Development of a composite system reliability evaluation program

This report describes in detail the mathematical models and techniques used in the development of a new composite reliability evaluation tool, CREAM. The methodology is based on Monte Carlo sampling of equipment availability (generators and circuits, including weather-dependent and common-cause or dependent outages) and load level (including system load or bus load vector scenario). The analysis of a sampled scenario is based on the linearized power flow model. Remedial actions and load curtailment are carried out by an efficient linear programming routine. Besides calculating adequacy indices (LOLP and EPNS) both at system and bus levels, the program produces the sensitivity of the reliability indices with respect to incremental reinforcements of generation capacities and circuit capacities. The required number of Monte Carlo samples is reduced through a new variance reduction technique, which uses analytical models (such a single-bus generation capacity reliability model) as a regression variable.'' The component of reliability indices explained'' by these variables is then factored out'' of the Monte Carlo sampling scheme. The program is demonstrated in case studies with two 500-bus utility-derived test systems. 12 refs., 1 fig., 7 tabs.