Dose Optimization -Reply

[Charles Willis <CAW@nrc.gov>]

Kevin,

I would be most interested in the responses you get.  I do not believe that I have
ever seen an optimization analysis as suggested by ICRP-22.  For the most part,
the NRC analyses consider two or three options, offer estimates of the costs and
benefits of each, and leave the choice to management.  In many instances, the
choice is simply a yes-no decision.  About the closest thing to optimization that
comes to mind is the EPA analysis of the uranium fuel cycle back in '73 ( see
EPA-520/9-73-003-B, -C & -D plus a later supplement EPA-520/4-76-017);
interestingly, the actual cost-benefit curve was not published but was included in an
internal document by Jim Martin and Allen Richardson entitled "Statement of
Considerations: Environmental Radiation Standards for the Uranium Fuel Cycle"
September 12, 1973.  The internal document was said to be available upon request,
but that was a long time ago.  

The problems with optimization in the real world include vast uncertainties in both
costs and benefits, the limited number of  feasible options and the fact that the
relationship between costs and benefits generally is not well represented by a
smooth, single valued, and continuous function.

It is an interesting question and I wish you well.  One of our graduate students is
writing his thesis on the subject, from another perspective.  Any solid information
on optimization that you might send would be appreciated.

Charlie Willis
caw@nrc.gov