Utility-scale wind and solar generation is now putting coal and nuclear plants out of business and is price-competitive with natural gas plants — but what would happen if rooftop solar was that cheap?
The retail residential electricity price varies regionally, but the national average in November 2017 was $0.13/kWh. There are ways to bring the cost of rooftop photovoltaic (PV) solar to $0.05/kWh by 2030, according to a new Department of Energy (DOE) paper. That value proposition could be appealing enough to change the already falling demand for grid-supplied electricity noticeably.
Getting to that low cost would require major business model innovation by solar installers. It would also require significantly more use of “integrated” solar roofing products still being developed. But if the solar industry makes those changes, it will offer electricity customers big new opportunities to which utilities may be forced to respond.
Rooftop solar is already growing fast and these projected low prices can only accelerate adoption. High penetrations can create challenges for utilities because of midday over-generation and an accentuated evening demand peak. But policies that encourage the use of battery storage and smart technologies can turn those challenges into opportunities.
The rise of solar
The levelized cost of energy (LCOE) for residential PV declined from $0.52/kWh to $0.15/kWh from 2010 to 2017, according to “Cost-Reduction Roadmap for Residential Solar Photovoltaics (PV), 2017–2030” from the DOE's National Renewable Energy Laboratory (NREL).
Effective partnerships between solar installers and roofers, now only beginning to emerge, could, at scale, bring the 2030 LCOE for PV on new roofs to $0.055/kWh. Similarly efficient and scaled up hypothetical partnerships between new home builders and installers could bring the 2030 new home LCOE for PV to $0.05/kWh.
NREL Solar Analyst and paper lead author Kristen Ardani said this thought exercise was to explore how solar’s 2030 LCOE could get to the target $0.05/kWh recently set by the DOE Solar Energy Technologies Office. The potential is there, but it would require significant changes in the solar installation, home building and roofing industries, she told Utility Dive.
“Between 2017 and 2030, an average of 3.3 million homes/year will be built or require roof replacement,” NREL estimated. Capturing even a small part of that technical potential of approximately 30 GW per year of rooftop PV “could have a significant impact on the evolution of the U.S. electricity system.”
Vikram Aggarwal is CEO for EnergySage, which offers a version of its solar market data and online marketplace to utilities. They are beginning to realize that “solar is disruptive, customers are looking to utilities for guidance with solar, and other companies are going after their customers,” he told Utility Dive.
The numbers in the NREL paper will only accelerate the utilities’ reaction, he said. “It could be a classic case of ‘if you can't beat them, join them.’”
Not everybody is convinced. Jim Piotrowski, director of solar generation for Arizona Public Service (APS), said the assumptions about price drops in the NREL report can be challenged. “There is only so much cost you can squeeze out of today's prices and we are getting to the limit,” he told Utility Dive.
The thought exercise
Ardani said there is growing interest by new home builders and roofers in the solar opportunity. “Given those trends, we wanted to see what the potential savings could be through 2030.”
The researchers first estimated 3.3 million rooftop solar installs/year and 30 GW/year of total estimated installed capacity in the U.S. from all existing solar suitable roofs.
From those figures, NREL cost modeling estimated today’s $2.66/watt PV installed cost could be reduced to $1.21/watt for replaced roofs and to $1.10/watt for new builds in 2030. Using those numbers and other data, the researchers calculated the LCOEs.
NREL Senior Solar Analyst and paper co-author Robert Margolis was also co-author on a recent and more comprehensive NREL assessment of U.S. rooftop solar. “The Ardani paper accurately represents the long run sustainable market for the residential sector,” he emailed Utility Dive.
Ardani's thought exercise modeled four options. Two “aggressive” alternatives assume a less efficient merging of solar installers' business models with those of roofers and new home builders. Two “visionary” models assume a more complete merging and “a higher level of innovation,” the paper reports.
All drive “significant installed-system price savings” from the 2017 benchmark of $0.151/kWh. Complete merging of solar installation with roof replacements provides slightly less of a 2030 'visionary-level' LCOE reduction than merging with homebuilders, to $0.055/kWh.
The projected 2030 LCOE reduction to $0.05/kWh comes from a 'visionary-level' merging of solar with new home builds. New home builders have the potential to get to higher levels of cost efficiency in things like labor and overhead than roof replacements can achieve.
About 65% of total installation savings comes from reduced soft costs. The biggest cost reductions come from the maturing market’s lower supply chain costs, combined home and solar sales and marketing costs, overhead costs, and costs of installation labor.
The researchers acknowledge four barriers to achieving the projected savings. First, Ardani said, the paper’s 'visionary' cases “assume a high level of business model integration across industries that doesn't exist today." It also assumes procurement, permitting and regulatory efficiencies that don't exist today.
Second, a “low-cost integrated PV and roofing product” will be necessary, the paper reports. Today’s integrated solar-roofing products are competitive in some cases, but realizing the paper’s 'visionary' models will require a more broadly competitive product. Developing it will be “very challenging” and require “significant investment,” the paper adds.
Third, the limited merging of solar with roofers and with new home builders to date must be transformed. “Homebuilders, PV installers and roofing companies will likely require clear examples of the benefits of business model integration before embracing it widely,” NREL acknowledges. Only fully integrated businesses lead to the necessary reductions in cost and regulatory hurdles.
Finally, "additional research is needed to clarify the cost savings that economies of scale could provide,” the researchers conclude.
“There are two things important to a utility audience here,” Ardani said. First, PV price declines are expected to continue. “As it gets more affordable for a larger number of homeowners, utilities can anticipate the need to process a higher volume of interconnections.”
More significantly, solar’s increasing cost-effectiveness “makes the business case for storage and other enabling technologies more compelling,” she added. “It raises questions about the utility’s role, about how the utility should value PV and PV plus storage, and about how the utility should be compensated for providing the distribution system.”
The view from the solar marketplace
EnergySage’s just-released Solar Installer Survey reports that demand for solar-plus-storage systems is already surging. “Nearly one in three solar shoppers are also interested in a home battery, according to respondents.”
Demand is rising even with battery costs ranging from $5,000 to $9,000 or more, which translates to $400/kWh to $750/kWh, and total system costs that are even higher, according to EnergySage. A new Tesla Powerwall 2.0 battery is $5,500. Supporting hardware adds another $700 and installation costs can be as high as $8,000 or more, EnergySage reported.
Declining solar-plus-storage price curves could make systems as common as refrigerators and air conditioners, Aggarwal said. “Utilities are increasingly interested in providing an online shopping platform to help guide their customers through their purchases.”
National Grid has successfully piloted the EnergySage solar marketplace platform for some of its Rhode Island customers, he said. It is preparing to open the platform to all its 440,000 Rhode Island and 1.3 million New York customers. New York State Electric and Gas is now piloting the platform for 20,000 customers.
The thinking is that it will increase customers’ satisfaction and maintain their loyalty, he added. Customers will then be more likely to turn to the utility when they think later about buying a broader suite of energy products and services.
“The next questions are whether the utility can generate significant revenue from those transactions,” Aggarwal said. “Our financial model shows that these marketplaces could generate a meaningful contribution to the utilities’ bottom lines.”
Customers are already spending on energy products and services, he said. If a significant portion of the NREL paper’s calculated technical potential is realized, “a large utility that acts as the procurement platform for its customer base, could see billions of dollars in transactions.”
As the marketplace shifts in response to customer demand, “utilities may want to start thinking about dramatic changes to their core business model and about these adjacent opportunities to remain relevant and generate revenues close to what they are making today,” Aggarwal said.
APS’s Piotrowski raised several questions about NREL’s assumptions. “It will be difficult to go from today's total installed cost of $2.66/watt to the projected total installed cost of $1.10/watt because people still need to make a profit,” he said.
Integrating solar installation into new construction and new roof builds could bring the marketed price point down, “but it could go away if there is a contractor involved,” he added.
Piotrowski also questioned the basic value of rooftop solar, even at $0.05/kWh. “It is great that solar prices are coming down, but in today's market, in the spring and fall shoulder months, we are curtailing plants so we can get paid to take California's over-generation of solar energy. Midday over-generation is the bigger issue we have to solve.”
Piotrowski leads the APS residential rooftop solar pilot projects approved by Arizona regulators in 2014. The 1,700 installations are giving the utility a better understanding of the impacts of large concentrations of rooftop solar on the distribution system, he said.
“If rooftop solar does go to $1.10/watt, the resulting higher penetrations would require us to strengthen our distribution grid because the solar output is in the middle of the day, when we don't need it,” he said. “Our rate design has moved our peak demand to the 3pm to 8pm period, and our solar-owning customers expect the utility to be their battery when there is no sun. That requires a reliable grid.”
Solar is “just one piece” in the operation of a distribution grid and it is driving APS to other technologies to meet peak demand needs, he said. “Storage may be the most economically feasible way to do that.”
The NREL findings tell Piotrowski a workable solution that includes storage and other load control technologies may be even more urgent for utilities like APS already facing midday over-generation and spiking peak demand, he said. “The exciting thing about the paper is that the solar-plus-storage package will be more cost-competitive in 2030 than it is now. That will allow us to harvest solar energy and have it when our customers need it.”
APS recently initiated a 75-home pilot incorporating solar, storage, smart thermostats, hot water heaters, home energy management systems and other customer-sited technologies. “Each home will receive a different package, Piotrowski said. “The small-scale study will allow us to identify what works to save energy and what works better.”
National Grid New Energy Solutions Program Manager Karsten Barden had a different perspective. About 300 Rhode Island utility customers used the pilot EnergySage online marketplace program to shop for solar and 75 purchased solar through the platform, he told Utility Dive. “That goes to customer satisfaction and that is the motivating factor.”
As the utility expands the platform to all its Rhode Island and New York customers, National Grid will use it to learn more about solar customers, he said. “Our ability to provide relevant and actionable information to customers will help us offer other types of products and services that allow them to evolve their control over their own energy choices and usage.”
The numbers in the NREL paper show why Barden and his National Grid team are “absolutely preparing for a future in which solar and storage are widespread,” he said. It could put more demands on the utility’s system but “we want to understand how all the customer-sited and utility-owned pieces of the distribution system can fit together.”
Offering access to new technologies and new rate designs that meet customer demand will allow National Grid to earn the performance incentives now being developed in New York’s landmark grid modernization initiative, Barden said. But those technologies and rate designs will also “allow customers to become a part of the solution to an integrated grid. That is the long game.”