ASSESSING THE TMDL APPROACH TO WATER QUALITY MANAGEMENT
KENNETH H. RECKHOW
Professor, Duke University
Chair, Committee To Assess The Scientific Basis Of The Total Maximum Daily Load
Approach To Water Pollution Reduction
Water Science And Technology Board
Division On Earth And Life Studies
National Research Council
Subcommittee On Water Resources And Environment
Committee On Transportation And Infrastructure
U.S. House Of Representatives
JUNE 28, 2001
Good morning Mr. Chairman, Representative DeFazio, and members of the Committee.
I am Dr. Kenneth Reckhow, Chair of the National Research Council (NRC) Committee to Assess the Scientific Basis of the Total Maximum Daily Load Approach to Water Pollution Reduction and a professor at Duke University. I am accompanied by the director of the NRC’s Water Science and Technology Board, Stephen Parker, the NRC study director for the project, Laura Ehlers, and Dr. Leonard Shabman, who is a professor at Virginia Polytechnic Institute and State University and who worked closely with the Committee and assisted with the writing of the final report.
As you know, in October, Congress suspended EPA’s implementation of the TMDL rules until further information could be gathered on several aspects of the program. In particular, Congress requested that the National Research Council examine the scientific basis of the TMDL program. The National Research Council is the operating arm of the AffiliateMarketIngtools of Sciences, AffiliateMarketIngtools of Engineering, and the Institute of Medicine, chartered by Congress in 1863 to advise the government on matters of science and technology. The U.S. Environmental Protection Agency sponsored this study of the TMDL program. I am here today to report on some of the principal findings and recommendations in our report.
Our central message can be summarized in a short sentence—we have the scientific capability to identify the polluted waters and develop plans for their cleanup. Underlying this general message are specific key points:
(1) Current scientific knowledge and techniques are sufficient to proceed with the TMDL program.
(2) Using current science, we recommend several changes in how EPA and the States conduct the TMDL program that can be immediately implemented, leading to improvements in the TMDL program. Many of these recommendations simply involve changes in the techniques used in the TMDL process, not development of new techniques.
(3) In the long run, science advances through a process of research and improved understanding. We can improve the science used to support the TMDL program over time, and we recommend research activities to do that.
The first and third points are worth restating by analogy with the field of medicine. We would all agree that today’s medical knowledge is sufficient for the successful practice of medicine. Still, we fund medical research, knowing that this will improve the practice of medicine, leading to improvements in health care over time. As much as we might hope that medical practice will eventually be perfect—all diseases correctly diagnosed and cured—in reality we know that medical uncertainties will always remain. Those uncertainties, however, do not prevent us from benefiting greatly as a result of current medical knowledge.
In a similar manner, our NRC Committee strongly believes that water quality knowledge is sufficient for the practice of water quality assessment and management within the context of the TMDL program. As with medical science, water quality science will never be perfectly known; there will always be uncertainties, but those uncertainties do not prevent us from making good decisions based on existing scientific knowledge. To improve practice over time, we continue to engage in medical research; likewise, we should continue to fund and conduct water research, with the expectation that such research will improve the TMDL process over time.
Let me now elaborate on my second point, which was that there are several changes that EPA and the States could immediately implement to improve the science underlying the TMDL program. I provide two brief examples:
(1) As part of the TMDL, EPA requires a “margin of safety” factor that is protective of water quality. This factor is intended to reflect the uncertainty in the TMDL forecast; however, EPA allows the margin of safety to be arbitrarily chosen, resulting in a variable level of protection. We can do better; techniques exist now to base the margin of safety on an explicit calculation of uncertainty.
(2) A TMDL is required for waterbodies listed by the states as in violation of a water quality standard. Unfortunately, many states have not required well-designed monitoring programs with statistical hypothesis testing to diagnosis standard violations. We know how to do this now, but states are just not using the best science in many situations. Using the medical analogy, this issue is equivalent to using outdated techniques for diagnosing a disease.
This second example addresses a vital feature of the TMDL program—the listing, or diagnosis, of waterbodies in violation of a water quality standard. Clearly, what we want to do is correctly identify all waterbodies in need of a TMDL, and we believe that the statistical hypothesis testing procedure will add the necessary scientific rigor to that identification task. At this point, the States already have identified, and listed (on the 303d list), waterbodies they believe to be in violation of water quality standards and hence in need of a TMDL. However, due to limitations in state water quality monitoring and assessment programs, many of these lists appear to be in error. That is, we believe that some waterbodies have been identified as “in violation” when in fact standards are being met, while other waterbodies that truly are in violation of a standard have not yet been identified. To correct this problem, we propose that EPA approve a two-list process in which States would create a “preliminary list” of waterbodies suspected of being in violation of a water quality standard. They would then have a limited period of time to undertake the necessary monitoring and statistical hypothesis testing to refine the preliminary listing decision. Those waterbodies thus determined to be in violation would then go on an “action list” (the 303d list) and require a TMDL to meet the standard and achieve the designated use.
In conclusion, it is important to recognize the tremendous variability among the estimated 40,000 projected TMDLs. Some are accompanied by a great deal of data and existing knowledge; for others, very little may be presently known. Some have self-evident solutions, while for others the problem is exceedingly complex and solutions are unclear. Again, consider the medical analogy—there are many different diseases and a great variety of treatments; thus doctors must be prudent in diagnosis, make appropriate treatment choices, and then monitor patients until health is achieved. Likewise, the variety of water quality problems under the TMDL program means that the States must strive to correctly diagnose standard violations, they must make judicious choices on TMDL treatments, and they must continue to monitor the waterbody until the designated use is achieved. Both the medical doctor and the TMDL scientist face uncertainty in their tasks. We believe that the NRC report outlines a strategy for using the best science to move ahead with the TMDL program.
Mr. Chairman, I would like to thank you and this Subcommittee for your interest in this vital program to improve the quality of our nation’s waters. My colleagues and I would be glad to answer any questions.