A Strategic Electricity Generating Reserve: Not Enticing Enough to Add to the Menu

The Trump administration just announced an unprecedented federal intervention on behalf of uneconomic coal and nuclear power plants in the name of promoting “national defense and maximiz[ing] domestic energy supplies.” The proposed intervention came to light last week due to a leaked draft memo written by the Department of Energy (DOE), which was quickly followed by a presidential directive ordering the Secretary of Energy to “prepare immediate steps to stop the loss of these resources.”

While the presidential directive is light on details, the leaked memo suggests that the requested DOE order would aid those plants through two rules. First, it would require grid operators to purchase electricity from a government-determined list of “fuel-secure” generators. In practice, this list is likely to nearly exclusively comprise coal and nuclear plants. Second, it would establish a Strategic Electric Generation Reserve (SEGR), which is a collection of generators set aside to supply power when electricity markets or other sources fail to deliver sufficient power to meet electricity demand.

Much ink has been spilled in the past week over the alleged need and justification for the proposed order, generally focusing on its first component—the mandate for grid operators to purchase electricity from coal and nuclear plants. Commentators have argued, and we agree, that this action appears to be a solution in search of a problem, and that it would simply create large costs for American energy consumers with negligible benefits for grid resilience.

The supposed problem to be solved is the potential unreliability of the electricity system. However, as has been widely discussed, the primary causes of electricity outages are transmission and distribution problems such as trees coming into contact with power lines. Historically, only 0.00007 percent of major disruptions have been due to fuel insecurity and only 0.009 percent have been due to generation inadequacy, meaning a solution focused on “fuel-secure” generation is ill-suited to address the purported problems highlighted by the memo.

While the proposed order seems inappropriate to solve the problem that it is supposedly designed to solve, in this post we consider the economic issues around the implementation of the proposed order. Since the first component of the proposed order—the electricity purchase mandates—have received much attention already (see, e.g., Forbes, CNBC, Utility Dive, and the Examiner), we instead focus on the second component: the Strategic Electric Generation Reserve (SEGR).

The Strategic Electric Generation Reserve (SEGR). How Might It Work?

The DOE memo contains very limited information about the SEGR, leaving many questions about how it would function in practice. That said, key design features include:

1. Reserve Size: How big would the reserve be?
2. Rules for Eligibility and Procurement: What kinds of capacity are eligible? How would resources be procured?
3. Rules for the Reserve’s Use: What is the rule for tapping the reserve?
4. Rules for Participation in Other Markets: Can reserve units participate in the energy market? Put differently, do they routinely generate electricity, or only in exceptional circumstances?

One analogous policy comes to mind that can shed some light on these features—the United States has long held crude oil reserves in the Strategic Petroleum Reserve (SPR). (RFF researchers, including one of the authors of this post, have written recently offering guidance on the use of the SPR and on factors affecting its optimal size.) The SPR is a reserve of physical barrels of crude oil. By contrast, a strategic electricity reserve would not physically store electricity because electricity cannot be cost-effectively stored at scale. Instead, a SEGR would involve accumulating a group of power plants to be available to generate incremental power on demand. However, the SPR is not the only reserve with which to draw a comparison.

A Comparison: The German Strategic Reserve

As it turns out, Germany is beginning to experiment with a strategic reserve as part of their ongoing energy transition (“Energiewende”), which includes a complete phase-out of nuclear power. Germany’s energy transition is much further along than the changes in generation mix experienced so far in the United States. To the extent that Germany’s transformation reduces the reliability of its grid (and it is not clear that it does), there is perhaps a stronger case for an electricity reserve there. From 2008 to 2018, Germany has retired more than half of its nuclear capacity (from 21 to 10 gigawatts (GW)) while more than tripling its non-hydro renewable capacity (from 29 to 101 GW), leading to fully 50 percent of its capacity comprising non-hydro renewables, compared to only 27 percent coal and nuclear and 14 percent gas. Generation inadequacy has not led to heightened supply disruptions thus far, but the expected continuation of these trends has raised concerns about tighter reserve margins in the future, suggesting a potential role for strategic reserves.

Even if there is a case for strategic reserves in Germany, the argument is far weaker for the United States. Indeed, the US experience looks tame by comparison: we have plenty of spare generating capacity, only about 13 percent of capacity is non-hydro renewable, and even with recent coal and nuclear retirements, 33 percent of capacity is coal and nuclear and 44 percent is gas (see EIA Electric Power Annual, Tables 4.2a and 4.5).

If the United States is going to pursue a reserve similar to Germany’s, it would behoove us to learn from the German policy. Fortunately, the Germans have been developing this concept for several years (see Keles et al. 2016 for an academic discussion of strategic reserves) and have identified a number of economically-minded design principles that aim to achieve a relatively efficient system. These principles address the four points mentioned above:

1. Reserve Size: Initially, the reserve will be capped at 2 GW, which is about 2 percent of annual peak load in Germany. After 2020, the reserve is allowed to increase to as much as 5 percent of peak load, or about 5 GW.
2. Rules for Eligibility and Procurement: Reserve capacity will be procured through competitive, technology-neutral auctions. A wide variety of resource types—not only power plants but also demand response providers—are welcome to offer their services to the strategic reserve.
3. Rules for the Reserve’s Use: The reserve will “be used only after all market-based solutions are fully exhausted.” These plants are kept as a last resort.
4. Rules for Participation in Other Markets: Plants that participate in the reserve will not be eligible to participate in, or return to, the energy market. In other words, strategic reserves are for emergencies and are not meant to simply be a stop-gap that props plants up long enough so they can return to the market.

How do these design features stack up to what we can glean of DOE’s proposed SEGR? Regarding the first feature, the reserve size, Germany’s 2 percent reserve is not a large one. By comparison, the US electricity grid operators already retain reserve margins (i.e., generating capacity in excess of anticipated peak demand) typically between 15 percent and 20 percent. Since the case for such a reserve is stronger for Germany than for the United States, this suggests a US reserve—if it should exist at all—should be smaller than Germany’s 2 percent of peak load. DOE’s memo gives no indication for the proposed size of the SEGR.

Regarding the second design feature, eligibility and procurement, economists favor competitive, technology-neutral auctions, as the German plan entails. Competitive auctions ensure that consumers are not overpaying, and technology-neutrality promotes economic efficiency by ensuring that no low-cost capacity options are left by the wayside. In addition, technology neutrality is consistent with longstanding FERC and NERC principles.

It is not clear from the scant description in the memo how the SEGR would be procured, but the heavy-handed approach for the electricity purchase mandates suggests that competitive auctions are probably not under consideration. It seems more likely that plants would be chosen in the same way that they would be chosen for the electricity purchase mandates—based on a federally-determined list of “fuel-secure” generators (best interpreted as coal and nuclear plants). Such administratively-determined procurement based on a subset of available resources would likely fail to identify the lowest cost providers of reserve services thereby undermining economic efficiency, resulting in higher costs for consumers.

Regarding the third feature, economists tend to favor relying on market mechanisms as much as possible before resorting to administrative solutions like deploying the reserve. Unfortunately, the DOE memo offers no clues for how the reserve would be used in practice.

Finally, strategic reserves should not distort energy markets by suppressing prices, which discourages firms from investing in new capacity or demand response. If the reserve is seen as a stop-gap measure that will prop up unprofitable generators temporarily and for them to later re-enter the market, new capacity investment will be deterred. Germany’s plan addresses this concern by prohibiting reserve capacity from later re-entering the market. Unfortunately, DOE’s proposed order is specifically designed to send the message that government policy will find a way for unprofitable plants to return to the market, even calling its own order a “stop-gap measure” (page 3 of the DOE memo).

You Can’t Have Your Cake and Eat It Too

This last point illustrates another internal tension in DOE’s proposed order: it wants to have its reserve and eat it too. The first component of the rule requires grid operators to purchase electricity from a list of power plants. The second component requires setting aside generation reserves that can provide incremental power in the event of an emergency. But a plant that is already generating cannot be relied upon for incremental power as a reserve. Buying generation from a plant eliminates that plant’s usefulness as a generating reserve. Germany’s plan solves this issue by prohibiting reserved plants from participating in the energy markets, only calling upon them in an emergency. Will DOE’s proposed order do the same? If so, reserved plants will generate power only rarely, resulting in little impact on the demand for coal that the administration wants to spur.

But before they start to bake this cake, policymakers at DOE need to ask themselves if a strategic generation reserve is really something we need to add to the national energy policy menu. We may not need to turn to Europe for lessons here when we can take a lesson from our own back yard. We already have recommended reserve margins that vary by electric reliability region across the United States. Most of the regional grid operators report realized current reserve margins that exceed the recommended levels and in some cases by substantial margins, suggesting very limited (if any) value to paying to keep an additional set of generators in reserve. So, what is the problem again? Arguably, in fact, consumers are already paying too much for reserve capacity in many locations, and adding to that burden without a demonstration of actual net benefits would be a bad deal for consumers.