A probabilistic study of electric power systems considering the uncertainties of renewable generation and loads

This paper presents an approach based upon probabilistic load flow to study the effect of wind generation and variation in loads on an electric power grid. A probabilistic approach was adopted due to the uncertainty associated with wind generation and power consumption. A new method of selecting random variables from historical data is implemented by finding the times of day where power generated or consumed is above or below a given threshold, and correlating the data to incorporate the effects of sites that are geographically close to each other. This is in contrast to other methods where data have been separated into monthly, weekday and weekend intervals. By breaking up the loads into these smaller intervals, a narrower (lower variance) distribution is created from which random load and wind speed data are selected, and in turn has the potential to create narrower distributions for results. The approach is demonstrated on a modified version of the IEEE 24-bus reliability test system. Historical wind speed data was obtained from NREL archives, and load consumption data was acquired from two towns in New Mexico.