Policy commentary

The Cooling Water Intake Structures Rule: An Experiment with More Flexible Regulation

May 11, 2009 | Winston Harrington

Typically, technology-based (TB) regulation involves the Environmental Protection Agency (EPA) identifying a technology that meets some conception of “best” performance (as defined in legislation) and then establishing a standard that achieves this level of performance. The costs, expected environmental improvements, or the value of those improvements are not taken into account in setting the standard. In one recent case, however, EPA took an alternative approach, calling instead for minimization of adverse environmental impacts, which gives regulated plants more flexibility than usually permitted.

The case involved Section 316(b) of the Clean Water Act, which regulates water withdrawals for cooling purposes and the accompanying return flows. A steam-electric plant, for example, may draw millions of gallons per day and the intake flows may cause mortality among crustaceans, fish, and even diving birds, by pinning them against screens (impingement) or sweeping them into the cooling system (entrainment). In fact, these processes can affect entire aquatic ecosystems by killing eggs, juveniles, and small organisms at the bottom of the food chain. Moreover, the water itself discharged from cooling systems can further affect aquatic ecology, by eliminating species sensitive to heat and favoring more heat-tolerant species that may not be natural to the local area.

As required under the statute, EPA identified—as the “best technology available”—closed-cycle cooling, which minimizes thermal releases, impingement, and entrainment through the use of cooling towers that draw much less water. Following executive orders mandating regulatory impact analyses (RIAs) on major rulemaking processes however, the Office of Management and Budget recommended removing this technology requirement and suggested a compliance option based on a plant-specific comparison of benefits and costs. The final rule, issued in 2004, was a complicated but flexible approach to TB regulation, involving several components.

Among the most important of those were a baseline against which performance was to be measured, namely the estimated mortality of marine organisms at a facility with “once-through” cooling and no controls on impingement or entrainment; a performance standard requiring both an 80 to 95 percent reduction in impingement mortality (compared with the baseline) and a 60 to 90 percent reduction in mortality from entrainment; and the identification of two designated technologies that EPA felt would meet the performance standards: a closed-cycle cooling system and a special screen designed to minimize mortality from withdrawals. Unlike most TB performance standards, these standards were based not on the capabilities of the technology but directly on the estimated effect of the technologies on the natural environment. EPA’s rule would allow a steam plant even further departures from the usual practice in TB regulation, including investment in ecological restoration measures that would, on net, reduce the mortality involved in water withdrawals and return flows, or a demonstration that it was entitled to a site-specific determination of compliance technology because the cost of adopting the designated technology would be significantly greater than the costs estimated in the rule or the expected benefits at the site.


These measures were both innovative and controversial. EPA recognized that the costs and biological effectiveness of abatement technologies for cooling water intake systems depended on local configurations and conditions, and on the local aquatic environment and its species. The desire to bring environmental effects into the rulemaking led the agency to an unusual definition of performance standards. Customarily, EPA defines the performance standard in terms of the performance of the technology itself, such as percentage reduction in emissions compared with no treatment. For this rule, the standard was written in terms of the effects on natural organisms—percentage reduction in mortality from impingement and percentage reduction in entrainment.

Other features were equally novel. While the use of compensatory restoration had for years been an option for developers seeking permits from the Corps of Engineers to alter wetland environments, this was among the first attempts to use it in more traditional regulation. And the site-specific benefit–cost analyses had a rough parallel in the “footprint” approach to the CAFE regulations for light trucks promulgated in 2005. Those regulations set manufacturer-specific standards based on the expected cost to manufacturers of modifying each model in their truck fleet. This feature exceeds even the requirement for regulatory impact analyses—that the total benefits of a rule justify the total costs. Indeed, the CWIS rule considered the potential of not just total but marginal benefit–cost comparisons. This is much closer to economists’ conception of how benefit and cost information should be used.

For EPA, estimating expected costs and benefits was complicated by the site specificity of cooling water intake systems. On the cost side, EPA was uncertain whether the lowest-cost compliance alternative would actually meet the performance standards at particular plants; thus, a more costly technology had to be assumed.

On the benefits side, EPA had to determine the physical and biological effects of the regulation, quantify those changes, and then estimate (in dollars) the value of those changes. Some categories of benefits resisted the final valuation step, and some could not even be quantified.

Ultimately, the only benefits valued were the benefits to commercial and recreational fishing. Costs of the rule exceeded benefits by a factor of about five. As this ratio makes clear, the nonmonetized benefits did receive consideration in the analysis, but necessarily were left out of the benefit–cost comparison.

The Legal Challenge

The cooling water intake system rule was challenged by states, environmental groups, and the utility industry. The individual appeals were merged into a single case (Riverkeeper, Inc. et al. v. EPA), which was decided in January 2007.

The U.S. Court of Appeals for the Second Circuit ruled that EPA’s use of benefit–cost analysis was an incorrect reading of the statute. The “best technology available” performance standard precluded the balancing of benefits and costs, it said; the only legitimate question here was whether the cost of meeting the performance standard was something that industry could reasonably bear (and the court observed that several plants had already installed the designated technology).

The court then called on EPA to tighten up the ranges in the performance standards so that a plant could not get away with minimum performance: the plant should do its best, not the minimum. The court also rejected the use of restoration as a compliance alternative, ruling that restoration was not “minimization” but impermissible “compensation” for environmental impacts, and in any event, restoration was not “technology.” Finally, the court remanded the site-specific compliance alternatives—the cost-cost test and the cost-benefit test. Thus, most of the rulemaking innovations were either rejected outright or remanded to EPA for clarification.

EPA subsequently suspended its cooling water intake rule and has not yet issued revisions. Meanwhile, industry petitioners have appealed the decision to the Supreme Court.

The story of this regulation illustrates not only the legal difficulty of building flexibility and cost-benefit consideration into technology-based rules, but also the conceptual difficulty of basing regulatory decisions on the likely consequences if the knowledge base for determining those consequences is deficient. This is not to say that a conventional technology-based standard would perform any better. It is difficult to determine whether by limiting the flexibility of plants in meeting environmental standards, the court improved matters or made them worse. 




Further Reading:


Harrington, Winston, Lisa Heinzerling, and Richard D. Morgenstern, editors. 2009. Reforming Regulatory Impact Analysis. Washington, DC: Resources for the Future.

U.S. Court of Appeals, 2nd Cir. 2007. Riverkeeper et al. v. U.S. EPA. January 25.

Newbold, Stephen C., and Rich Iovanna. 2007a. Ecological Effects of Density-Independent Mortality: Application to Cooling-Water Withdrawals. Ecological Applications 17(2): 390–406.

Newbold, Stephen C., and Rich Iovanna. 2007b. Population Level Impacts of Cooling Water Withdrawals on Harvested Fish Stocks. Environmental Science and Technology 41: 2108–14.


Views expressed are those of the author.
RFF does not take institutional positions on legislative or policy questions.

The views expressed do not necessarily represent those of U.S. Environmental Protection Agency or ot