Risk levels associated with wind generation penetration using Monte Carlo simulations

Abstract

An adequate installed capacity in generation system is an important criterion to know the reliability of power system in supplying the demand. The introduction of renewable resources of different types such as wind, solar, wave and tidal energy, has increased the generation capacity. However, these renewable resources are intermittent in nature which can significantly disrupt system reliability. This research is focusing on evaluating how intermittent wind generation contributes to the generation system reliability. By using the reliability indices, the performance of new generation capacities can be decided. Traditionally, a method called Capacity Outage Probability Table (COPT), probabilistic-based, was used to determine risk associated with generating units in a power system. But, by using the COPT method, system operator is unable to get a meaningful interpretation, since the loss of load probability (LOLP) is a tiny probability number. The objective of this research is to develop a new risk-based assessment method by using Monte Carlo simulation to quantify the risk amount in MWs. Value at Risk (VaR) is a method, used in finance as a risk measurement tool, which can summarize the expected maximum loss (or worst loss) over a target horizon within a given confidence level. Thus, in this thesis, an evaluation using the VaR is developed and presented extensively for the first time in the context of power system reliability assessment. A Wind Turbine Generation (WTG) model is constructed. A multi-state model is developed for modelling wind generation due to the randomness of wind speed. Probability models describing uncertainties in system generation and projected load demand are first constructed. By using Monte Carlo Simulation (MCS), Loss of Load Probability (LOLP) has been calculated by applying the proposed study on two practical test systems; RBTS (6 bus) and IEEE RTS (24 bus). LOLP is further used to quantify the exact amount of lost load and the Expected Demand Not Supplied (EDNS) due to contemplated uncertainties. The effect of WTG penetration is discussed in the case studies such as variable penetration, increasing load demand, number of multistate WTG, carbon emission and different wind distribution. The analysis shows that the capacity credit of a WTG is not equivalent to the conventional generator. An exact capacity of conventional generator need to be replaced with multiple WTGs unit in order to have same reliability. Besides, high penetration of WTGs in a power system shows a great risk for the system load to be unsupplied. On top of that, from the environmental point of view, tons of CO2 emissions can be avoided if conventional generation is replaced by WTGs. The study also shows that the evaluated VaR produces close results as compared with the EDNS in the same case studies. The benefit of using the VaR is that the system risk is simply reflected in a single quantity in MW. The proposed approach can be a great tool for power system operators in decision making concerning uncertainties arising in the generation side of the system.