This study extends the agent-based 8-Zone ISO-NE Test System to include wind turbine agents, each characterized by location, physical type, and an output curve mapping local wind speed into wind power output. Increases in wind power penetration (WPP) are modeled as build-outs of investment queues for planned wind turbine installations. The extended system is used to study the effects of increasing WPP under both stochastic and deterministic day-ahead market (DAM) formulations for security-constrained unit commitment (SCUC).

For each tested WPP, the expected cost saving resulting from a switch from deterministic to stochastic DAM SCUC is found to display a U-shaped variation as the reserve requirement (RR) for deterministic DAM SCUC is successively increased. Moreover, the RR level resulting in the lowest expected cost saving systematically increases with increases in WPP.