A reserve forecast-based approach to determining credit collateral requirements in electricity markets

In electricity markets, credit collateral requirements for participants have traditionally been set based on historical price data that may not properly reflect future risks. A new predictive approach to determining credit risk is proposed in this paper. For any market that prices reserves in the real-time market, correlation exists between available reserve levels and real-time energy prices. This paper shows that it is possible to forecast hourly system-wide available reserves in a realistic system such as the Electric Reliability Council of Texas (ERCOT) market for up to a week ahead with high confidence. A credit collateral call can then be issued based on predicted system conditions due to the strong correlation between system-wide reserves and real-time energy prices. This in turn, would lower the risk of default for participants as it better reflects their actual risk. Case studies based on a representative ERCOT simulation show that potential scarcity conditions can be successfully identified as far as four days out.