Electricity Planning Quagmire: Marginal Cost Pricing & Renewables

So even though renewables can provide a benefit in lowering the overall clearing price for energy, there will come a point – if we have not already reached it – where there will be so many wind and solar resources bidding into the RTO markets that marginal clearing prices no longer benefit customers … [by making] dispatchable-on-demand generation resources needed for system reliability unable to survive economically. (Bernard McNamee, below)

A recent article at RealClear Energy by former FERC commissioner Bernard McNamee, “Why Marginal Pricing in Wholesale Electric Markets May Need Reform” (June 20, 2021) recognizes a problem with regulated pricing that results in a very inconvenient truth: renewable energy has blown up the neoclassical planning model for electricity in the Texas ISO (Independent System Operator) construct. (And it is hurting reliable generation in the RTO (Regional Transmission Organization) construct.

I am surprised that the RTO/ISO experts I have been in debate with (Robert Borlick; Lynne Kiesling; Joe Polalsky; Eric Schubert) have not emphasized (recognized?) the second of two major problems of renewables:

• Disappearing at the peak (“expected”)
• Badly compromising the day-to-day economics of the reliables (gas-fired, coal-fired, and nuclear generation) via marginal cost pricing (below).

The Texas grid is badly wounded for the first time in its long, storied history. The cause is interacting government intervention, as I outline here and here. The experts/regulators/planners are running scared with hopes that mild weather and strong baseload (non-renewable) performance will prevail in Texas this summer and beyond.

But with much more wind and solar coming on the grid, and not a whole lot more legislative/regulatory reform (other than criminalizing non-preparedness such as weatherization), the new normal may be grid unreliability.

McNamee’s article is excerpted below in the relevant part (his argument that rates are worse in RTO/ISO “competitive” markets versus traditional utility service is not of interest here).

… we need to pay attention to the warning signs in places like Texas and California where reliability is decreasing.

Unlike traditional regulation of vertically integrated utilities – in which public-service commissions establish rates on a cost-of-service basis – RTOs use competitive marginal-cost auction markets to set a clearing price for energy (some also do so for capacity).

Under this market process, the amount of electricity that the RTO [ISO] predicts is needed at any given time is published, and electric generators then bid to serve that need. The last megawatt of generation needed sets the price for energy paid to all the generators who clear the market.

When the RTOs [ISO] and their marginal pricing models were developed in the late 1990s and early 2000s, most of the electric grid got its energy and capacity from resources that used fuel to generate electricity – coal, nuclear, and natural gas, along with hydro.

Because the demand for electricity would rise and fall depending on the time of day (the load curve), generation resources were generally stacked in the bidding to serve load. Coal and nuclear tended to run 24/7 and served the base load. Their capital costs were generally high and their operating costs, including fuel costs, relatively low.

Natural gas would serve the intermediate and peak load. Natural gas plants had lower capital costs but relatively more expensive fuel costs. Because natural gas generation at the time was usually the most expensive and most flexible resource, the last selected natural gas generating unit would usually set the energy clearing price.

This meant that all the generation resources selected would be paid the natural gas resource clearing price for energy regardless of their costs – including at the peak times of day, when scarcity would drive up the overall price for energy. The economic benefit of this system was that the markets would motivate the development of more efficient generation technology and the cheapest fuel to meet the needs of the grid.

Arguably, and as many studies concluded, this incentive model helped reduce overall wholesale [retail] electric rates. These savings were also spurred by a combination of cheap natural gas, additional subsidized renewables, and low load growth.

But things have changed in the electric markets since they were first developed. Today, we see more and more intermittent renewables resources entering the energy markets. These resources have no fuel costs and receive substantial government subsidies such as production tax credits, investment tax credits, and state subsidies.

Because marginal pricing means that the energy price set by the last resource picked is paid to all resources that clear the market, renewables with little to zero costs are paid the same amount of money as dispatchable on-demand resources – usually a natural gas plant. (Of course, this is a little more complicated because markets use locational marginal prices (LMP), which include things like transmission congestion costs and line losses, in addition to the system energy price.)

And because of the nature of the RTOs, this means the actual lower-cost benefits of renewables – no fuel, tax credits, and RECs – do not always flow through to end-use customers. This arrangement can be compared to states with rate-regulated utilities under a cost-of-service model, where the utility is usually obligated to pass fuel, tax, and operating savings to customers.

So even though renewables can provide a benefit in lowering the overall clearing price for energy, there will come a point – if we have not already reached it – where there will be so many wind and solar resources bidding into the RTO markets that marginal clearing prices no longer benefit customers. Paradoxically, though all the financial benefits of these resources may not be flowing to customers, the overall effect on the grid may be to suppress prices – and make dispatchable-on-demand generation resources needed for system reliability unable to survive economically.