A Newbie Ventures into the Weird World of the Safe Drinking Water Act
(An opinionated, noncomprehensive look at why the Environmental Protection Agency [EPA] did what it did)
Toward the end of February, a colleague at Stanford Linear Accelerator Center (SLAC) asked me a simple question: "What is the concentration limit for Be-7 in drinking water?" Though my health physics career started back when Jimmy Carter was president, I'm a newcomer to both SLAC and the environmental side of radiation safety. So once again, I had to say those tiresome words: "I don't know, but I'll find out and get back to you."
The EPA adopted a revised set of regulations called the National Primary Drinking Water Regulations for Radionuclides in late 2000. These regulations include the following [40 CFR 146.66(d)] requirements:
I had to re-read the second section a couple times. In 2000, EPA published regulations based on a dose-risk model described and published in 1963?! How could EPA possibly defend using NBS Handbook 69? Hadn't that model been replaced twice-over in most of the world (ICRP 26, ICRP 60)? I'd never accuse the U.S. of being a leader in adopting international standards, but even so, this case seemed extreme.
But the question I'd been asked was the current federal limit for Be-7 in drinking water, not why the EPA did what it did in 2000. Given SLAC's terrific libraries, finding NBS Handbook 69 was easy. Slogging through the text was less fun. Of course, there was an easier path to the answer. An appendix to the immense EPA document, Implementation Guide for Radionuclides, EPA 816-F-00-002, March 2002, available on-line, lists radionuclides and the maximum contaminant levels for compliance with the 4 mrem/year limit, based on data from NBS Handbook 69.
As usual, the simple question had a simple answer. As usual, I could have found it quickly, if I'd just known where to look.
Coincidentally, on this day, I started getting e-mail messages with subject lines like "Court Upholds EPA Drinking Water Standards for Radionuclides." Descriptions of court cases aren't at the top of my list when it comes to recreational reading, but then there was the puzzling fact of EPA's use of a 1963 NBS dose model . . . I started reading.
On February 25, 2003, the U.S. Court of Appeals ruled against the City of Waukesha, the Nuclear Energy Institute (NEI), and the National Mining Association in their petition for a review of the EPA's regulations related to the Safe Drinking Water Act (SDWA). And, predictably enough, the City of Waukesha and its fellow petitioners had not failed to question the EPA's use of NBS Handbook 69.
The write-up of the court's decision filled in a number of gaps. Despite the irritating backwardness of it all, one could begin to see reasons why EPA might want to use NBS Handbook 69 as the basis for most of their radionuclide concentration limits.
Since 1996, the SDWA has contained an "anti-backsliding" provision. Under this provision, it is permissible to revise the regulations adopted in support of the act as long as the revisions keep the "protection of the health of persons" at the same level. Alternatively, revisions may change the requirements to an even more protective level. But under this "anti-backsliding" provision, no revisions may be made that would cause the regulations to decrease the "protection of the health of persons."
The SDWA also requires EPA to do a cost-benefit analysis whenever setting a radionuclide limit (that is, a maximum contaminant level [MCL]) in the national drinking water regulations. A grandfather clause exempts the MCL values in effect on June 19, 1986, and a cost-benefit analysis is required only if it is proposed to alter a radionuclide MCL from its 1986 value. (Though there may be exceptions, the MCL values in effect in 1986 were generally based on the concentration that would give a standard man drinking 2 L/day an annual dose of 4 mrem, using the NBS Handbook 69 model from 1963.)
Between the anti-backsliding provision and the grandfathered MCL values, EPA obviously has a strong ongoing incentive to keep as many radionuclide MCLs unchanged as possible. And EPA did exactly that when they published the new radionuclide standards in late 2000. They added limits or other requirements for uranium and radium, but otherwise left existing limits unchanged.
Current risk models calculate a different and generally lower health risk per unit activity ingested than the health risk implied by the NBS Handbook 69 methodology. Remember the court case against EPA? The petitioners argued that in 2000, EPA had, by sticking with MCLs based on the 1963 model, effectively issued stricter standards and that EPA should be required to do a cost-benefit analysis to justify that action (or maybe inaction).
But the court didn't buy that. The grandfather clause is specific about exempting the MCLs in place as of June 1986 from the cost-benefit analysis requirement. The clause does not give any exceptions to that exemption. I could be wrong, but I'm guessing this grandfather clause would apply even if someone found clear evidence of a simple math error in the original calculation of an MCL.
The petitioners also argued that EPA failed to use "best available science," citing the fact that MCLs derived from NBS Handbook 69 meant the health risk from water at an EPA concentration limit wasn't consistent from radionuclide to radionuclide. Again, the court disagreed, noting that EPA had indeed reviewed risk results from more modern risk models. "Best available science" did not, the court said, dictate that EPA select drinking water MCLs to yield uniform risks and ". . . whether to insist on uniformity is a policy judgment that the SDWA leaves to the EPA's discretion."
The City of Waukesha and its fellow petitioners failed to achieve their objectives in challenging the EPA and its continued use of a long out-of-date dose-risk model for radionuclide ingestion. But they did succeed in generating a surprisingly entertaining record that seems to explain why EPA hasn't adopted limits based on modern risk models and why their position is a legally defensible one. And they did add a small piece to the education of this health physicist, struggling to understand the Safe Drinking Water Act.