How can markets set efficient prices in peak demand hours?

Prices during peak demand hours are critical for the economic efficiency of electricity markets.  During peak times, quick-starting peaking units are usually started and dispatched.  In the tightest hours, operators may take emergency actions to increase the real-time supply or decrease demand.  Efficient pricing in peak hours requires that the costs of these actions be reflected in the real-time market prices.

Real-Time Pricing of Combustion Turbines.  Peaking generating resources are generally natural gas-fired combustion turbines known as “gas turbines”.  The cost of starting and operating these units in any instance reflects the real incremental cost of supplying load and should be reflected in marginal prices.  Most real-time market dispatch models can only set prices based on the offers of online units.  Historically, most markets did not have pricing rules that allow the costs of starting and dispatching high-cost peaking resources to be reflected in real-time energy prices.  As a result, such units often require guarantee payments to cover their costs that must be allocated to loads.  Moreover, the real-time energy pricing algorithm sets price at the highest-cost unit to serve load, but excludes units that are considered non-dispatchable, like quick-start units.  Therefore, energy price is understated because it does not reflect the cost of starting and operating theses units to meet energy demand.

Well-designed quick-start pricing models allow real-time prices to include the cost of committing and running peaking units when they are the incremental source of energy.  Such models have been implemented by NYISO, MISO, and ISO-NE.  This helps ensure that the marginal prices reflect the full cost of serving load.  Such pricing rules help improve key incentives for efficient long-term and short-term efficiency.  They reduce reliance on uplift payments and send more efficient economic signals to guide commitment decisions.

Such pricing models are needed when gas turbines are utilized because gas turbines constitute most of the resources at the high-priced end of the supply curve.  When they do not set prices, the prices are often set by a much lower-cost unit.  If the portfolio of higher-cost resources in the real-time market included a mixture of flexible and inflexible units, this price-setting issue would not be as large a concern because one could expect high-cost flexible units to set prices when the inflexible units could not.

Real-Time Pricing of Demand Response and Emergency Actions.  This approach can be adapted for pricing demand response and other out-of-market operator actions taken during emergencies.  These factors alter the supply and demand balance outside real-time market dispatch, so their effects are very similar to an operator deciding to start a gas turbine.  As capacity margins fall in most centrally-organized markets, the frequency with which the markets rely on demand response or other operator actions should rise and increase the importance of pricing these actions efficiently.  Operators often avoid shortages by taking actions that may be as costly as the value of the shortage.  Therefore, if these actions are not priced, it can significantly reduce the shortage (or near-shortage revenue) produced by the energy and ancillary services markets.