When California's lawmakers mandated 60% renewables by 2030 and targeted zero emissions by 2045 in Senate Bill 100, passed in September, they took on never-before-answered questions about reliability.
One of those questions is how to keep the lights on without natural gas at the most supply-constrained times and places. Instead of a capacity market, which many systems use to meet these spikes, California imposes a resource adequacy (RA) requirement on load serving entities (LSEs) that adds 15% extra generation to their portfolios.
The natural gas generation that dominates today's RA supply cannot be part of a 100% emissions-free future. With its investor-owned utilities (IOUs) on track to reach 50% renewables by 2020, policymakers want to move natural gas out of the RA mix, despite resistance from natural gas interests.
"The fight in commission proceedings over natural gas is where the transition to a zero-carbon grid will happen," according to former utility executive Jim Caldwell, technical director for the Center for Energy Efficiency and Renewable Technologies (CEERT). "We know where it will end, but not how to get there."
California's natural gas fleet supplied only 33% of the state's 2017 electricity and is not growing, according to an August paper on California's natural gas from the Union of Concerned Scientists (UCS). But almost half of natural gas plants serve supply-constrained times and places and will continue to be needed for RA unless clean energy alternatives can replace them.
That has created a heated debate between policymakers. Some side with natural gas plant owners who say using unproven alternatives risks letting the lights go out. Their opponents say portfolios of distributed energy resources (DER) are ready to meet the most challenging demand spikes. And, they say, keeping natural gas owners in the RA business risks not achieving the state's clean energy targets.
Three types of RA
The 2000-2001 California energy crisis was caused, in part, by independent power producers (IPPs) withholding supply during energy market demand spikes to drive prices up. In response, California lawmakers decided to not rely on market forces for critical capacity. The RA program ensures the California Public Utilities Commission (CPUC) oversees system reserves.
The CPUC requires LSEs to prove procurement of their required RA, which is divided into three categories, well-ahead of when it will be needed. The bulk is system RA, which is each LSE's share of the capacity that might be necessary to meet California's peak load.
Local RA is each supply-constrained locality's minimum local capacity requirement (LCR) for energy needs under extreme weather or transmission outage conditions, according to the CPUC's 2017 Resource Adequacy Report. Because LCR is for areas with a more constrained supply, its average price was 40% higher than that for system RA in 2017.
When LSEs fail to meet their local requirements, the California Independent System Operator (CAISO) procures on their behalf through higher-priced "last resort" mechanisms. Together, they were 13% to 20% of the 2017 RA procurements.
A flexible RA product was added to LSE obligations in 2015 in response to the state's increasingly sharp evening demand period peak. To date, flexible RA makes up a small part of RA capacity.
California's almost 200 utility-scale natural gas plants have about a 39 GW capacity, according to UCS. Many are peaker or combined cycle turbines that can provide dispatchable generation to smooth renewables' variable supply curves and meet unexpected system events. They therefore remain the primary RA for local and emergency needs.
"To transition to a low-carbon future, support for natural gas has to stop because, even if they don't run often, they get paid a lot of money to show up, and that money is not available to alternative resources."
Jim Caldwell
Technical director, CEERT
But some plants in California's aging natural gas fleet now require costly emissions control or maintenance upgrades. Plant owners say the standard one-year RA contracts they now rely on for compensation do not provide the long-term certainty to finance upgrades.
They want California regulators to restructure RA compensation to allow them to keep the plants in service. Opponents say that would lead to continued reliance on natural gas and leave little opportunity for proving DER portfolios' viability as RA.
"All the new natural gas plants authorized in the last 10 years have been for local capacity needs," CEERT's Caldwell said. "To transition to a low-carbon future, support for natural gas has to stop because, even if they don't run often, they get paid a lot of money to show up, and that money is not available to alternative resources. That makes LCR procurement the critical battle."
The local RA challenge
"California already has too much natural gas and the first priority is making sure it continues to retire," Laura Wisland, UCS Western Energy Director and co-author of the paper on California's natural gas told Utility Dive. "In some places clean technologies can replace gas plants as local RA, in some places they cannot."
CAISO is releasing a study of LCR areas in November that will identify where replacing natural gas is feasible and cost effective, Wisland said. The study will consider transmission upgrades, utility-scale renewables plus storage and DER portfolios as alternatives.
The only current challenge for RA is meeting local needs, CPUC Energy Division Director Edward Randolph agreed. But much of that comes from confusion over who is responsible for RA procurement.
Until recently, long-term energy procurements by California's three dominant IOUs, Pacific Gas and Electric (PG&E), Southern California Edison (SCE) and San Diego Gas and Electric, met all system needs. But new LSEs, including community choice aggregators (CCAs) and Direct Access providers, have complicated things.
"CCAs don't have the ability or willingness to commit to longer-term natural gas resource adequacy agreements."
Edward Randolph
Director, CPUC Energy Division
Because the new LSEs are taking away much of their load, the IOUs have all the local RA they need and are not willing to make new commitments, Randolph said. "CCAs don't have the ability or willingness to commit to longer-term natural gas resource adequacy agreements."
The most recent integrated resource plan (IRP) proceeding assumed existing natural gas plants currently meeting local needs will stay online, Randolph said. But some of those plants face increased operating expenditures or expensive environmental upgrades that their owners say make it uneconomic to keep them in service without extended compensation guarantees.
"That is why proposed RA reforms would introduce longer term RA contracts for plants that meet local needs," Randolph said. But with both IOUs and CCAs not procuring, it is unclear who would commit to those contracts.
Natural gas plants meet "a significant amount" of local RA and will continue to be needed for that, according to Jan Smutny-Jones, former CAISO board chair and current CEO of the Independent Energy Producers Association (IEPA). "But short-term contracts do not provide the financial certainty to make long-term upgrade investments, which is why natural gas plants are closing."
"Some gas companies are crying wolf when they say they need longer contracts to stay in service, and others are telling the truth."
Laura Wisland
Director, UCS Western Energy
CAISO has been using the higher priced last resort contracts more in the last two years, which indicates the RA program "is missing its mark," Smutny-Jones said. The numbers agree. In 2017 alone, 3,851 MW of older natural gas capacity retired and only 438 MW of new capacity was added, according to the CPUC RA report.
"A three-year to five-year forward RA program would give plant owners enough certainty to make investments that would keep those reliability resources online," Smutny-Jones said. "The product could also support investments in demand response and storage. Any resource that is competitively procured and meets the RA need should be paid to do it."
"Some gas companies are crying wolf when they say they need longer contracts to stay in service, and others are telling the truth," UCS's Wisland said. "Contracts for all resources that provide RA might be better, but no contract should be longer than three years because the market is so dynamic that something we don't think is possible may be possible in five years."
An alternative in storage and solar
UCS found 23% of the combined cycle capacity and 24% of the peaker capacity, or 28 of California's 89 operating flexible plants, "could be retired as early as 2018 without negatively affecting grid reliability." The more ambitious emissions reduction target which SB 100 provides is likely to drive much greater use of battery energy storage and lead to the retirement of 87% of California's peaker capacity, the report added.
But to fully kick its natural gas habit, California has to do more than build storage, UCS added. It must shift demand away from the evening peak, build more storage for its solar over-generation, link to a wider grid to access out-of-state renewables and use DER portfolios in place of the natural gas generation now serving local RA.
The question is how to meet reliability needs with as little natural gas as possible, UCS's Wisland said. Solar plus storage, when cost-effective and not limited by battery duration, may beat longer-term natural gas contracts in the market. Policies that support alternatives or limit emissions may also be necessary.
"The most cost-effective solution will be a portfolio of distributed resources," Wisland said. RA compensation may be inadequate to support a DER portfolio "but it can also be compensated for serving customers' behind the meter needs."
Both Randolph and Caldwell described recent successes in using DER portfolios to meet local RA needs. SCE's preferred resources pilot, initiated in response to the 2013 closure of the utility's San Onofre nuclear plant, is farthest along. It has 112 MW of DER in place and is on track to have 265 MW serving local needs by 2022, according to SCE Senior Advisor Sergio Islas.
"Too many people still say reliability is spelled G-A-S, but solar plus storage prices are now low enough to compete."
Jim Caldwell
Technical director, CEERT
The CPUC has approved two newer proposals. SCE will use a 125 MW DER portfolio to replace the Puente natural gas plant in Oxnard and the Oakland Clean Energy Initiative, being developed by PG&E and East Bay Clean Energy will use a 45 MW DER portfolio to replace an aging diesel plant in Oakland.
PG&E's proposal to replace the 567.5 MW Metcalfe natural gas plant with a DER portfolio made up primarily of battery storage is scheduled to be decided by the CPUC in November, Randolph said. "We are taking almost every opportunity to meet local capacity needs with preferred resources," he added.
Natural gas plant owners are fighting back by saying they can't keep their plants online with one-year contracts, CEERT's Caldwell said. "They say subsidies to renewables are driving them out and putting reliability at risk. But distribution system operators must start working with DER portfolios or there will be no change."
To meet the SB 100 zero emissions goal, it will be necessary to burn half as much natural gas in 2030, Caldwell said. If existing plants are left in operation, it will keep new DER technologies off the grid.
"Too many people still say reliability is spelled G-A-S, but solar plus storage prices are now low enough to compete," Caldwell said. "It is time to face the future by planning orderly natural gas plant retirements and creating incentives for alternatives. Getting to 100% in 2045 must start now. It is not a problem of technology or cost, it is a problem of people and business practices from the past."
