Premium
A decision problem under uncertainty for nuclear power system development
Author(s) -
Pavelescu M.,
Ursu I.
Publication year - 1982
Publication title -
international journal of energy research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.808
H-Index - 95
eISSN - 1099-114X
pISSN - 0363-907X
DOI - 10.1002/er.4440060107
Subject(s) - ignorance , nuclear power , econometrics , preference , optimal decision , mathematical optimization , computer science , operations research , economics , mathematics , statistics , decision tree , ecology , philosophy , epistemology , artificial intelligence , biology
Decision making under uncertainty is a further step in comparison with decision under risk in the more realistic approach to decision problems concerning, for instance, nuclear power system development. In this paper the theory developed is, however, based in a great measure on that of risk preference. The theory of decision making under uncertainty is applied to a nuclear power system NPS consisting of PHWRs and PWRs integrated with LMFBRs. Nine development alternatives of the system which evolves for a period of 40 years are considered. The fast reactor integration is accomplished beginning in year 15 with a variable time delay so that for every alternative, six final states are possible. An econometric model of the system offers the cost price of annual energy generated by the system at the end of the given time interval for every possible state of any alternative. Further, the complete ignorance case is considered, resulting from the principle of insufficient reason, and the risk preference theory is applied. Then the partial ignorance case is taken into account and finally it it shown how we can infer a plausible a priori optimal probability distribution to have an optimal decision characterized by an optimal selected development alternative, for which a minimum certain equivalent of cost price of annual energy is realized with an accepted level of risk and a determined value of risk averter.