The staff of the California Public Utilities Commission (PUC) has released a proposal aimed at fixing the emission shortfalls of the state's premiere incentive program for behind-the-meter (BTM) energy storage, suggesting a state-wide partnership to develop a greenhouse gas (GHG) signal.
One of the top policy goals of the Self Generation Incentive Program (SGIP) was to reduce GHG emissions. But evaluations have shown that the SGIP storage projects have, in fact, increased greenhouse gas emissions.
As the state is grappling with this reality, California lawmakers extended the program at the end of August by passing SB 700. The bill extended SGIP for another five years and refreshed the $800 million in funding for the program. The governor has yet to sign the bill.
Depending on the outcome, revisions to the SGIP could have application at other GHG reduction programs and incentives across the country.
"The issue is important because no other state is facing this yet," Alex Morris, vice president of policy for the California Energy Storage Association (CESA), told Utility Dive.
States like New York, Massachusetts and New Jersey are touting the GHG reduction benefits of their energy storage policies. California did the same only to discover that the program was adding to, not reducing, GHG emissions. "Other states and commissions are trying to figure this out, and they could leverage what California" learns from this process, Morris said.
Starting off on the wrong metric
When the SGIP program was set up, it was assumed that California's rate structure would encourage the operators of energy storage facilities to charge their systems when solar power flooded the grid with cheap power during the middle of the day, and to discharge their batteries when rates rose in the evening as gas-fired generation ramped up to fill the gap as solar power faded. That was not what happened.
The only metric storage operators are required to meet under SGIP is a lifetime round trip efficiency (RTE) level of 66.5%. That requirement was put in place as a proxy for GHG reductions and to restrict the eligibility of batteries designed for emergency backup purposes.
As more polluting fossil generation sources typically have higher marginal costs, it was assumed that optimizing a storage system for avoiding high time-of-use (TOU) or demand charge rates would bring GHG reductions.
An Itron report on SGI found that theory holds true some of the time, but not all of the time. If a battery charges only when zero-marginal-emissions resources are on the margin, the result is a net reduction in total emissions, regardless of the RTE. But Itron's analysis also found that "when dispatch is optimized to maximize customer bill savings, almost all projects increase GHG emissions regardless of their RTE." That is because there is an "imperfect alignment between the retail rate signals and actual marginal emission rates."
"We have a strong agreement: RTE is not a good proxy."
Policy director, California Solar & Storage Association
Itron concluded that "RTEs alone are not sufficient to guarantee GHG emissions reductions." Or, as the PUC staff said it in its recommendations: TOU rates have not coincided with times of high grid emissions, Instead, retail rates incentivize customers to prioritize demand charge management over TOU arbitrage.
In essence, the fix is a simple one. PUC staff is recommending incorporating a GHG signal into SGIP rules. As usual, the devil is in the details. "We have a strong agreement: RTE is not a good proxy," Brad Heavner, policy director for the California Solar & Storage Association, told Utility Dive. "Let's live by the actual performance, and it will solve the problems. We are on board with the general concept. We just have to work out the details."
So, the question then becomes: if TOU rates do not reflect GHG intensity, what can be used as a GHG signal? PUC staff's solution is to have SGIP administrators contract with WattTime or another qualified entity to provide a GHG signal that reflects real-time, marginal GHG emissions rates at five and 15 minute intervals.
WattTime is a non-profit subsidiary of the Rocky Mountain Institute that was formed to find a way to determine how clean the energy being consumed at any given time is. It was formed to tackle that question for energy storage, but also for electric vehicles, water heaters and "anything that consumes electricity and is flexible about when it uses energy," Gavin McCormick, WattTime's co-founder and executive director, told Utility Dive.
The basic concept is simple: it is like a locational marginal cost for emissions, McCormick said. The GHG signal WattTime devised is embedded in software. It is like an app, he said, that can be added to the algorithm that a battery operator uses to run its storage system.
The WattTime app is constructed using Environmental Protection Agency data on actual emission and generation output from individual power plants. That data is compared with real time prices in the wholesale market and adjusted for factors such as weather, temperature, wind speeds and fuel costs. That data is then used to come up with day-ahead emissions for any point on the grid. WattTime currently supplies emissions forecast that look ahead 24 hours. It is working on extending that forecast to 72 hours.
McCormick said he wants the technology to be available to anyone who wants to use it. The company offers the use of the data for sale. But in California, WattTime is taking a different approach. The company intends to offer an open source version of the software for the California market, in part because it is an easier market to forecast because there are fewer variables.
If the price of natural gas is known, it can be used to calculate the marginal emissions from a gas plant. If the marginal price of power is lower than that, the power is coming from a cheaper generation source and in California that means a renewable power source because the state does not have coal or nuclear power.
That same calculation is harder in place such as the PJM Interconnection's wholesale market, where coal and nuclear power are also at play, said McCormick.
McCormick said WattTime's preliminary testing has shown that the GHG reduction benefits from using the company's algorithm could be even greater in more complex markets such as New York and Massachusetts.
The same software also can monitor GHG emissions in real time. That would give operators time to adjust their operating parameters if they are behind in PUC mandated GHG reduction targets, McCormick said.
"There is a balance that has to be struck to make sure the requirements aren't so onerous they don't create delays for customers."
Director of public policy, Sunrun
Under the PUC proposal, commercial storage projects would be under a performance based incentive with annual checks on performance. The first 40% of the incentive would be paid to operators upfront. The remaining 60% would be paid over five years, but the incentive would be reduced if a facility failed to reduce GHGs by 25 kilograms of carbon dioxide per rated energy capacity of the system. After five years, compliance with GHG performance would be done on a fleet basis, according to the PUC proposal.
The proposal also calls for operators whose fleets are found to increase GHGs to be temporarily suspended from submitting new applications for the incentive program. And operators could be required to purchase revenue-grade meters at a cost of about $2000 per project.
As with commercial storage projects, the PUC proposal calls for ending the RTE requirement for residential storage projects. Instead, residential storage projects would be required to enroll in a time-varying rate with peak starting at or after 4 p.m., pair with and charge at least 75% from a solar system, and to have a single-cycle RTE of at least 85%.
While many parties seem to agree about the value of using a GHG signal, some in industry are concerned about some of the other details of the PUC proposal.
"Broadly speaking, we don't want to be increasing GHGs," Melicia Charles, director of public policy at Sunrun, told Utility Dive. "There is a balance that has to be struck to make sure the requirements aren't so onerous they don't create delays for customers."
Sunrun is still examining the PUC's proposal, Charles said, but one of the issues the company is looking at is whether reporting of performance targets should be done on an individual or aggregated basis. "There is value in looking at the aggregated performance," she said, noting that an aggregated approach gives individual customers more flexibility in how they use their systems.
"There is a risk that the SGIP will face further slowdowns if industry and finance can't easily calculate their impacts to the customer."
Senior manager of regulatory affairs, Stem
That approach is seconded by Stem, which develops energy storage projects for commercial and industrial customers. From Stem's perspective, the PUC should preserve customers' ability to use energy storage for many uses, such as wholesale market contributions during heat waves or distribution infrastructure deferral, rather than simple energy arbitrage.
"We would like to see reasonable, flexible options for compliance that can accommodate these nascent but successful business models until the Commission can design electric rates that will give the appropriate GHG signal and incentive for emissions-reducing storage operations," Jim Baak, senior manager of regulatory affairs at Stem, told Utility Dive via email.
Customers' load shapes vary widely and, he added, SGIP projects have "a very tiny impact on GHG emissions, less than a small fraction of a percent of GHG emissions in California," he said.
For reference, in 2016, non-residential SGIP projects increased GHG emissions by 726 metric tons of carbon dioxide, according to the Itron report. That same year, routine GHG-emitting activities statewide totaled 429 million metric tons of carbon dioxide.
Meanwhile, until the PUC finalizes the new SGIP rules — not expected until the second or third quarter of 2019 — the SGIP market faces uncertainty. "There is a risk that the SGIP will face further slowdowns if industry and finance can't easily calculate their impacts to the customer," Baak said.