ML19256A346
| ML19256A346 | |
| Person / Time | |
|---|---|
| Issue date: | 12/27/1978 |
| From: | Minogue R NRC OFFICE OF STANDARDS DEVELOPMENT |
| To: | Mills W ENVIRONMENTAL PROTECTION AGENCY |
| References | |
| NUDOCS 7901050054 | |
| Download: ML19256A346 (2) | |
Text
t w,N g *8 UNITED STATES v
l' &
+1 NUCLEAR REGULATORY COMMISSION h
[
~
WASHING TON, D. C. 205LL y
e 1,n 3D.
-/d a
N
'f
(
DEC L i l',
Dr. William Mills, Acting Deputy Assistant Administeator for Radiation Programs Office of nadiation Program (ANR-458)
U.S. Environmental Protection Agency 401 M Street, S. W.
L!ashington, D. C. 20460
Dear Dr. Mills:
We have been in close contact with your staff since last August, in discussions of the HLW standards which EPA and NRC are scheduled to issue soon.
I believe it would be useful at this time for me to set down some of our ideas on the specific structure and implementation of regulatory standards.
There are three important regulatory elements for HLW disposal:
(1) the EPA environmental radiation standard for HLW, (2) the NRC regulation for disposal of HLW, and (3) the NRC review and licensing process by which a specific repository is authorized.
The NRC elements must-be based on the EPA standard or, if they precede it, must be brought into conformity with it when it is promulgated. The NRC regulation and licensing action must implement the specific require-ments of the EPA standard. This close relationship between the EPA standard and the NRC regulation and licensing actions makes us especially sensitive to the structure of the EPA standard and its explicit require-ments.
We feel strongly that a deterministic method should be used to regulate nuclear facilities. We are aware that you are considering a substan-tially different type, a probabilistic standard which requires quanti-tative risk assessment. Based on our understanding of the virtues and the weaknesses of quantitative risk assessment, we are convinced that it can and should be used to provide insight on the quality and effec-tiveness of HLW disposal regulation, but it cannot be the explicit basis of the regulation which requires rigorous satisfaction because:
1.
The analytical techniques are complex a' d there are many areas in them which are the subject of wide ficagreement in the technical community.
/
e Dr. William Mills DEC " 71973 2.
These quantitative techniques are greatly dependent on the quantity and quality of the data upon which they are based.
3.
In most cases where one confronts the analysis of low probability events, statistical uncertainties make rigorous use of the quantitative results impossible.
Standards for protecting public health and safety can be expressed as limiting levels of physically meaningful parameters, such as materials released, radiation dose, health effects (a ueterministic standard),
or as a probability of certain parameter levels being reached or exceeded (a probabilistic standard).
In the first instance, the implementor is required to demonstrate compliance with physical limits on consequences, taking into account the effect of important potentially disruptive events such as floods, faulting, etc.
Compliance in the second instance hinges on demonstration of the probability of occurrence (as well as the consequences, i.e., risk) of those events. Although there are no laws of science which preclude the possibility of performing such risk assessments on the long-term isolation of radio-active waste, the capability to perform such risk assessmenta in a manner sufficiently rigorous to serve as the primary basis for licensing decision does not now exist and there is no assurance that it will (or can) be developed in the next several years.
__ In the past two weeks we have been working on possible forms for a deterministic EPA standard which would be consistent with your analyses and with our need to implement its specific requirements.
I suggest
_ _ that we meet soon to discuss this matter further.
Sincerely,
&h Robert B. Minogue, Director Office of Standards Development
,