یک روش احتمالاتی برای ارزیابی بهره برداری از وضعیت جغرافیایی نیروگاه های برق آبی

A procedure to evaluate efficiency in the exploitation of the geographic situation of hydroelectric plants is developed here. It is based on the probabilistic composition of criteria. A comparison of 80 plants is carried out, with volume of water flow at the location and transmission rates paid measuring the potential of the geographic situation and installed power and assured energy measuring the employment of such potential. An analysis based on a new index of quality of approximation and a new measure of importance derived from Shapley value is used to select the criteria that enter a second stage of the efficiency evaluation.

The main goal of this study is to build a methodology for evaluating the efficiency of hydroelectric plants (HEPs) in energy generation. The main tool employed is the probabilistic composition of evaluations according to multiple criteria. The measures of preference according to each criterion are calculated in terms of probability of reaching an excellence frontier, as proposed by Sant’Anna and Sant’Anna (2001). Four attributes are considered. The suitability of the geographic situation is evaluated by two independent criteria measuring the importance, respectively, of the water resources exploited and of the proximity to the markets supplied. On the other hand, the capacity of production of the plant is accessed by measuring two indicators, installed power and assured energy. The comparison based on variables representing the physical production structure provides an orientation to the expansion of the electric energy generation system, either by expanding active HEP or by building new HEP. It also enables the evaluation of the effect on the economic viability of the enterprises of the form of measuring variables such as transmission tariff and assured energy. The situation attributes are measured in an inverted form. The availability of water resources is measured by the inverted inflow, that is, in terms of time per volume of water. The proximity of the market is measured in terms of transmission tariff, that is, by the ratio of total cost charged for the transmission of energy from the plant to the consumer market to power transmitted. Thus, as desirable in an efficiency analysis, efficiency increases as the values of the variables measuring inputs decrease. The opposite occurs with the production attributes. The higher the values of installed power or assured energy, the higher the output generated and the efficiency in the exploitation of the available resources. A second stage is based on the application of rough sets theory (RST), of Pawlak, 1982 and Pawlak, 1991, to investigate the importance of each variable. To measure such importance, a probabilistic variant to the index of quality of approximation of Pawlak is employed. This measure is based on the computation of probabilities of concordance between classifications according to different sets of attributes. Thus, instead of disregarding the cases of indiscernibility according to the decision variable, this new measure considers all alternatives of indiscernibility or dominance. Following Greco et al., 1998 and Greco et al., 2001, the evaluation of the importance of each attribute and of the interaction between attributes employs the computation of a variant to Shapley values (Shapley, 1953). In the following section, the probabilistic approach is discussed. Section 3 presents the data of 80 HEPs operating in the Brazilian territory and preliminary results of their probabilistic comparison. In Section 4 the new indices of the quality of approximation are presented. In Section 5 they are applied to the selection of criteria to compare the HEP. Final comments are presented in Section 6.