Abstract:
Electricity price is crucial to the operations of many industrials and the occurrence of infrequent `price spikes' in the NZEM can have an adverse effect on these consumers. A better understanding of the mechanisms that drive prices can help industrials improve decision making to hedge against price spikes. The key to mitigate the consequences of high price spikes is to identify the causal mechanisms and modify consumption behaviour accordingly. We use the vSPD model which is a replica of Transpower's SPD market clearing engine to investigate the dispatch solutions that contained high prices. This research isolates a database of price spikes and identifies reserve requirements, frequency-keeping, HVDC link, outages and demand as the key factors that cause these high prices. We further investigate the demand factor by analysing an industrial, residential and rural demand pro le respectively and propose a demand model based on the results of the demand analysis. The demand is modelled in two parts; first, a combined seasonality model with temperature effects and interaction terms and second, an ARIMA process to model the serial correlation in the remaining demand residuals. We then present the out-of-sample forecasting performance of the model. Ensemble simulations are used to generate a distribution for the forecasted demand that can be used to predict the probability of a high price spike occurring.