California has already postponed and even canceled plans to build new natural-gas-fired power plants in favor of distributed energy. But it hasn’t proposed to replace an existing power plant with them -- until now.
Early this month, the California Public Utilities Commission issued a resolution that would direct utility Pacific Gas & Electric to open competitive solicitations for DERs -- energy storage, mostly, but with room for other carbon-neutral “preferred resources” like demand response -- to cover the grid capacity and voltage needs now being served by three Northern California natural-gas-fired power plants.
Choosing which power plants are vital to keep the transmission grid running is the domain of the California Independent System Operator, not the CPUC. And the grid operator has designated the 580-megawatt Metcalf Energy Center south of San Jose, as well as the 47.6-megawatt Yuba River and Feather River generators as reliability must-run (RMR) resources.
These RMR contracts come with secure, high payments for running relatively few hours per year -- mainly during hot summer afternoons when air conditioning loads surge, demand spikes, and CAISO struggles to maintain the reserves to cover potential emergencies. Calpine, the owner of the plants, said they aren’t cost-competitive without RMR, and intends to seek that status with federal regulators, having passed over opportunities to bid their energy or capacity into resource adequacy or bilateral agreements.
But California regulators argue that these must-run contracts are too expensive, and that distributed resources can replace them at lower cost to ratepayers. It also states that Calpine’s contracts failed to go through CAISO’s procurement process for flexible capacity, which could “lead to market distortions and unjust rates for power” in years to come.
That, the CPUC contends, gives it broad authority to allow solicitation of resources that can "fill local deficiencies."
While the resolution doesn’t set a specific megawatt-hour target on what PG&E should procure, it does set some clear deadlines. If the resolution is passed, PG&E will have no more than 30 days to issue its first solicitations. Any resources the utility does procure will have to prove they’ll be ready in time to assure the RMR contracts won't be renewed in 2019.
Over the past five years or so, California’s push to replace natural-gas power plants with a portfolio of solar PV, batteries, thermal storage, demand response and energy efficiency has grown from an experiment to a cost-effective reality. The first round came in 2014, when Southern California Edison contracted for about 260 megawatts of DERs to help mitigate the loss of the San Onofre nuclear power plant -- although SCE and neighboring San Diego Gas & Electric were also allowed to contract for more than twice as much new gas-fired generation, against objections from clean energy groups and community environmental advocates.
The next test came in 2016 after the leaking Aliso Canyon natural-gas storage facility was shut down, leaving the region’s power plants open to being overwhelmed by winter and summer peaks in energy demand. This proceeding led to SCE procuring more than 100 megawatts of energy storage and demand response, delivered within six months of CPUC order. The Aliso Canyon deployment was a record turnaround for large-scale energy storage from Tesla, Greensmith and AES, and a proving point for Google’s Nest that its existing customer base could be tapped for grid needs.
The growing economic advantages of energy storage versus peaker plants came to a head this fall, when the California Energy Commission said it planned to reject NRG Energy’s proposed Puente natural-gas power plant near Oxnard, leading NRG to suspend its application process.
The decision came after analyses from GTM Research indicated that CAISO had used outdated 2014 cost figures in determining that energy storage would cost more than twice as much as the Puente gas-fired power plan. An analysis from the Clean Coalition found that a combination of solar and storage would be slightly cheaper than the $300 million proposal.
While the CPUC has yet to issue any guidance on whether it will ask SCE to seek out DER alternatives if the Puente power plant doesn’t get built, CAISO has indicated that an open solicitation would be the only way to find out whether storage, solar, demand response and other assets are cost-effective in its place.
Ravi Manghani, GTM Research’s director of energy storage research, noted that the CPUC’s latest resolution doesn’t face the same compressed timeline as the Aliso Canyon procurements did. But it is the first time California regulators have directed DERs as an alternative for an existing gas-fired power plant, rather than for an unbuilt plant.
This could put the CPUC’s intentions in conflict with the state's grid operator and the Federal Energy Regulatory Commission, he noted. CAISO will likely want to play a role in deciding whether distributed energy resources are reliable enough to meet its needs.
At the same time, the grid operator has been increasingly involved with the CPUC and utilities in developing alternatives for its capacity requirements and energy markets, through avenues such as the Demand Response Auction Mechanism. Last month, CAISO and PG&E unveiled another proposal to find distributed alternatives that can, alongside transmission substation upgrades, replace a diesel-fired peaker plant in Oakland. It's the first time CAISO has proposed inserting DERs into its transmission investment plans.
CPUC spokesperson Terrie Prosper noted that the resolution doesn’t set a strict timeline for how quickly procurements are expected. She estimated a few additional months to bring the resources on-line, compared to Alison Canyon.
California is one of a few markets where energy storage is competitive with gas-fired peaker plants today. But that's changing.
“I can’t see a reason why we should ever build a gas peaker again in the U.S. after, say, 2025,” Shayle Kann, senior adviser to GTM Research and Wood Mackenzie, said at Greentech Media’s Energy Storage Summit this month.
Current lithium-ion battery and balance-of-system costs are low enough to compete with simple-cycle combustion turbine peaker plants in select markets today, and will be broadly competitive by turn of the decade, he said.
Over the next 10 years, the U.S. needs to add 20 gigawatts of peaking capacity to its grid, more than 12 megawatts of it expected to come on-line between 2023 and 2027, giving energy storage more time to build an economic advantage.
Listen to Shayle Kann and Stephen Lacey talk about the economics of battery storage versus gas peakers on a recent episode of The Interchange.