Multi-attribute decision theory and energy systems analysis

This paper reports on the application of new developments in multi-attribute utility theory (MAUT) to energy systems analysis. In applying MAUT, the analyst has to extract decision maker's utility functions for individual attributes and preference weights for aggregating the individual utility functions. Such information is obtained from lottery-based inquiries (Keeney and Raiffa, 1976). While a decision maker can provide consistent and reasonable responses for single attribute lotteries, he/she often is unable or unwilling to provide precise multiattribute tradeoffs. It is possible, however, to obtain partial or imprecise information that is either represented by linear inequalities among the preference weights, or by semantic fuzzy responses. The author develops multi-attribute decision models and solution techniques for cases where partial and imprecise preference information is available. The methodologies outlined are then applied to two problems of energy systems analysis: siting of nuclear waste disposal facilities, and policies to reduce sulfur dioxide emissions from coal-fired utilities to mitigate the impact of acid deposition