LINEARIZED TECHNIQUE AND MONTE CARLO SIMULATION FOR STOCHASTIC POWER FLOW STUDIES OF ELECTRIC POWER SYSTEMS.

This paper describes two methods for the power flow study of an electric power system which is subject to random load and generation schedules. The first method is a Monte Carlo simulation in which load and generation data are stochastically modeled and pseudorandom multivariate bus power injections are generated digitally. The method is characterized by long computation times in order to get line loading data within reasonable confidence limits. The second method presented is a closed-form power flow method in which the nonlinear power flow equations are linearized, so that the line loading becomes a linear vector-matrix function of multivariate random variables of presumably known statistics. The method is characterized by rapid computation times but potentially unacceptable accuracy. The Monte Carlo simulation and linearized power flow technique are compared with regard to their effectiveness in calculating the statistics of line power flow from the statistics of the bus demand and generation.