Analysis of electricity price in competitive markets a study in Singapore market

Market participants have to deal with constantly changing electricity price in competitive electricity markets in the course of their everyday system operation. Although these prices in a particular market do exhibit some characteristics they are stochastic in nature and the prices cannot be predicted or forecasted for an individual hour. Yet, reasonable predictions or representations of such prices are very crucial for the operations of many market participants, such as Gencos. Many studies have adopted suitable statistical representation of the price in order to carry out system operation problems. However, justification for such models on the basis of real market prices is rarely found. This paper presents the initial findings of the study conducted to characterize the statistical behavior of electricity price in competitive markets. The published electricity prices from Singapore market have been used in these studies. The preliminary analysis is carried out using the price data from the first three years after the introduction deregulation in the market. It is used to identify the general behavior of the mean values of electricity price which show that it is possible to establish reasonable relationship between the market price and many system related factors. Recent historical data from the same market are used to investigate the statistical nature of the electricity price. Several statistical distribution functions are identified through the analyses of the historical price data which are likely to represent the stochastic nature of the electricity price. Graphical and statistical analyses are used for this purpose. Statistical analysis and testing have indicated that the behavior of the market price may resemble log-normal distribution. Further refinements in the representation of the market price are being studied.