Maintaining efficient operation of facilities that have been in operation in excess of their standard service life is among the most significant and complex problems faced in modern electric power systems. The urgency of this problem stems from the systematically growing fraction of such facilities, which has reached as high as 60% by now. The difficulties of solving it are caused by the necessity to develop methods and algorithms for quantitatively evaluating the operative performance efficiency (OPE). A method and an algorithm are proposed for quantitatively estimating the integral OPE indicator for unique facilities, i.e., facilities that do not have analogs for the specified combination of attribute varieties. By using the proposed approach, it becomes possible to obtain not only error-free estimates of technical and economic indicators (TEIs) but also, what is most important, a physical interpretation of the integral indicator. The multidimensional nature of the monthly average values of TEIs and nonrandom nature of samples from the totality of TEIs are factors that pose serious limitations to the application of the classic hypothesis tests. A new criterion that successfully overcomes these obstacles is developed. The critical values of integral indicators appearing in this criterion are determined by simulating possible realizations of the integral indicators. A smaller risk of obtaining an erroneous solution on the maintenance of unique facilities is achieved, due to which reliable methodical support for the enterprise management staff is ensured. The sequence of calculations carried out for a gas- and oil-fired 2400-MW condensing thermal power plant (CTPP) is illustrated. To make the data transformation manipulations more compact with concurrently retaining the possibility to compare them with the results from a quantitative estimation of the OPEs for facilities of the same type, the calculations are carried out only for certain levelized monthly average TEI values of the CTPP power unit boiler plants. By using the monthly average values of unique old power facilities, so-called oldtechs (UPOTs), it becomes possible to perform operative monitoring of their variation with time, and the changeover from actual to normalized values makes it possible to estimate the change in the UPOT technical state from the average wear and the degree of its maladjustment.