EPA Maximum Contaminant Level for Tritium cont'd

DOE-HDBK-1129-99

the World Health Organization and Canadian levels are approximately 200,000 pCi/L (189,000

pCi/L for Canada and 210,000 pCi/L for the World Health Organization. These are based on a

value of 10 percent of the public dose limits as calculated in accordance with ICRP-60

recommendations. [12]

EPA calculated a maximum contaminant level (MCL) for tritium of 20,000 pCi/L in 1976 based on

values for worker exposure contained in Handbook 69, "Maximum Permissible Concentrations of

Radionuclides in Air and Water for Occupational Exposure," published by the National Bureau of

Standards in 1959 and amended in 1963. This calculation assumes that an extra dose resulting

from organically bound tritium, equal to a 20 percent increase over that determined for worker

exposure dose in Handbook 69, should be factored into the MCL. The resulting MCL (20,000

pCi/L) is based on not exceeding 4 mrem/year on a total body basis, and assumes a daily rate of

ingestion of 2L of water.

In July 1991, EPA proposed revised MCLs for radionuclides based on its dosimetric model

RADRISK, which included many of the concepts and assumptions included in the effective dose

equivalent method (ICRP 30). In the proposed regulation, EPA determined that the MCL for tritium

was 60,900 pCi/L. DOE provided comments to EPA suggesting that the appropriate value for the

tritium MCL should be 80,000 pCi/L, based on dose to soft body tissues. This value is consistent

with the Derived Concentration Guide (DCG) value in DOE Order 5400.5, "Radiation Protection of

the Public and the Environment." Apparently, EPA used its RADRISK model to calculate the

60,900 pCi/L value, which differed from the models and parameters recommended by ICRP 30.

The proposed regulation has not been finalized. Projections from EPA are that a re-proposed rule

is expected no sooner than December 2000.

In early 1997, EPA prepared a Direct Final Rule that would have updated the methodology used in

the 1976 final NPDWR. This methodology would be consistent with the method contained in ICRP

30 and Federal Guidance Report 11; i.e., the effective dose equivalent conceptrather than the

"total body or organ dose equivalent" method used in 1976. This direct final rule would have

retained the 4 mrem/year standard, which is based on daily ingestion of 2L of water and

corresponds to a lifetime cancer risk of 10-4. In April 1997, however, EPA determined that the Safe

Drinking Water Act Amendments of 1996 contained a provision that would preclude EPA from

promulgating a standard that results in increased risk, in the context of revising an existing

standard. Although the direct final rule was described as a technical modification to the

methodology used to calculate a regulatory compliance level, and that the basic standard (4

mrem/year) was unchanged, the concentration levels for many of the radionuclides would

increase, thereby increasing risk. The level for tritium is the most dramatic example, since the

change in methodology would increase the regulatory compliance level from 20,000 pCi/L to

86,000 pCi/L. To date, EPA has not promulgated this direct final rule.

It is the current understanding of the Office of Environment Safety and Health (EH's) that EPA is

considering a risk-based approach to setting NPDWRs for radionuclides, rather than the dose-

based approach currently followed. No decisions have been made. EH is not aware of any plan

for providing an advance or draft rule for review and comment on this approach. If a risk-based

approach were to be adopted, it is likely that the MCL for tritium would be on the order of 30,000

pCi/L rather than the 60,900 pCi/L proposed in 1991. A summary of the rule and proposals is

given in Table 3-1.