A Comparative Study of Capacity Market Demand Curve Designs Considering Risk-Averse Market Participants

While capacity remuneration mechanisms are emerging around the globe, a variety of capacity demand curves is being used. Poorly tuned capacity demand curves can, however, lead to expensive over- or underprocurement of capacity. This work provides a methodology to directly compare different capacity demand curve design methodologies, accurately reflecting the net cost of new entry in the definition of the demand curves. To this end, a long-run equilibrium with energy-only and capacity markets with risk-averse market participants is considered. We find that sloped capacity demand curves, setting the net cost of new entry at the reliability target, and inelastic capacity demands, that are set to the reliability target, perform ideally assuming perfect foresight. Considering erroneous assumptions on the value-of-lost-load and risk-aversion, the results obtained using these sloped capacity demand curves are more sensitive to uncertainties than those using inelastic capacity targets. In contrast, conservatively tuned marginal reliability impact-based curves perform suboptimal assuming perfect foresight, but are less sensitive to erroneous parameterization.