9.2.1 The Rationale for Capacity Payments
Like many interventions in market systems, the rationale for capacity markets in electric power arises from some attempts to correct for perceived market failures. These interventions create distortions in the electricity market, and capacity payments have been the favored mechanism to correct for these distortions. In the case of electric power generation, there are three important such distortions or market failures:
- Power grid operators are required to maintain a certain level of excess power generation capacity beyond what they expect to need to meet peak demand. This requirement grew out of a large blackout in the 1960s, when the North American Electric Reliability Council (NERC) was created to develop reliability standards to which utilities and grid operators agree to adhere. This NERC requirement is known as "resource adequacy" and the level of excess power generation capacity required is called the "installed capacity margin."
- Following the California power crisis, the Federal Energy Regulatory Commission placed price caps on restructured electricity markets in the US, limiting how high that prices in the day-ahead or real-time market could get. For many years these price caps were set at $1,000 per MWh. Some regional markets have started to increase these price caps in recent years.
- Recall from earlier in the course that LMPs can be very volatile, changing in some places every five minutes. Most retail electric rates, however, are fixed at a certain level for months or even years at a time. This means that electricity users tend to overpay for power when it is cheap to produce, and underpay for power when it is expensive to produce. In particular, the retail price of power during the highest-demand times of year may be substantially lower than the marginal cost to produce that power.
To see how these three distortions can cause problems in electricity markets, we'll use a simple example. Let's say that PJM needs a new power plant built for the system to meet its installed capacity margin requirement. The price cap in the market is $1,000 per MWh. The cost of the power plant is $400,000 per MW of capacity, and its marginal cost of producing electricity is $240 per MWh. The plant would run 1% of the time during the year (a capacity factor of 0.01) and the relevant discount rate is 15% for a 15-year time horizon.
All PJM needs is for someone to build this plant, run it during peak hours when LMPs are very high, and make enough profit doing so. There's a problem, however. If you calculate the LCOE for this power plant, you should find that the cost for the power plant to break even is around $1,021 per MWh.
The figure below plots the LCOE of the power plant as a function of the capacity factor, with the $1,000 price cap shown in the figure. Because the plant needs to charge more for its electricity than is allowable under the price cap, no sane investor would build this plant. Moreover, during those peak demand periods, consumers might pay a retail rate of $100 per MWh for electricity that costs $1,000 per MWh to produce. With other goods, we might expect demand to go down during times that prices go up (think about the demand for Uber rides during surge pricing, for example). But this doesn't happen in electricity because retail prices do not always reflect marginal costs very well.
PJM's problem is that it needs this plant to be built in order to have enough of an installed capacity margin. But no one has any financial incentive to build the plant. Remember, PJM can't own any generation assets and so can't build the power plant itself.
This has become known in the electricity business as the "missing money" problem. It has been exacerbated recently by very cheap natural gas and a large contribution to the grid from renewable power generation in places like Texas and California. These states have seen sustained periods of day-ahead and real-time energy market prices at $0 per MWh or even negative prices. The capacity market is one solution to the missing money problem -- one that has been adopted in every US Regional Transmission Organization except the ERCOT system in Texas.