On the Co-Optimisation of Reserve Markets: With application to New Zealand

Abstract

Electricity markets bring together buyers and sellers of electricity in order to dispatch the electricity grid. To accomplish this, market exists where generation companies compete by submitting offers to a System Operator who clears the market, subject to security constraints. To ensure these security constraints are satisfied, reserves are procured through a co-optimised reserve market. This reserve market is solved simultaneously with the energy market to determine the least overall cost. This approach, which has long been implemented in markets such as New Zealand (Alvey et al., 1998) and Singapore (Lu and Gan, 2005), is also becoming popular throughout the large North American markets (PJM, 2014). The simultaneous procurement of reserve introduces new mechanisms through which reserve constraints may influence the final energy price. These constraints may limit the dispatch of risk setting generation or transmission assets. In New Zealand, more than 10,000 trading periods over a five year period were identified as exhibiting symptoms of reserve constraints. These constraints also influence the optimal decision criteria for participants. A Supply Function Equilibrium model has been presented to investigate competition between suppliers located at either end of a reserve constrained transmission line. The optimal decision for these suppliers is to withhold reserve in order to limit energy market competition. This behaviour was observed during the winter months of 2012 in New Zealand. Consumers may participate in the reserve markets by offering Interruptible Load. Reserve market participation by Consumers modifies incentives regarding the optimal response to highly priced trading periods which are reserve constrained. A kNN model has been developed to identify trading periods ex ante where reserve constraints cause high energy prices. An Interruptible Load consumer following a curtailment strategy based upon this model exhibited increased profits compared to alternative methods. This price taking approach has been extended using a numerical stochastic optimisation model to determine the optimal combined consumption and reserve offer using simulations and dynamic programming. A consumer using this model showed profitable results which approached the theoretical perfect response strategy. This work constitutes the first cohesive assessment of the effects of co-optimised reserve markets on pricing mechanisms, generators, and Interruptible Load consumers.