Last week, the California Public Utilities Commission took its latest crack at fixing the state’s resource adequacy program, which sets the rules for how the state procures enough generation capacity to assure its grid stays running even in times of peak energy demand.
The resource adequacy (RA) program has served that purpose well in the decade-plus since it was created after the California energy crisis. But the rise of renewable energy, the looming closure of many natural-gas-fired power plants, and the increasing share of customers being served by community-choice aggregators are creating new challenges that require some significant changes, according to the decision (PDF) passed unanimously by the California Public Utilities Commission.
We’ve been covering the transformation of California’s energy system over the past decade, including the rise of solar power as a significant, and sometimes excessive, contributor to the state’s generation profile, and the concomitant decline for the natural-gas-fired power plants that have traditionally served as the resource of last resort for the state’s capacity needs.
We’ve also been covering the rise of community-choice aggregators (CCAs), from a handful of scattered counties and cities procuring their own renewable energy portfolios, to a scale and scope that threatens to remove millions of customers from the rolls of the state’s investor-owned utilities — most notably Pacific Gas & Electric, which is also undergoing a transformational bankruptcy proceeding.
For the RA program, these changes have led to an increasingly unstable situation, as Commissioner Liane Randolph wrote in a blog post: “The law requires us to avoid costly backstop procurement, but that has become more frequent in recent years.” According to California Public Utilities Commission (CPUC) President Michael Picker, last year saw 11 waivers issued to CCAs or small electricity providers that could not fully meet their RA requirements.
State grid operator CAISO must procure “backstop” resources to fill those gaps, under the requirements set by the Federal Energy Regulatory Commission, and with fewer and fewer gas-fired plants in the state remaining to provide that reliability, “plant owners are increasingly able to exercise market power,” Randolph wrote.
Over the past few years, various CPUC commissioners, including President Picker, have raised the concern that increasing migration of utility customers to CCAs is undermining the RA program. The program now relies on a complex set of rules to govern how and when CCAs take over RA responsibilities from the utilities they’re being carved out of.
We’ve seen some far more contentious policy battles over the role of CCAs in the state’s energy landscape. In October, the CPUC unanimously voted to boost the Power Charge Indifference Adjustment fees that CCAs, as well as competitive energy providers under the state’s Direct Access program, pay utilities when they take over their customers — a move decried by CCAs as potentially adding hundreds of millions of dollars to their costs.
But in the matter of how to fix RA, CCAs are more aligned than at odds with the state’s utilities. In fact, the various parties to the RA proceeding have shown a remarkable level of agreement on what’s needed to improve the state’s capacity mechanism:
- First, extending the current one-year procurements to multi-year contracts to lock in more certainty at better prices.
- Second, creating some sort of central procurement mechanism to manage the increasingly fractured landscape of electricity providers in the state.
Last week’s decision addressed both of those goals, but was only able to reach consensus on the first. That's largely because creating a central procurement entity for the state’s capacity needs is too complex, and too fraught with uncertainties, to proceed at this point.
But on the issue of multi-year procurement, the CPUC was able to find consensus on a structure for three-year RA procurements starting in 2020. The specifics of how it’s to be implemented, however, may be counter to some parties’ wishes.
Multi-year RA contracts: A consensus idea, with the devil in the details
California’s reliance on one-year-ahead RA procurements is unusual among the country’s various capacity markets and mechanisms, which tend to look several years ahead. Multi-year contracts offer benefits for both the “load-serving entities” (LSEs) such as utilities and CCAs, and the power plants and other capacity resources such as energy storage or demand response, in the form of increased certainty of supply and price.
Extending today’s one-year RA contracts is such a popular idea, in fact, that the only big disagreement between parties was on whether to increase it to three years, as most advised, or to five years, as a handful of parties including PG&E and San Diego Gas & Electric suggested. The CPUC chose three years on the grounds that it reduced the risk of over-procuring for a future that could change quickly, citing the example of PG&E’s commissioning of the 2.2-gigawatt Moss Landing battery park, which was approved last year and is set to come online later this year.
But there are complexities involved with extending a system designed for year-ahead procurements to cover three years, the CPUC noted, particularly with the fluid status of LSEs in the state. Last year, the CPUC issued a Track 1 decision in its RA proceeding that set new rules for how CCAs’ share of RA responsibility is managed, requiring them to register and participate in the process at least one year before launching. But the CPUC is “unable to anticipate when new LSEs will form or how load will migrate among LSEs beyond the one-year timeframe” — and with the current growth rate of CCAs, this could amount to a significant shift in load served by the third year of a three-year RA contract.
To solve this problem for 2020 procurements, the CPUC settled on allocating LSE’s local requirements for all three years based on its first-year load share as provided by the California Energy Commission load forecasting process, and updating requirements for years 2 and 3 “during the following year’s year-ahead allocation process.” And on the advice of CAISO, it extended the current one-year waiver and penalty structure for LSEs that can’t meet their RA requirements to three years.
But regulators are also planning to “revisit the LSE-based component of multi-year local procurement in a decision to be issued in the fourth quarter of 2019,” indicating it’s not done fine-tuning on this issue.
In addition, the CPUC addressed a potential conflict between increasingly localized RA procurements, which is good for assuring that local grid constraints or challenges are effectively mitigated, and expanding the scope of procurements to avoid putting too much market power in the hands of localized resources. Specifically, last week’s decision undid a previous decision to combine six of PG&E’s “local areas,” namely Sierra, Fresno, Humboldt, North Coast, Stockton and Kern, into one called “PG&E Other Areas,” in order to break them up into their own local areas again.
This decision folds into a broader one about how granular the state’s future RA procurements should be, the CPUC noted. CAISO is arguing for breaking down procurement into “sub-local areas,” as opposed to the bigger “local areas” they’re part of, to “more closely tie procurement requirements with local capacity needs and operational requirements, reducing the potential for inefficient local procurement and CAISO backstop procurement.”
But at the same time, it noted, “we are not convinced that this level of disaggregation is workable in the current bilateral market,” particularly in areas where market power may lie in one or a few resources’ hands. This situation “may lead to LSE deficiencies and inevitable backstop procurement,” the very thing CAISO is seeking to reduce by going to the sub-local level.
The CPUC did not make a final decision on this issue — but it did cite the disaggregation of the "PG&E Other" local area as “a necessary first step toward addressing inefficient procurement that may lead to backstop procurement. This level of disaggregation will also provide useful feedback to the commission in assessing further disaggregation to the sub-local area level” in future proceedings.
Central procurement: A good idea, but who’s going to do it?
Centralizing procurement of resource adequacy into a single entity has been a possibility for the CPUC for the past few years, and was the subject of much discussion and analysis as part of last year’s Track 1 decision on RA. Today’s system relies on each LSE securing its own RA capacity, a structure that’s coming under increasing strain with the rise of CCAs and direct access customers.
Last year’s decision found that “central procurement system, at least for some parts of the local RA requirement, was ‘most likely to provide cost efficiency, market certainty, reliability, administrative efficiency, and customer protection,’” indicating it’s worth pursuing.
Almost all the parties involved in the RA proceeding, from utilities to CCAs to CAISO to environmental groups like the Natural Resources Defense Council, also support some form of central buyer for at least a portion of future RA needs, the decision noted. And a majority of them agree that the most likely party to take on that responsibility, at least for the short term, would be the investor-owned utilities themselves.
The problem is that none of the parties — including the utilities themselves — want them to actually serve that purpose, at least not beyond the short term. For CCAs and renewable energy advocates, letting the investor-owned utilities (IOUs) procure capacity on their behalf opens up problems of conflicts of interest and lack of transparency, as well as putting long-term authority in an entity that may well be shrinking compared to alternatives such as CCAs, or the rising share of customer self-generated solar energy.
At the same time, “the distribution utilities themselves are either unwilling to take on this role or agree to do so on an interim-only basis,” the CPUC wrote. Southern California Edison was the most open to the idea, but only if “certain conditions are met, such as durable cost recovery and equitable cost allocation” and “only on an interim basis,” while PG&E and SDG&E proposed the creation of a new entity to serve the role.
Utilities also had “concern with the potential financial costs and risks associated with the central procurement function,” the CPUC noted. In particular, “the increased financial commitment associated with large-scale procurement could raise debt equivalency issues.” Debt equivalency is a term to describe how credit ratings agencies account for the ongoing costs of buying power from third parties through power-purchase agreements — while it’s not a debt, it is an ongoing commitment to the company’s balance sheet.
As SDG&E wrote in its comments to the proceeding, “Debt equivalence applied to a utility’s balance sheet…without corresponding increase in equity or compensation could negatively impact the utility’s credit standing and financial stability.” This is a particular concern for California’s IOUs in light of the state’s inverse condemnation legal doctrine, which holds utilities responsible for damages from fires or other disasters caused by their equipment whether or not they’re at fault.
Inverse condemnation likely accelerated PG&E’s descent into bankruptcy in response to the potential tens of billions of dollars of liabilities it faces in association with wildfires of 2017 and 2018, and “numerous parties cite concerns for PG&E’s precarious financial position with respect to exposure to wildfire damages and solvency issues” as a reason why it shouldn’t serve as a centralized RA buyer in particular.
“On the other hand, parties who support designating the distribution utilities (and even some who oppose) acknowledge that the investor-owned utilities are likely the only candidates who can serve the central procurement function in the immediate term,” the CPUC wrote. Even the typically anti-utility nonprofit The Utility Reform Network wrote in comments that the IOUs are the “only feasible entities” to serve as central buyers, since they “have the resources, the knowledge and experience to take on this task effectively.”
There are other options for central procurement, but they offer their own problems. For example, SDG&E and PG&E advocated for a special purpose entity — “a new state agency or private entity selected through a competitive solicitation process or through legislation” — to collaborate with CAISO and the CPUC “to select an optimal portfolio to meet local needs.” But as multiple parties noted, creating a new government entity would involve “substantial time and expense,” as well as legislation authorizing it. All of these factors would have to be filled in before a new entity could be considered as an option.
The CPUC also looked at a proposal from CalCCA and power plant operator Calpine for making CAISO itself the central buyer, on the grounds that it’s already doing the core work behind the RA program, is governed by tariffs, operates independently, and has relatively transparent procurement compared to utilities. As CalCCA noted, CAISO also has “tools and legal authority to spread costs across the utilities’ service territories on cost-of-service rates, if contract negotiations fall through.”
But other parties raised concerns with the CAISO serving as central buyer — most notably, the potential to open California’s capacity market and environmental goals to federal oversight, starting with the stakeholder initiative process to design a new market structure and the related tariff amendments for approval by FERC. These same concerns applied to CPUC’s decision to reject a proposal for California to create a centralized capacity market like those operated by mid-Atlantic grid operator PJM and other independent system operators, put forward by the Alliance for Retail Energy Markets and energy services provider Shell, which is active in the state’s Direct Access program.
Given this set of conflicts, the CPUC determined that it “does not find a viable central buyer at this time” and delayed its decision on the issue. Instead, it’s asked parties to the proceeding to “propose central buyer structures” and bring them to discuss at a series of workshops over the next six months.
Bringing transparency to an opaque system
One of the challenges in assessing the RA program’s effectiveness is its lack of transparency to outsiders — a problem the CPUC acknowledged, and promised to fix, in last week’s decision.
“We recognize that certain information regarding the broader RA procurement outlook is not publicly available and only visible to Energy Division staff,” regulators noted. This information includes how many megawatts of different resources are being procured in different regions by different LSEs, and importantly, where deficiencies are occurring and which LSEs are asking CAISO for waivers.
To address this, CPUC staff will be preparing two reports, the first due within 60 days and the second next year, that will provide much more public detail on the RA program than has been available to date. The list includes:
- Total megawatts for any/all resources procured (gas, storage, renewal/DER) to meet RA requirements
- Development of preferred resources in local and system areas
- Information regarding local deficiencies, including the number of LSEs that are deficient, type of LSE, location of deficiencies, amount of deficiencies (in MW), number of local RA waiver requests, and anonymized statements from the LSE as to the reason for the deficiency (such as which generators bid into the solicitation, whether the bids included dispatch rights or other terms addressing how local resources bid in the energy market)
- Information regarding system and flexible capacity deficiencies, including anonymized statements from the LSE as to the reason for the deficiency
- Resources on the Net Qualifying Capacity list that are not shown in RA filings as under contract to an LSE(s)
At the same time, the CPUC’s decision keeps confidential much of the data on RA procurements, particularly that considered to be market-sensitive, such as current program share or price-per-megawatt for specific resources. In a compromise, the CPUC took up PG&E’s proposal to “disclose all resources used to satisfy an LSE’s RA obligation in the previous year without identifying the number of megawatts associated with the resources.”
This method will meet the requests of environmental groups such as the Sierra Club for information on which resources different utilities or CCAs are tapping for resource adequacy, while protecting “sensitive information such as an LSE’s load share and open position.”