Regulation and Utility Performance

Last week’s outages in Houston due to Hurricane Beryl were agonizing and frustrating. The Category 1 hurricane made landfall on July 8 and maintained hurricane strength until it reached Houston, delivering 10-15 inches of rain in some areas and resulting in 2.7 million power outages in the region. Four days later, over one million customers were still without power. Today, nine days after the storm hit Houston, 40,000 customers still don’t have power.

Criticism abounds and is pointed at various parties — the Governor and Lieutenant Governor for mishandling the federal disaster declaration, Houston’s mayor (who then blamed his predecessor), local officials, and CenterPoint Energy (which we’ll come back to shortly).

One important factor in analyzing the storm response and assessing performance is the confluence of events: flooding and outages from a storm in May, then Beryl, and now a heat wave, almost like a summer version of the combination in 2021’s Winter Storm Uri. This pattern is called a compound disaster (this link is a really informative National Academies report if you want to dig deeper), in which multiple events cascade and affect a system’s resilience, its ability to rebound from damage and return to full functionality. When compound disasters occur in complex cyber-physical-social systems like a region and its electricity system, both the physical infrastructure and the social capital that make an area resilient can be stressed beyond their limits, especially when the compound cascade is a pattern that is extremely rare or hasn’t been experienced before.

Beryl struck a region that’s still recovering from destruction caused by wind and water this year. The Houston area experienced severe flooding at the beginning of May after a torrential downpour drenched the area in upward of 23 inches of rain. A band of intense thunderstorms known as a derecho then galloped across Texas on May 16 and charged into Houston, tearing off roofs, shattering windows, and knocking down power lines. Another burst of thunderstorms and tornadoes erupted over Memorial Day weekend, lashing Houston with winds up to 88 miles per hour.

These back-to-back storms stressed infrastructure and depleted emergency resources. They also left little time to rebuild, let alone harden streets, power lines, and sewer systems against future disasters.

Then Beryl brought its own suite of destruction. It pushed a life-threatening storm surge onto the coast, dumped up to 15 inches of rain, and even spawned tornadoes that pushed trees over into homes and utility poles. Power companies are warning residents to stay at least 35 feet away from downed power lines.

Now a heat wave is baking the region, threatening vulnerable Houstonians who don’t have power right as they need it the most. At a city council meeting this week, officials pressed a representative from CenterPoint Energy, the main power provider in Houston, about the ongoing outages. (Vox July 12, 2024)

If you want to read more from an electricity policy perspective, I recommend Doug Lewin’s Texas Energy and Power Newsletter highly; Doug has had several recent articles about Beryl and Houston, one of which I’ll come back to shortly.

In addition to offering sympathy for the people in Houston affected by these events, I want to use the example of this compound disaster and a conversation on X/Twitter to broach this question: even in a state like Texas that has both retail and wholesale competition and regulates the wires companies, is traditional rate-of-return regulation a good institutional fit with ensuring resilience in the face of such events?

A regulatory system predicated on rewarding capital investment doesn’t provide incentives to do things that aren’t capital intensive. And since the utility has a monopoly, other people are legally prevented from doing many other things that would increase both resilience and reliability. The best people to make investments for reliability and resilience aren’t always the utility.

CenterPoint Energy is the regulated distribution wires company in the Houston region, and they have received considerable criticism for how they have handled the Beryl recovery. They didn’t have updated outage maps, some areas had power and then lost power for extended periods days after the storm, and their lack of communication with customers has left many frustrated (and the video of a CenterPoint representative delivering a message sitting under a thermostat set and working at 70 degrees was some pretty tin-eared post-disaster communications). The regulator, the Public Utility Commission of Texas, is opening an investigation into their performance with the rest of hurricane season looming.

The form of regulation that electric utilities face in the United States is called rate-of-return (ROR) regulation. The idea of ROR regulation draws on natural monopoly theory, the argument that the high fixed cost structure of costs in this industry lead to declining average costs and a tendency toward consolidation into a monopoly with associated high (and inefficient) pricing.

Here’s how ROR works: The idea is to control the profits of regulated utilities, ensuring that they earn a fair return on their investments while protecting consumers from excessively high rates. It’s a a cost-based form of regulation, focused on ensuring that public utilities cover their operating costs (labor, tree trimming, etc.) while also earning a fair return on their capital investments. Their operating costs, plus depreciation and taxes, are cost components that along with the value of their capital assets (a.k.a. rate base) and their allowed rate of return determine their revenue requirement.

The ROR represents the utility’s opportunity cost of capital as the fundamental idea underlying the “fair” rate of return that regulators set. The idea is to compensate them for their capital that could be used in some other productive and profitable way if they weren’t using it to deliver electricity reliably. The allowed rate of return typically reflects a combination of the cost of debt (interest rate), the cost of equity (usually estimated through a discounted cash flow or capital asset pricing model), and looking at the utility’s combination of debt and equity to calculate its weighted average cost of capital (WACC). They earn profit on capital investments, not on operating expenses, not on tree trimming.

This financial accounting implementation of ROR regulation uses cost-recovery-based rules to control price and quantity by controlling return on capital inputs, theoretically keeping regulated prices close to costs and quantity as high as feasible to achieve universal electrification. The other relevant part of ROR regulation that’s outside of the ROR formula is the determination of the output/product/service that the regulated utility provides. Regulation defines what the utility does and the quality level at which it’s expected to do it. For electric utilities quality is measured in terms of reliability, with two general metrics measuring the duration of outages (SAIDI) and the frequency of outages (SAIFI).

You may have already spotted one incentive problem in this setup (there are more!): the utility has an incentive to provide higher quality if they can invest capital in it and increase their rate base, but regulators (and associated stakeholders like consumer advocates) who work to keep regulated rates affordable have an incentive to either reduce allowed investment or the allowed ROR, or both, which may reduce the utility’s effort toward quality/reliability. Of course regulators are charged with the mission of enabling safe, affordable, and reliable utility service. Increasingly both utilities and regulators are focusing also on resilience, which is different from but related to reliability because resilience is the ability of a system to absorb a shock and rebound from it.

The utility can misjudge its effort, and regulators can make mistaken judgements of the prudence of a given investment, which can lead to protracted outages like those seen in Houston. In this specific case, CenterPoint had applied for a $100 million Department of Energy grant to harden their distribution grid, but their application was rejected and they have reapplied in the next round. The PUC regulators are also on schedule by October to review CenterPoint’s hardening plan for investments including metal poles rather than wood. But those errors are even more likely in these compound disaster settings. How can/should utilities and regulators account for and plan for such compound disasters (and the correlated risks embedded in them)?

Rather than looking at the utility or the regulator and saying “do better!”, we should ask whether the structure of ROR regulation itself is the underlying problem. Think about it: the utility’s primary incentive is to build stuff because that’s what they earn their ROR on. For the past century, building power plants and poles and wires and transformers and substations and meters delivered on safety, affordability, and reliability.

Wires distribution companies still should be and will be capital intensive, as the proposal to replace wooden poles with metal poles illustrates. But there’s also a lot that can contribute to reliability and resilience that doesn’t require a lot of capital. Like tree trimming.

This is where an interesting little exchange on X/Twitter got me thinking:

Although quality requirements exist (such as SAIDI/SAIFI performance), regulated utilities aren’t really penalized for outages, so actions like tree trimming are not profitable (and on top of that customers do not get compensated when they suffer losses in outages, so customers bear considerable outage risk). Replacing wooden poles with metal poles is profitable because poles are capital and count as rate base.

Cost-recovery-based, capital-based ROR regulation is part of the problem because of how it structures the business model and the incentives of the regulated utility, and how it precludes others from taking beneficial actions. For at least the past 15 years there’s been some attention to changing ROR regulation so it’s less focused on capital investment and on the quantity of electricity the utility sells and more focused on performance based on some metrics. This performance based regulation (PBR, not the beer) involves defining certain performance metrics and thresholds for meeting them, and tying utility revenues to its performance on those metrics. The National Council of State Legislatures has a useful primer on cost of service regulation and performance based regulation.

So, for example, if a utility is not meeting outage metrics it pays a penalty/has its revenue reduced in much the way that my interlocutor King of Power described it in his post above. 17 states and DC are exploring PBR, but bureaucratic processes move slowly, especially ones that upend a century-old status quo.

I see PBR as on a continuum toward another regulatory model: price cap regulation.

As economist David Sappington says, price cap regulation does what it says on the tin:

Price cap regulation incorporates a form of regulatory lag while delegating some pricing discretion to the regulated firm. The typical price cap regulation plan specifies a price cap period (often 4 or 5 years) and a maximum rate at which the regulated firm’s inflation-adjusted prices can rise annually, on average, during this period (Sappington and Weisman, 2010). The firm’s realized costs and earnings are usually reviewed at the end of the specified price cap period, and often employed to update estimates of the pricing restriction that would allow the regulated firm to secure a reasonable level of earnings in the next price cap period. When the firm’s realized performance is employed to set future standards, the firm’s incentives for exceptional performance are dulled. However, price cap regulation provides some incentives for cost reduction by instituting an explicit regulatory lag. (Sappington 2015)

This style of regulation has been implemented most notably in telecommunications and in the UK’s electricity regulation, known as RPI-X. The utility’s cost and capital investment estimates are incorporated into establishing what’s essentially the utility’s revenue cap, and in the UK that cap ratchets down by an X factor per year over the term of the cap. The X factor gives the utility an incentive to innovate when innovation will reduce their costs and increase their profits. When paired with definitions of product and service quality, it counters the perverse incentive that ROR regulation gives utilities to choose the capital-intensive approach to their operations. If a regulator is concerned about quality degradation, then the RPI-X structure can include performance incentive mechanisms, so I see price cap + PBR as a both-and, at least conceptually. Thus it’s possible, or I would even say likely, that price cap regulation with outage penalties would be more likely to result in lower-cost resilience-enhancing maintenance and the utility’s ability to be more nimble in the face of compound disasters.

Price cap regulation doesn’t do anything, though, about the second obstacle: the legal barrier to other people making investments that can enhance resilience and reliability. Here Doug Lewin is absolutely right when he observes that microgrids would solve multiple problems arising in situations facing Texas and other states:

Distributed energy resources, or DERs, are collections of solar panels, batteries (including those in EVs), connected devices and appliances, and gas generators. They’re spread among homes and buildings and configured as highly resilient microgrids. When trees fall on transmission lines, these virtual power plants continue to create power and keep people and communities safe. …

Microgrids also add power to the grid in times of scarcity. Other times, microgrids produce power for homeowners and business owners, lowering consumers’ energy bills year-round.

DERs and microgrids aren’t a panacea; unfortunately, there is no panacea for the grid. But they’re among the only solutions that address nearly all of the state’s energy challenges simultaneously. They’re also languishing far down policymakers’ priority list, behind things like new gas power plants that take longer, cost more, and don’t work as well. …

Texas has adopted a groundbreaking policy to incentivize microgrids and received accolades for it, but it doesn’t amount to much if it’s not implemented.

It’s past time for the PUC to get moving and prioritize microgrids consisting of distributed energy resources — plus other solutions such as home weatherization and energy efficiency — that improve resilience, help solve resource adequacy, and lower bills.

Microgrids are smaller grids that can be self-contained and therefore can island, or detach from the distribution grid in ways that can benefit both, because they harness the robustness of diverse DERs that are digitally connected and can operate flexibly to serve both individual and grid requirements. Whether it’s resilience in the face of a hurricane or building new supply and new capabilities in the face of growing demand from population growth and data centers, microgrids contribute to making power more abundant and affordable.

But they challenge a status quo in which regulated utilities have the legal monopoly over wires networks that cross public rights of way. Office park microgrids, campus microgrids, hospital microgrids may not run into that constraint, but anything bigger than that does.

Is traditional rate-of-return regulation a good institutional fit with ensuring resilience in the face of such events? 20th century public utility rate of return regulation creates incentives to invest in capital when capital may not be the only or the best approach. It also stifles the development of private non-utility investments like microgrids that can complement the utility’s grid and bolster its resilience. It gets in the way of using 21st century technology to create resilience, abundance, and affordability. It gets in the way of innovating to reduce the likelihood of Houston’s experience happening again when the next compound disaster occurs.

This post was originally published on Lynne’s Substack, Knowledge Problem. If you enjoyed this piece, please consider subscribing here.