Decarbonizing the electricity sector requires a massive shift away from fossil-red generation, towards a diverse suite of newer technologies. Demand response(DR) is one of these promising technologies. It encompasses a range of tech-niques that adjust demand levels in response to system conditions. It couldmake the grid more exible and reduce peaks in demand, potentially helpingintegrate renewables and operate the grid more eciently. DR can have oper-ating characteristics that dier from those of traditional generators. To realizethe potential value of DR to a decarbonizing grid, we must understand howthese unique properties aect its system-wide value. This study contributesto our understanding by characterizing the relative value of dierent possibleproperties of DR, so that market participants can focus their eorts on the mostuseful types of DR resources. We use a two-stage stochastic unit-commitmentmodel, with ERCOT as our test system. Features examined include advancenotication requirements, restrictions on when DR is available, the number ofstartups, the number of hours of operation, and the amount of energy shed. Re-sults suggest that inexpensive DR that requires advance notication may stillbe quite valuable to the grid, and these limitations aect the value of DR lessthan other usage restrictions. Availability for early afternoon ramps and peaksis key for realizing reductions in system costs and ramp rates among thermalgeneration, and may be more important than the ability to respond to the realtime market. The understanding gained from this study can guide the devel-opment of new DR products that provide higher system-wide value and betterconsumer satisfaction.
展开▼