An analysis of price spikes and deviations in the deregulated Turkish power market

Abstract

The successful operation of a real time market is related to the planning in the day ahead market. We analyze the day ahead and real time market data for the Turkish power market for the period 2012–2015 to classify price spikes and their causes. We also focus on the levels of deviation between the day ahead market values and the real time market values. We define price deviation and load deviation ratios to measure the level of deviation both in price and demand. The analysis for the load is based on load shedding and cycling values. We analyze the mean and standard deviation in market prices and we determine the price spike as a two sigma deviation from the mean value. It is shown that 60% of the price deviation ratios are in the range of (±20%), while 44% are in the range of (±10%) and 35% are in the range of (±5%). We also show that 56.9% of the spikes are due to problems in the generation of natural gas based power plants which affect the day ahead and real time prices. A total of 29.2% of the spikes are due to power plant and system failures that affect only real time prices. The share of high temperature based spikes is 13.9% which is a result of air conditioner usage. Keywords: Market clearing price, System marginal price, Price spike, Load shedding, Load cycling