Analysis of locational marginal prices in look-ahead economic dispatch

This paper studies the impact of introducing look-ahead dispatch on locational marginal prices in real-time electricity markets. Due to the high forecast error and inter-temporal variability of renewable resources (e.g., wind and solar), many system operators are in the process of upgrading their market clearing engine from static economic dispatch-based ones to dynamic look-ahead economic dispatch-based ones. Compared with conventional static dispatch that calculates the optimal dispatch for the current operating interval, look-ahead dispatch calculates the optimal dispatch solution over multiple future time intervals. Look-ahead dispatch can obtain cost saving by pre-ramping some of the dispatched generators so that the system can more economically match forecast changes in net load (i.e., system load - renewable power output). In addition to the cost saving benefit, system operators in RTOs claim that look-ahead dispatch results in lower price volatility than static dispatch. In this paper, we attempt to demystify this industry claim and provide counterexamples to this claim. Specifically, we investigate the relationship between price volatility and some key factors such as system-wide ramping capability, the number of time intervals in look-ahead dispatch horizon, load pattern and forecast uncertainty. The simulation results are illustrated with numerical examples in the IEEE 14-bus system.