Application of statistic methods in smart distribution grid concept

Abstract

Contemporary distribution power utilities (DPUs), which are faced with the electricity market, have to work on the edge of their efficiency. From technical point of view, SCADA systems were not sufficient to achieve the desired efficiency in DPU business. However, Smart Grid Concept (SGC) with integrated Distribution Management System (DMS), including SCADA system, can achieve that. The quality of SGC and DMS application depends on the quality and quantity of both realtime data provided by SCADA system and historical data. Among historical data, the most needed, but the least reliable, and for the final results most critical, are the data about electrical consumers' loads. Individual modeling of each consumer is practically unfeasible, i.e., economically and technically unjustifiable. Therefore, this paper offers a procedure for derivation of load profiles with required precision on the basis of relatively small number of measurements. In addition, the paper proves the normal distribution is the best assumption for statistical modeling of the loads of large number of consumers.