Application of wind generation capacity credits in the Great Britain and Irish systems

Abstract

The concept of capacity credit is widely used to quantify the contribution of renewable technologies to securing demand. This may be quantified in a number of ways; this paper recommends the use of Effective Load Carrying Capability (ELCC, the additional demand which the new generation can support without increasing system risk), with system risk being measured using Loss of Load Expectation (LOLE, this is calculated through direct use of historic time series for demand and wind load factor). The key benefit of this approach is that it automatically incorporates the available statistical information on the relationship between wind availability and demand during the hours of very high demand which are most relevant in assessing system adequacy risk. The underlying assumptions are discussed in detail, and a comparison is made with alternative calculation approaches; a theme running through the paper is the need to consider the assumptions carefully when presenting or interpreting risk assessment results. A range of applications of capacity credits from Great Britain and Ireland are presented; this includes presentation of effective plant margin, ensuring that the optimal plant mix secures peak demand in economic projection models, and the Irish capacity payments system. Finally, new results comparing capacity credit results from the Great Britain and Irish systems using the same wind data are presented. This allows the various factors which influence capacity credit results to be identified clearly. It is well known that increasing the wind load factor or demand level typically increases the calculated capacity credit, while increasing the installed wind capacity typically decreases its capacity credit (as a percentage of rated capacity). The new results also show that the width of the probability distribution for available conventional generating capacity, relative to the peak demand level, also has a strong influence on the results. This emphasises further that detailed understanding of risk model structures is vitally important in practical application.