Joint analysis of power system reliability and market price considering the uncertainties of load forecasts

Abstract

Reliability evaluation plays an important role in power system planning and operation. In electricity market, the uncertainties of market price and load forecasting will also greatly affect system reliability. This paper presents a novel model and algorithm for joint analysis of power system reliability and market price considering the uncertainties of load forecasts. After a process of reliability evaluation and market simulation, the results include system reliability indices and the probability distribution function of price can be achieved. If the forecasting load is a certain value, the probability distribution of the market price, as well as Expected Value and Standard Deviation, can be obtained directly. Furthermore, a stochastic load model is established to describe the uncertainty of future system load. Taking the uncertainty of load forecasting into consideration, we can suppose the forecasting load to be a random variable. Under different load levels, the market price has different probability distributions. So, the full distribution is the weighted sum of each distribution. In this algorithm, units' uncertain bids and output are incorporated together as well as the probability of forecasting load. In order to show the efficiency of the proposed models and algorithms, two numerical samples are studied in details in this paper. A simple system which contains 4 units is analyzed firstly. The calculations of system reliability indices considering uncertain load forecasting are demonstrated within this example. Consequently, IEEE Reliability Test System (RTS) composed of 32 units is utilized to simulate the real electricity market. The probability distribution of market price under different load levels are calculated and then depicted by figures. Through the study, the expectant unit output and its income are also analyzed and some instructive conclusions are yielded which are useful for the participants of the market. Based on these results, it is also possible to evaluate the risk involved by generating units when system load forecasting is uncertain. The sample studies indicate that the model established is efficient.