How California can get to a reliable, 85% clean grid by 2030

Can California really rely on gigawatts of solar and wind power and batteries, plus long-duration energy storage systems and ​“firm” carbon-free resources like geothermal power plants, to replace the need for most of its fossil-fueled power by the end of the decade? And can it do so without driving power prices through the roof or exposing the state to the risk of major blackouts?

Yes, new modeling suggests — but the state is likely to be more successful if it dramatically ramps up offshore wind and geothermal power and depends less than previous forecasts have suggested on new utility-scale solar farms.

A better way to ramp up clean power in California?

Last week, California officials warned that the state faces many more summers of potential grid electricity shortfalls, casting the viability of its long-term grid-decarbonization goals into question. More clean energy is needed to mitigate the heat waves, droughts and wildfires intensified by climate change that are causing the threat of major blackouts. At the same time, solar and battery supply-chain disruptions are threatening to stall deployment of the clean energy resources the state has targeted to add over the coming decade.

But new in-depth modeling from energy analysis firms GridLab and Telos Energy indicates that despite these risks, California can reliably increase its annual share of carbon-free electricity from about 55 percent today to 85 percent by 2030. California’s most recent procurement goals are roughly on target to reach a similar level of year-round clean energy by setting goals of tens of gigawatts of solar and wind farms and batteries by 2032, along with gigawatts of offshore wind farms, pumped-hydro storage facilities and geothermal power plants.

The remaining 15 percent of the state’s annual electricity needs by 2030 could come either from in-state gas-fired power plants or from electricity imported from other states, the GridLab-Telos modeling indicates. That would mean the state could retire roughly one-third of the gas power plants it had been expecting to still be online in 2030, beyond those already set to retire over the next few years, all while maintaining a level of reliability that could minimize the risk of rolling blackouts like those California experienced in the summer of 2020.

“California can reach an 85 percent clean electricity standard and can do so while keeping the lights on,” said Priya Sreedharan, program director at GridLab, in an interview. ​“There are a lot of details that go into how we interrogate that question, but that’s the bottom line.”

But in a nod to California’s ongoing struggles, the analysis also finds that ​“the pace of deployment required to reach this mark by 2030 will be difficult to sustain without new policy.” Specifically, it warns that policy changes are needed to prevent overreliance on solar and batteries, which could lead to a less reliable grid.

A companion policy report from think tank Energy Innovation calls for increasing the amount of offshore wind power to provide energy in the evenings when California’s grid faces the greatest stress, as well as adding more geothermal power plants that can run around the clock and provide valuable grid stability. It also suggests boosting energy-efficiency and demand-response programs to reduce demand during hot summer evenings.

And California should target clean-energy deployments that would help close down gas power plants in disadvantaged communities that suffer disproportionate harms from the pollution they cause, the report contends.

Mike O’Boyle, director of electricity policy at Energy Innovation, said the report provides ​“more confidence that California can exceed its current statutory goals for clean energy by 2030.”

But the report also highlights that current planning methods might not reveal the most cost-effective and reliable mix of resources the state should be supporting today to ensure they’re built and ready to serve by the end of the decade, he said. Finding that optimal mix requires ​“having a more iterative planning process, adding in these [more diverse] resources and changing the portfolios to assess the trade-offs,” he said — something that the GridLab and Telos modeling does by testing the viability of three different scenarios for California’s future energy mix.

California may have trouble meeting its clean-energy goals under the current policy

GridLab and Telos modeled three futures for California to demonstrate these trade-offs. The first is a ​“base case” that closely matches the resource mix to be developed under current state policy. The second is a ​“diverse clean resources portfolio” that increases offshore wind and geothermal power compared to current policy. The third is a ​“high-electrification” portfolio that builds on the diverse clean portfolio but also incorporates increased demand from electric vehicles and electric building heating. The state now has policies in place that promote EV adoption and building decarbonization, but they haven’t yet been integrated into state grid-planning processes.

The base case hews closely to the mix of clean energy procurement targets laid out in a February decision from the California Public Utilities Commission as part of its long-range integrated resource plan. The CPUC decision calls for adding about 17.5 gigawatts of utility-scale solar power and 3.5 gigawatts of on-shore wind power by 2032, along with 13.6 gigawatts of battery storage, 1.7 gigawatts of offshore wind, 1.2 gigawatts of geothermal power and 1 gigawatt of long-duration pumped-hydro energy storage.

February’s decision boosts the state’s clean-energy deployment ambitions from prior targets set in 2020, and it is projected to achieve 86 percent carbon-free electricity on an annual basis by 2032. But according to some climate advocacy groups, it may not be aggressive enough to get the state on track to reach 100 percent carbon-free energy by 2045, the target set by a 2018 state law.

California has been breaking records when it comes to supplying nearly 100 percent of its electricity from solar and wind power for brief moments during sunny and windy spring afternoons. In March, state grid operator CAISO saw its supply of renewable power briefly peak at 96.4 percent of its overall demand. Last month, it hit a new record of 97.6 percent, and last week it set yet another record of 99.9 percent.

But California still depends on the roughly 40 gigawatts of gas-fired power plants that provide both baseload and peaking power reserves during other parts of the year. As of 2020, California got about 54 percent of its annual electricity supply from renewable energy, hydropower and carbon-free nuclear power, according to data from the California Energy Commission.

Meanwhile, California is facing grid power shortfalls during hot summer evenings after solar power fades away but air-conditioning electricity use remains high. On two evenings during a regionwide heat wave in August 2020, state grid operator CAISO was forced to call on utilities to impose rolling blackouts that left hundreds of thousands of customers without power for up to two hours at a time.

Governor Gavin Newsom (D) imposed emergency measures to forestall the need for more rolling blackouts later that year. Drought conditions and wildfires that brought the state close to similar grid emergencies in the summer of 2021 led to further emergency measures meant to reduce the risk of summer shortfalls this year and in future years.

But despite these emergency measures, California officials warned Friday that the state still faces a shortage of about 1.7 gigawatts of power to meet potential peak summer evening demands this year. Wildfires or regionwide heat waves could widen that gap by several thousand more megawatts, CAISO Chief Operating Officer Mark Rothleder said.

Delays in building new solar and battery projects are worsening this midterm outlook, said Siva Gunda, vice chair of the California Energy Commission. The state is still about 600 megawatts short of hitting a target to procure 3.3 gigawatts of new clean resources by 2023, as ordered by the CPUC in 2019. The shortfall is largely due to Covid-related supply-chain disruptions and compressed timelines for completing projects, he said.

Ongoing supply shortages and rising prices for solar panels and lithium-ion batteries have been compounded recently by the U.S. Department of Commerce’s investigation into imposing tariffs on solar panels imported from Cambodia, Malaysia, Thailand and Vietnam, Gunda pointed out. These combined impacts could lead California to be 1.6 to 3.8 gigawatts short of the more than 13 gigawatts of solar and batteries it has ordered to be deployed through 2025.

CPUC President Alice Reynolds highlighted an array of emergency measures instituted in December to mitigate the risk of shortfalls, from programs that pay customers to reduce power use during grid emergencies to extending the planned retirement dates of a number of coastal gas-fired power plants.

“We know that reliability is going to be difficult in this time of transition, and we know that climate change is continuing to put Californians at risk,” she said. 

Last month, Newsom floated another possibility: keeping the Diablo Canyon nuclear power plant open past its planned 2025 closure date. The plant currently provides about 10 percent of the state’s annual electricity supply. But it’s unclear how the state might reverse course at this late juncture, or how Diablo Canyon’s owner, utility Pacific Gas and Electric, would be able to cover the costs of keeping it open and unwind the complicated closure settlement agreement it reached with communities and environmental groups.

The case for aggressive action to build a more diverse portfolio 

These are some of the reasons why Energy Innovation is highlighting the value of the ​“diverse clean resources portfolio” modeled by GridLab and Telos. That portfolio includes 4 gigawatts of offshore wind power and 2 gigawatts of geothermal power by 2030.

This is quite a bit more offshore wind than the 195 megawatts called for in the CPUC’s latest resource plan, and it’s almost double the 1.2 gigawatts of geothermal power being called for. But reaching these aggressive targets could significantly reduce how much solar power and batteries California will need by decade’s end to maintain grid reliability, according to the modeling.

“You can think of diverse clean resources bringing two types of benefits,” Sreedharan said. ​“One is on the reliability side, and the other is on the deployment-feasibility side.” 

Offshore wind farms would generate more power in the evenings when California’s grid is under the greatest stress, while geothermal power could be relied on around the clock, she said. They could replace a much larger amount of solar power and batteries that would otherwise be needed to fill those gaps, as the chart on the left below indicates.

More offshore wind and geothermal power could cut nearly in half the utility-scale solar needed by 2030 compared to the base-case scenario, as well as cutting about 4 gigawatts of battery storage needed to shift that solar power from when it’s produced to when the grid needs it, GridLab’s Sreedharan said. That could reduce the risk that continued delays in building enough solar and battery farms to meet the state’s goals will leave the grid short of what it needs to maintain reliability, she said.

That increase in future reliability could also allow California to shut down up to 11.5 gigawatts of gas power plants, Energy Innovation’s O’Boyle said. Those closures would be above and beyond the roughly 4 gigawatts of gas power plants already under state mandate to close in the coming years to protect ocean ecology, but whose retirements have been delayed for fear of undermining grid stability. Speeding the closure of gas plants is important not just to fight climate change but also to reduce the environmental and health harms those plants impose on nearby communities, he said.

But current state policy doesn’t support such a rapid buildout of offshore wind and geothermal, he said. The gargantuan buildout ordered by the CPUC last year to help replace Diablo Canyon only calls for 1 gigawatt of geothermal power or other ​“firm” carbon-free resources by 2026, half the amount the new report’s diverse-resources scenario calls for by 2030.

And legislation passed last year to promote offshore wind development calls for 10 gigawatts by 2040, but it doesn’t provide aggressive-enough financial incentives or offshore transmission grid development support to meet the 4-gigawatt-by-2030 target in the GridLab-Telos model, the report points out.

Nor does existing state policy support the buildout of enough transmission grid capacity to ensure that current state procurements can be rapidly brought online, according to the report. CAISO’s recently released long-range transmission plan calls for about $30 billion in investment over the next 20 years. But given that solar and battery projects already face years-long backlogs in connecting to CAISO’s grid, it’s unclear whether near-term transmission buildout can support even faster growth in coming years.

More aggressive actions to encourage households, businesses, factories and farms to reduce peak power use could help mitigate the risk that California won’t be able to deploy enough solar power and batteries in time to maintain grid reliability over the coming decade, said Daniel Esposito, senior policy analyst at Energy Innovation.

Better coordination outside the state’s borders could also help, he said. States across the U.S. West are facing the same heat, drought and fire risks as California, and an increasing number of them are setting carbon-reduction goals that will impose similar challenges in devising a reliable mix of clean energy resources to meet them, he noted. Efforts to link Western states into a regional transmission organization could bring more certainty to California energy planning, helping to ensure that power imports are ​“available on those high-demand days when the grid is most stressed,” Esposito said.

All of these steps could be critical to allowing California to electrify its transportation and buildings, Sreedharan noted. GridLab and Telos’ ​“high-electrification” scenario encompasses the state’s grid needs if it were to successfully transition to 100 percent electric vehicle sales by 2035 and decarbonize building heating at a more rapid pace than is anticipated in current state grid planning.

Modeling of this scenario indicates that if California were to attempt to support those types of electrification needs without expanding its supply of offshore wind and geothermal power, it could require building up to 40 gigawatts of utility-scale solar — a ​“level of deployment that could be challenging to achieve.” A study from Southern California Edison, the state’s sole all-electric investor-owned utility, revealed similar gaps between California’s EV and building decarbonization policy and its current energy plans.

O’Boyle noted that the high-electrification scenario is already baked into state policy, by 2018 law AB 3232​’s mandate to reduce greenhouse gas emissions from residential and commercial buildings by 40 percent below 1990 levels by 2030 and by Newsom’s 2020 executive order calling for an end to sales of gasoline-powered light-duty vehicles by 2035. ​“We don’t want to be underprepared for electrification,” he said.

A model to help California plan its clean energy future? 

A number of reports have assessed the economic potential for reaching high levels of renewable energy in different states and across the country. GridLab participated in one such study in 2020 that indicated the U.S. could reliably and cost-effectively run a power grid with 90 percent carbon-free energy and 10 percent gas-fired power by 2035.

But the new GridLab-Telos study provides a more rigorous analysis of the grid reliability factors that will come from such a transformation than other studies conducted in the U.S. to date, Sreedharan said. In fact, the unique methodology it employed to measure the energy resource portfolios against the grid stresses California faces could be a valuable approach for the state’s agencies to adopt, she added.

The key to this value is the ​“stress-test methodology” that GridLab and Telos used, she said. GridLab and Telos used the same modeling software tools that the CPUC uses. Those include the Resolve capacity expansion model to create the portfolios of resources that can achieve the 85 percent clean energy targets, as well as the Plexos least-cost expansion planning model to determine the reliability of those resources across a range of potential futures by modeling the hourly commitments and dispatches of all those resources across the grid system interconnecting the U.S. West.

But rather than subjecting a single portfolio to a range of probabilistic analyses, GridLab and Telos concurrently ran their three portfolios through a specific set of predictable ​“stress tests,” such as the risk of droughts reducing available hydropower, wildfires disrupting specific transmission lines connecting California to other states, or coal plants being retired across the West. A specific test was conducted to emulate the same Western heat wave that led to the August 2020 rolling blackouts.

The distinction between these two kinds of analyses actually matters a lot for resource planning because of the enormous computing power required to run these kinds of models, Sreedharan said. The probabilistic analyses used by California agencies today are so computationally intense that they can’t be conducted at reasonable cost or time for more than a handful of scenarios, she explained.

GridLab and Telos’ stress-test methods are far less costly and time-consuming from a computing standpoint, she said. To be clear, they don’t offer the same level of certainty that the more hefty probabilistic methods do, she noted. But they can be used as a precursor to those methods to test out a far broader set of potential portfolios.

“The kinds of insights we gained through our analysis of the different portfolios could only be gained by this step, where you’re testing different portfolio compositions against these different what-if scenarios, these stress-test factors,” she said.

Another valuable insight gleaned from the GridLab-Telos method is the need for more precise modeling of the grid impacts of closing specific gas-fired power plants, O’Boyle said. The GridLab-Telos modeling was able to show that California could retire about one-third of those plants by 2030 and run the remaining ones less frequently than it does today.

But it wasn’t able to drill down into the detail that would indicate which of those plants could be closed without potentially risking grid instability in certain constrained parts of the grid, he noted. This issue of local resource adequacy is a significant challenge for CPUC, CAISO and the state’s utilities and community choice aggregators, and it has been a major sticking point for efforts by communities and environmental justice activists to close power plants that are the source of harmful local air pollution.

“About half the gas plants in California are in or near disadvantaged communities” or in regions with air pollution that exceeds standards of public health and safety set by federal and state authorities, O’Boyle said. Those plants provide roughly the same amount of generation capacity as the GridLab-Telos study indicates could be retired if the issue of local grid reliability weren’t a factor, he added.

Back in 2016, community activists in Oxnard fought back against a plan to build a gas power plant in the region, leading the CPUC to oversee a process that led to the plant being canceled and replaced by a mix of batteries, demand-side resources and grid voltage support technologies.

That’s a model that could be replicated to find ways to close existing plants, O’Boyle said. ​“Agencies can begin the process of examining gas plants in disadvantaged communities on a more fleetwide basis,” he said, as well as pinpointing new clean-energy procurements that could come online ​“where there’s potential to displace that capacity. That’s an exercise that needs to happen, and responsibility needs to be assigned to the utilities” to make it happen, he said.