2009 State of the Market Report
bid price, indicating that the load should be scheduled regardless of the day-ahead price. We are generally interested in comparing the net load cleared (defined as the physical load, plus virtual load minus virtual supply) in the day-ahead market as a percentage of the actual real-time load. This relationship affects commitment patterns and RSG costs because units are committed and scheduled in the day-ahead to satisfy the net load.
When day-ahead net load is significantly less than real-time load, particularly in the peak load hour of the day, the Midwest ISO will frequently commit peaking resources to satisfy the incremental increase in load. As shown later in this section, peaking resources often do not set real-time prices, even when these resources are effectively marginal. This can contribute to suboptimal real-time pricing and can result in inefficiencies because lower-cost units that could have been committed through the day-ahead market will be displaced by peaking resources committed in real-time.
Additionally, when significant quantities of generation are committed by participants or by the ISO after the day-ahead market, this additional supply will lower real-time prices and create an incentive for participants to schedule net load at less than 100 percent. The most common sources of additional supply increases real time are:
Supplemental commitments by the Midwest ISO made for reliability after the day-ahead
Self-commitments by market participants after the day-ahead market; and
Wind output that was under-scheduled in the day-ahead market.
To show the net load-scheduling patterns in the day-ahead market, Figure 19 compares the monthly peak-hour day-ahead scheduled load to actual load. We show the peak hour of each month because this is when the Midwest ISO is most likely to require additional generation.