Concentrated solar power's failure to gain momentum in U.S. markets is a signal that traditional resource valuations may be slowing the energy transition, a February CSP conference made clear.
CSP, which converts the sun's heat to electricity, was once dominant, then faded when photovoltaic (PV) solar, which turns the sun's light into electricity, plummeted in price. But unlike CSP, PV, even with batteries, cannot provide the long-duration, dispatchable generation that high-renewables power systems will need, conference participants said.
"Renewables are now mainstream and fossil fuels are the alternative," California Energy Commission (CEC) Chair David Hochschild told regulators, utility executives and analysts at the conference. With new zero-emissions mandates, "we will need a diversity of renewable resources to keep the system reliable, and we will need CSP, particularly, because of its long duration storage [potential]."
Ambitious 100% renewables mandates drive indiscriminate procurement of the lowest-cost renewable kWh, utility executives and regulators said. But the transitioning power system requires broader value, even if the per-kWh price is higher.
Restructuring markets, policies and utility planning to compensate investments in resources with a higher overall grid value, despite higher capital expenditures, will be necessary to deliver a reliable, low carbon power system, they added.
How CSP works
CSP uses mirrors to concentrate the sun's heat within a single point containing a heat-retaining fluid. The captured heat creates steam that, like conventional generators, drives an electricity-generating turbine.
CSP tower technologies, like the 110 MW Crescent Dunes Nevada project and the 394 MW Ivanpah California project, focus the sun's heat on fluid flowing through a tower's apex. Trough technologies, like the 280 MW Solana project in Arizona, heat liquid flowing through the focal points of trough-shaped mirrors.
Heat-absorbing fluids include water, as at Ivanpah, or a molten liquid that more efficiently holds heat, as at Crescent Dunes. Insulated tanks store the heated fluid for on-demand dispatch of electricity.
Utility-scale PV panels release electrons when exposed to the sun's light. The electricity flows to the grid or can be stored in batteries. Cost and regulatory barriers have largely limited cost-effective battery storage to four-hour durations, although battery stacks and alternative battery chemistries that deliver longer duration storage have been piloted.
In 2010, the U.S. had 0.4 GW of CSP and only 0.1 GW of utility-scale PV. But cumulative CSP installations reached only 1.7 GW by 2020, while falling panel costs led to the installation of 35.4 GW of utility-scale PV by 2020, Wood Mackenzie Senior Analyst for U.S. utility-scale solar Colin Smith emailed Utility Dive.
Since 2014, "CSP technology has been plagued with performance problems and a high price point," Smith said. It has had trouble competing with PV or PV plus storage, "even with the promise of 12-hour-to-24-hour continuous power."
Globally, policy incentives as well as high electricity prices have kept CSP's high capital and per-kWh costs from being a barrier. At the end of 2019, 90 CSP plants representing about 6,000 MW of capacity were in operation around the world, Hank Price, managing director of CSP developer Solar Dynamics, told Utility Dive.
Questions about CSP grew in the U.S. after Crescent Dunes' contract with Nevada's NV Energy was canceled in 2019 due to flaws in the molten salt storage system. But other countries have found success with that storage method — six of the 14 global tower projects and over 30 trough projects use molten salt fluid for storage.
Repairs are reportedly underway for NV Energy and the project will likely find another off-taker, Price said.
But many at the conference told Utility Dive that doubts about CSP's technical feasibility and the low cost of PV could make a new contract difficult to find. NV Energy declined to comment.
CSP's first barrier: Getting paid
Early CSP projects had capital costs that reached billions of dollars and their average levelized cost of energy (LCOE) was $0.21/kWh. Although the upfront capital cost is still high, the U.S. Department of Energy estimated CSP's 2018 LCOE, with 12 hours of storage, dropped to $0.098/kWh.
A 2019 contract price for CSP with storage in Dubai was reported at $0.083/kWh, significantly less than the Lazard-reported LCOE of $0.15/kWh or more for a natural gas peaker plant that its flexibility would allow it to replace.
Monetizing CSP will require new incentives that value its unique set of system benefits "instead of valuing the least cost resource," former Nevada utility commissioner Rebecca Wagner told Utility Dive. Advocates need to make the case to regulators that "CSP may cost more, but its storage allows using renewables overgeneration to flatten peak demand and fill gaps when wind or PV are not producing."
If CSP advocates make that case, regulators will need to reconsider existing market rules and regulatory structures, according to analysts at the conference.
"Markets and regulatory structures were set up around a different set of electricity resources than those that will dominate the future," Energy Innovation VP Sonia Aggarwal told Utility Dive at the conference. "That has cascading implications."
Existing incentives assure compensation to a generation source for the value its energy delivers to the power system in cents per kWh, Center for Energy Efficiency and Renewable Technologies (CEERT) Senior Technical Consultant Jim Caldwell told Utility Dive. "If CSP got paid for all its services, it might be able to compete, but that would require different market rules."
CSP could play a role in controlling the rising cost to ratepayers of meeting peak demand ramps, but market rules define procurement separately by energy, capacity and ancillary services, Caldwell said. "CSP doesn't win in separate solicitations. Only pricing all three together makes CSP the least cost solution, especially now because capacity and ancillary services are becoming more valuable."
The need to accurately value the services resources deliver is an emerging challenge for system operators across the country, Caldwell said. "It's premature to value specific services now, and it might be inaccurate because the need for them is still limited, but it's not too early to think about how we value them so the method is ready when it is needed."
Better modeling and planning
The modeling done for integrated resource planning can identify the higher value of CSP and other low-carbon technologies that offsets its higher costs, National Renewable Energy Laboratory (NREL) Senior Analyst Trieu Mai told Utility Dive. "Portfolios can then be re-evaluated with production cost modeling for reliability and resource adequacy in an iterative process."
The models consider "essentially infinite options" to "make tough choices about how the system operates today and how it might operate in the future," he said.
Modeling can compare CSP projects, especially hybrid versions combined with PV or natural gas, to conventional peaker plants, and identify what plant designs fit best in a highly renewable grid.
Variable wind and solar interconnected with the grid through inverters can cause fluctuations in voltages and frequencies that create challenges for protecting system stability, power system consultant Debra Lew, a former NREL Engineer and GE Technical Director, told Utility Dive. "But there is a future coming with a lot of variable wind and solar, and there has not been sufficient planning for that future."
Work on using zero-emissions resources not interconnected through inverters to meet capacity and system balancing needs is advancing, but work on using them for system stability is not, she said. Instead, system operators are investing in hardware to correct and react to voltage and frequency fluctuations.
Zero-emission dispatchable resources like CSP, stored hydro, and geothermal can provide system stability and offer more value than a utility investment in hardware, because the utility gets clean energy with the expenditure for system stability, Lew said. But there are few incentives or market mechanisms to compensate utilities and developers for those investments, so "PV and batteries will be built because they are cheap and fast" and hardware will remain the main tool for system stability.
CSP's second barrier: Short-term thinking
In the future, planners may have "supermodels" that identify long-term value in high capital expenditures, but today's models often lead to short-term solutions, Lew said. "The alternative is to identify what resources will be needed in zero emissions scenarios and build them now, but that would require the long-term planning we are not doing."
The size of investment in resources like CSP and the time it takes to develop them are also "very daunting" for a regulated utility, Arizona Public Service (APS) Vice President for Public Policy Barbara Lockwood told the conference.
There are many new uncertainties about demand because of the emergence of new load serving entities (LSEs) and rapidly increasing customer adoption of distributed resources and energy efficiency, she said. As a result, APS has "defaulted to much smaller investments" like small PV plus storage installations that require lower capital expenditures and shorter times to develop.
Southern California Edison (SCE), in response to similar market signals, has also shifted procurement away from high capital expenditures, said SCE VP for Energy Procurement Bill Walsh. Smaller procurements "help control or smooth the system or fill in gaps as generation changes."
Utilities prefer "the incrementalism of [requests for proposals] for 50 MW or 100 MW investments" because it seems less risky, former Nevada commissioner Wagner said. "But that is short-term thinking and avoids evaluating and recognizing system needs over the longer term. To achieve 100% clean energy goals, we need to start thinking out of the box and explore all opportunities."
Electric utilities are moving "from what resource we want to what service we need and when we need it" and more suitable contracts might "be more about capacity than about energy," APS's Lockwood agreed.
Electrification and zero emissions mandates could increase the California system's load by 75% or more and that will require "big investments" in new generation, California Independent System Operator CEO Steve Berberich told Utility Dive. Counterparties, like investor owned utilities and new LSEs, for those investments should be part of designing new contracts that can secure the capital needed.
Contracts must support projects that offer "services from renewables that we got from the thermal fleet," Berberich said. A new accounting process being developed by California regulators should lead to metrics that can be used for those contracts, he added.
Value is becoming "a key part of planning," California Public Utilities Commission (CPUC) Energy Division Head Edward Randolph told Utility Dive.
The effective load carrying capacity of PV without storage was set at 17% in a 2019 CPUC ruling, which means only 17 MW of a 100 MW solar procurement counts toward meeting peak demand. That will likely increase procurements of resources that can better meet system peaks.
That type of "fundamental rethinking of what system value is and how to define it quantitatively is needed for zero carbon resources with high upfront capital costs," Aggarwal said.
"A new approach may seem somewhat visionary because market mechanisms must change to enable all resources to compete on equal footing for compensation for delivering energy, flexibility, and grid services."
Correction: A previous version of this article incorrectly conveyed two statements from power system consultant Debra Lew regarding variable wind and solar as well as work on using zero emission resources. In addition, the global capacity of CSP plants at the end of 2019 was incorrectly stated, It was 6,000 MW. The article has been updated.