by Julian Spector
November 26, 2018

Author’s note: We’re taking a break from the global storage series to dissect a landmark storage deal. Stay tuned for coverage of the exciting developments in South Korea and Australia in the weeks to come.

Earlier this month, PG&E won approval to build two record-breaking grid batteries.

The deadly Camp Fire quickly overshadowed that news, becoming the most destructive conflagration in California history. The possible role of PG&E’s grid equipment in starting the fire has prompted fears of bankruptcy.

Those developments complicate PG&E’s ambitious grid-modernization and decarbonization plans. The top utility regulator pledged support for the company; with California’s clean energy mission hinging on a functional utility, the state is unlikely to let the company slip into disarray.

In any case, the recent headlines have not changed the status of the 567.5-megawatt/2,270-megawatt-hour Moss Landing project, a PG&E spokesperson confirmed. Assuming that continues to be the case, this is a project any storage wonk needs to understand, so we’re picking apart the deal for this week’s Storage Plus.

Moss Landing's sheer scale demands attention. The four-project portfolio includes a 300-megawatt project developed by Vistra and a 182.5-megawatt battery PG&E will source from Tesla.

“It really is landmark in the sense that it's the biggest ever done,” said Charles Post, PG&E's energy storage program manager, in an interview shortly after regulators affirmed the project. “It will be the largest utility-owned [and] the largest third-party-owned [battery plant] in the world.”

Biggest battery records change hands annually. The project's more enduring legacy will hinge on proving out the role of batteries in a decarbonizing electrical grid.

Several groups opposed Moss Landing as an expensive and unnecessary use of ratepayers' money. To prove them wrong, PG&E will need to show that the battery really does reduce reliance on natural-gas plants. No plants are closing down yet, so the utility still has plenty to prove.

What will it do?

The stated goal of Moss Landing is to enhance local reliability and help increase renewables on the grid.

More specifically, the plants will serve resource adequacy with a must-offer obligation to the California Independent System Operator. Provided they fulfill that obligation, plant operators can monetize their assets in wholesale markets, too.

“TheNo. 1 obligation is the [resource adequacy], because if they fail there, then you end up in this contractual breach,” Post noted. “It's a very bad place that you don't want to be if you're a developer.”

The ability to play in markets like energy and ancillary services brings down upfront cost, because the utility contract doesn’t need to cover all of the revenue requirement. That introduces merchant risk on the part of Vistra, and exceedingly few entities are financing merchant battery plants in the U.S. right now. Vistra builds merchant power plants as its core competency, though, so it’s less of a stretch.

The company also reduced development costs by siting in an empty turbine hall with existing interconnection.

PG&E will own the project at a nearby substation, which Tesla will supply through an engineering, procurement and construction contract coupled with a 20-year performance and maintenance agreement. Once built, PG&E will dispatch it for resource adequacy and other needs.

That contract contains an upsize clause, which could expand the battery to 6-hour duration in the future, if the need arises and if regulators approve. That scenario reflects the modular nature of the technology.

“If our needs change, we own it, so we can pivot and use it for what we need,” Post said of the substation plant, in contrast to third-party contracts which procure a narrowly defined service.

The contracts also differ in their financial implications: The costs of third-party contracts are passed through to ratepayers; projects owned by the utility earn it a regulated rate of return.

What does this mean for storage project financing?

To fulfill this project, PG&E will build batteries at a scale never seen before. This will require a scaling up in how the industry delivers, much like the solar industry in its jump from projects with tens of megawatts to hundreds of megawatts.

The business models to support massive batteries have to evolve as well. The hybrid approach seen in Vistra’s proposal — combining contracted revenues with merchant revenues — could become a model for other markets.

Texas doesn’t offer any capacity contracts, so its storage developers need to become comfortable with more merchant risk. So far, that hasn't really happened.

But other markets, which have competitive wholesale markets and grid reliability or capacity contracts, could easily replicate Vistra’s model. New York and Massachusetts look promising in the near term, and have political leadership advocating for more storage deployments, too.

Will this displace gas plants?

The narrative that Moss Landing is the battery plant that vanquished gas peakers works better in rhetoric than in practice.

“People like to talk about batteries replacing gas plants,” said Toby Shea, a senior credit officer at Moody’s who tracks California utilities and generators. “It’s a little bit misleading, because it’s not a one-for-one replacement.”

Let’s unpack what functions the battery is and isn’t replacing.

Moss Landing came into being because the California Public Utilities Commission rejected out-of-market payments (known as "reliability must-run" or RMR) for three uneconomic Calpine gas plants, two of which are peakers. It would be accurate, then, to say the battery plants are replacing contracts that would have gone to existing gas plants.

One could be forgiven for assuming that development would result in the closure of the at-risk gas plants. Instead, the Federal Energy Regulatory Commission approved Calpine's RMR request anyway, RTO Insider reported. The plants could receive payments through 2020, although smaller sums than Calpine had hoped for.

After several tough years, conditions are improving for California gas plants.

“The market has really turned for gas plants,” Shea said. “California has been closing plants because of the once-through cooling requirement. Now, instead of a surplus, you really have a shortage of gas.”

The halting of several proposed gas plants, like NRG’s Puente plant in Oxnard, insulates existing plants from competition. The planned closure of the Diablo Canyon nuclear plant in the next few years will add a massive gap in the state’s capacity resources.

Even if the plants were disappearing, Moss Landing isn’t solving the immediate grid reliability shortfall: A transmission upgrade will fill that need in 2019, for a cost of $14 million. Several groups used this fact to argue the plant is no longer necessary (more on that later).

Moss Landing hasn't notched any kills, but it will challenge the incumbents as a new contracted capacity resource.

Supporters of the project say it will reduce reliance on the surrounding gas plants, most of which will be off of their contracts by 2019, so the state won’t need to pay to keep them open.

That role will resemble a peaker in some ways, by injecting power when state independent system operator CAISO says it needs more. The total power capacity actually exceeds a typical peaker, looking more like a combined-cycle gas turbine in scale. The difference is, Moss Landing runs out after four hours, making it hard to replace gas altogether.

Moss Landing could take care of shorter peak events in the region, denying business that would otherwise go to the surrounding gas plants. When a heatwave triggers high demand for longer than four hours, those gas plants will save the day.

The question to monitor here is what kind of a dent this battery will make in the gas plants’ revenue. Does it serve enough of the capacity need to box out one or more plants entirely? Do the plants stay online but operate for fewer hours, slowly bleeding revenue?

Expect the peaker replacement trend to materialize fractionally, rather than in absolutes: a steady weakening of gas plant economics, rather than discrete knockouts attributable to individual storage projects.

Will the opposition persist?

The CPUC overrode several opposing voices to approve Moss Landing, but those arguments linger in the air as PG&E brings these batteries to life.

Setting aside the wonky procedural complaints (see my "Showdown at Moss Landing" for more details), the main critique was that it wasn’t a necessary expenditure to deal with the problem at hand. The reliability need that Moss Landing was created to fix has already been dealt with, at least for 2019, by a $14 million transmission upgrade.

The commissioners didn’t stop there. Instead, they looked ahead to the market power the South Bay gas plants will exercise as possible retirements loom.

“Given that resources in this locally constrained area have market power and may retire at any time, there is risk of additional retirements in the future, which would reduce available capacity to meet Local Capacity Requirement (LCR) need,” the decision states. “The storage contracts presented in this [advice letter] will help mitigate the impacts of future retirements by adding capacity to the respective local subarea.”

For the ratepayer advocate, pushing the use case into the hypothetical future put the proposal on shaky ground.

We don’t know how exactly the 4-hour batteries will impact the market power of gas plants, said Julie Halligan, who manages the electricity policy and planning group at the Public Advocates Office (formerly known as the Office of Ratepayer Advocates).

“It’s not clear that these resources would be able to displace gas generation in the area,” she said. “If you don’t show us any analysis supporting what you’re claiming, it would be difficult for us to support it.”

This search for empirical proof clashed with the hopeful rhetoric of PG&E and the environmental groups backing Moss Landing, which cast it as a bold step toward California’s clean energy goals.

“They're trying to keep cost as low as possible without this looking toward the future,” Post said of the criticism.

The ratepayer advocates support grid storage as a tool for meeting California’s clean energy goals, Halligan said. They just want to see evidence that each individual project is a sound investment of ratepayers’ money.

If PG&E wants to build more projects like Moss Landing, it would do well to track its impacts on the surrounding gas plants, so it can provide more detailed evidence for future proposals. It would help if at least one of those at-risk plants actually shut down in the next few years.

Greenhouse gas tracker added

The final order did include a formal evaluation requirement: PG&E will report back to regulators annually on the greenhouse gas emissions impacts of the battery plant, a development that Halligan hailed as “an improvement.”

Operational choices dictate what sort of emissions profile the facilities will have. This created controversy for California’s Self-Generation Incentive Program (SGIP), which subsidizes small batteries for homes and businesses. The program has an environmental mission, but the batteries it subsidizes don’t always reduce carbon emissions.

“Behind the meter, the customer is the master, and sometimes that doesn't line up — sometimes that does increase GHG emissions,” Post said.

The stakes are higher for Moss Landing, given its enormous size. If the batteries charge exclusively from nearby gas plants, losing energy in the power conversion process, they’ll run dirtier than just using the gas power directly.

If they charge up on midday solar power (avoiding curtailments) and discharge to offset gas generation in the evening, they should reduce emissions overall. This is all hypothetical until the performance data comes in; that information will be enlightening for future mega-storage proposals.

In theory, Moss Landing won’t face the same pitfalls as the SGIP batteries, because it will participate in a market in which cheap power tends to be lower-carbon and expensive power tends to be higher-carbon. The only way to know for sure is to document actual performance.