Will electric power consumers facing smart-grid enabled real time prices have the potential to accidentally destabilize the power grid and cause a blackout? A paper presented at a recent IEEE conference says it is a possibility. The surprising culprit? Too much price elasticity in the market demand function.
It is a surprising culprit because consumer demand for electricity is currently notoriously inelastic (that is to say, not responsive to changing prices) in the short run, in part due to the way standard regulatory rate structures end up with consumers being presented with relatively unchanging prices reflecting a longer-term average cost of production. Prices don’t change much, so consumers don’t watch prices much. But this price inelasticity of demand doesn’t mean the quantity of electricity consumers want to consumer is unchanging – consumers want more or less electricity throughout the day in response to ordinary household schedules and in response to outside temperatures and building heating and cooling demands. Consumer demand for power responds to a lot of things, but rarely to the price of power itself.
Because of the way the current grid is designed, the quantity of energy supplied and demanded must be balanced continuously. Therefore, the grid is typically operated to take the quantity of power demanded as a given and make whatever adjustments in the quantity supplied to maintain system balance. (In brief, because prices can’t do much work coordinating supply and demand in the short-run, all of the coordination must be done by adjusting quantities. Grid operators can typically control suppliers but not consumers, so quantity-based supply side adjustment does most of the work of keeping the market balanced.)
The authors, three engineers at MIT, worry that if too many consumers facing real time prices pick similar high price points at which to cycle off appliances (or low prices as which to charge electric vehicles), that the market demand function will acquire highly price elastic segments in which quantity demanded will suddenly drop off (or spike up) at rates faster than the supply side can safely accommodate. Therefore, a blackout risk. To counter this possible risk, the authors suggest diversifying price signals sent to consumers, or employing hourly instead of 5-minute price signals, or using rolling-average prices to consumers rather than location-specific current marginal price. They admit their safeguards would hamper the efficiency of market results, the efficiency loss essentially the price paid to mitigate the possibility of a price-responsive demand shock to the system.
In my view, the idea of having so many real-time price-aware consumers responding in the market remains so far-fetched that I’m not willing to worry about that so many of them will coordinate their home energy management systems on the same price points and unwittingly bring down the system.
And well before this possibility of too-much consumer responsiveness comes about, I suspect most RTOs will be paying suppliers for ramping capability and charging consumers for using it in ways that will enable sufficient short-run system responsiveness. So I’m not ready to worry now about this problem, and don’t think that I’ll need to worry about it later, either.