative leaning of the D.C. Circuit, it seems more likely than not that the EPA’s CCS requirement will be overturned.

Delaying existing standards / Here’s why all this matters: Section 111 of the Clean Air Act provides that the EPA cannot regulate existing power plants unless new power plants are also regulated. In other words, lawful new power plant standards are a prerequisite to regulating existing power plants.

Pursuant to President Obama’s plan, the existing power plant rule is to be proposed in 2014 and finalized in 2015. The EPA is slated to finalize its new power plant proposal with the CCS requirement in September 2014. That means the D.C. Circuit’s decision on CCS will come in 2015 or 2016, the exact timeframe that the EPA is planning to start regulating existing plants. If the court overturns the CCS mandate for new plants, existing-plant regulation will have to halt until the EPA replaces the new plant rule or until the U.S. Supreme Court decides the case. Either situation would delay the existing-plant standards for years.

Betting on CCS is not worth the risk. From a climate perspective, the existing-plant standards are what matter. Brad Plumer of the Washington Post analyzed various legal studies—including my own—and noted that the EPA’s existing-plant standards could reduce power plant emissions by anywhere from 1 to 26 percent. I think the rules will end up on the lower end of that range, at about a 5 percent reduction. Regardless of the stringency, however, there is no dispute that existing-plant standards are where meaningful reductions could occur. The EPA’s overzealousness in regulating the construction of future power plants means that these reductions may be significantly delayed.

The underlying rationale for the EPA’s CCS policy is unclear. The EPA has already announced that it won’t be requiring CCS for existing power plants, and so the policy is clearly not an effort to stand tough on new power plants in order to adequately regulate existing plants. It is possible President Obama is using the rule to make an international statement that all new coal power plants built worldwide should use CCS, but this seems like a poor reason to risk meaningful U.S. emissions reductions.

What should the EPA do? There are no other add-on control technologies like CCS. The EPA’s only other option is to require that new coal power plants be built using the most efficient design, which is what the agency should do in its final rule. This position would be much more legally defensible because such plants are already operating in the United States today. Moreover, in the unlikely event that anyone builds a new coal power plant, at least the plant would be more efficient and lower-emitting than most of the plants currently operating. This change would probably have a negligible impact on emissions, and the EPA could always revise the new power plant rule and require CCS later down the line. Following this path would protect the agency’s authority to regulate existing plants and would give the EPA and the industry more time to see if CCS really works and at what cost.