The paper starts with the task of assessing different qualities (security, efficiency, matching ecological standards, etc.) of the complex units operating in electricity sector (power plants, current generators, electricity transmission lines, generating companies, etc.). Then the task considered as the problem of multi criteria assessment of the multidimensional objects, which are identified with a finite bundles of initial parameters. The paper gives brief exposition of different models applied for assessing the initial parameters of the operational units in electricity sector. Special attention is given to the methods for measuring objects' characteristics, quantified with magnitudes of different scales, and their consecutive standard normalization, which further allows for consistent description of the impacts they have upon the quality of the objects considered. Due to such normalization all separate quality parameters have the values within the interval between "0" (least preferable level of quality) and "1" (maximum feasible level of quality). The paper articulates the "incomparability problem" for the objects under assessment. It typically occurs when one has to make the objects' comparison with the full range of their parameters. In many cases it happens that one of the objects is superior to another in regard to some of the parameters, but it is inferior in regard to the rest of them. To break such multi criteria "incomparability" we offer to use the Method of Aggregate Indices (MAI) which is well developed by now and widely used as an effective tool of analysis in different subject areas. The paper studies main functional forms used for the synthesis (aggregation) of the separate quality indices into a single aggregate appraisal of the quality. The aggregate measure takes into account not only the magnitudes of the indices considered, but it also depends on how valuable they are for the object on the whole. Thus, the role of the "weight coefficients", which allow to assess the importance of each separate quality index, is discussed. The issues of selection of the most appropriate form of aggregating function are given brief consideration. These functions should reflect marginal substitution effects for each pair of separate quality indices.