Coherency Reduction in the EPRI Stability Program

The purpose of this paper is to present computational experience of the cumulant method of generation system produc¬ tion cost calculations This method has been recommended for use in analysis of generation systems with individual generators sized at less than 10% of total generation capacity and with forced outage rates greater than 10% However, smaller utilities have generation systems that do not have these characteristics This phenomenon can cause errors in the production cost results using the cumulant method This paper examines and compares equivalent load curves calculated by the cumulant method and conventional convolution Recommendations are made on generator state models and the number of cumulants used in the calculations in order to reduce cumulant method errors The results of these recommendations show that production cost calculations using the cumulant method are entirely acceptable for screening generation expansion plans