ORIGINAL REPORTING: Big solar is the cheapest solar but is it the best solar?
Utility-scale solar is significantly cheaper than rooftop. Does that make it better?; The cost numbers don't lie, but critics say they don't tell the whole story of rooftop solar's value
Herman K. Trabish, July 23, 2015 (Utility Dive)
Ro oftop solar costs significantly more than utility-scale solar, according to a new study that's stoking debate throughout the energy community, but critics say that price isn't everything.
The cost of utility-scale solar comes to between $0.066 per kWh and $0.117 per kWh, according to new calculations from The Brattle Group, while the cost for customer-owned rooftop solar is between $0.123 per kWh and $0.193 per kWh. And for customer-leased rooftop solar, the cost is reportedly between $0.14 per kWh and $0.237 per kWh.
“This doesn’t mean distributed solar doesn’t have a bright future. It is growing very rapidly, 74% year over year, is very popular with customers, and has many desirable features,” explained Brattle Group Principal Peter Fox-Penner.
“The point of the report is that utility-scale solar has other very desirable features and many of the same desirable features and it is cheaper,” Fox-Penner said.
“Most of the environmental and social benefits provided by PV systems can be achieved at a much lower total cost at utility-scale than at residential-scale,” is how the study puts it.
Brattle’s study, entitled "Comparative Generation Costs of UtilityScale and Residential-Scale PV in Xcel Energy Colorado’s Service Area" was commissioned by First Solar, the world’s biggest utility-scale solar developer, with support from the Edison Electric Institute, the trade group for American investor owned utilities.
Brattle Group researchers located two hypothetical 300 MW tranches of PV solar in the Xcel Energy Colorado system in 2019 to establish study specifics like insolation, Fox-Penner explained. Cost data was drawn from the National Renewable Energy Laboratory’s Open PV Database, which catalogues all projects nationwide, and from the Lawrence Berkley National Laboratory (LBNL) solar industry market reports.
One hypothetical tranche of PV was solar power plants that sold their output to Xcel Energy under standard long-term power purchase agreements. The other was 60,000 5 kW owned or leased residential rooftop systems.
A reference case was based on actual 2014 numbers. Five scenarios were run. In three, the investment tax credit (ITC) is 10%, 30%, or absorbed by the developer. In a fourth, the cost of PV drops. In the last, inflation rises:
A cost for the electricity from utility-scale solar and from both customer-owned and customer-leased systems was generated for each scenario.
“The generation cost difference between the utility- and residential-scale systems owned by the customer ranges from $0.067 per kWh to $0.092 per kWh solar across the scenarios,” the study reports. “To put this in perspective,national average retail all-in residential electric rates in 2014 were $0.125 per kWh.”
“The biggest two takeaways were surprising,” Fox-Penner said. “The generation cost per solar MWh purchased is half as large for utility-scale solar as for rooftop solar. And per MW of solar installed, the carbon savings and the fuel savings from a MW of utility-scale PV are 50% larger.”
Taken together, he added, that is a significant gap in value propositions.
The study attributes the gap to two major factors. One is a lower total cost per kW for installed panels at utility-scale arrays. The other is is that large scale arrays generate more power from their 300 MW capacity due to optimized panel placement, tracking, installation efficiencies and other economies of scale that come with large facilities.
Brattle also attributed a $195 million higher net present value (NPV) over 25 years to the rooftop PV.
“If the same amount of residential-scale PV systems (1,200 MW) were installed in 2019 as in 2014,” the study reports, “they would cost customers roughly $800 million more in NPV than a comparable purchase of utility-scale systems, under conditions assumed for the Reference Case.”
The study seems designed by EEI and First Solar to conclude that utility-scale solar is the best option, said SolarCity Communications Vice President Jonathan Bass. According to him, Brattle fails to consider all benefits to the grid of distributed solar and costs from central generation.
Bass points to another report prepared by the consultancy Crossborder Energy, a firm commonly tapped by the solar industry to prepare reports onrooftop PV issues. The study, entitled “Relative Benefits and Costs of Rooftop and Utility-scale Solar” offers a more comprehensive analysis, Bass said. While the report is not currently available online, its author covered its finer points in a presentation at the Southwest Clean Energy Transmission Summit.
Utility-scale solar offers higher capacity factors and significant economies of scale, Crossborder concludes. But rooftop offers location at the point of end-use, reliability benefits (especially when paired with storage), societal and customer choice benefits, and lower cost to customers than green pricing programs.
“To the extent that these added benefits of rooftop can be quantified, they essentially offset the cost advantage of utility-scale systems…[and] the higher costs of rate-based cost recovery erode some of the advantage of utility-scale’s lower costs, resulting in a modest advantage for rooftop solar,” Crossborder finds, though “both types of solar should have central roles.”
It is urgent to “de-carbonize” electricity generation, Bass agreed, so both utility-scale and distributed generation solar have important roles to play.
Strong evidence for community solar
The Brattle report is “strong evidence in favor of the community solar model, particularly because when it gets to the MW scale range community solar starts to look like utility-scale solar in cost,” Fox-Penner said.
First Solar’s recent ownership-scale investment in community solar developerClean Energy Collective and SolarCity’s recent commitment of $200 million to community solar projects in Minnesota suggest big things coming for the emerging sector. Community solar allows residential and non-residential utility customers without solar-suitable roofs to buy a small part of a large central array.
There is a point where economies of scale “break” in the favor of community solar developers and another — at around 5 MW, according to LBNL — where the advantages start diminishing, Fox-Penner explained.
“It is not that there is no advantage to building above 5 MW but that gains above that size diminish,” he said.
But the community solar equation should also factor in land costs, Fox-Penner said. Near population centers land can be expensive, and larger parcels to accommodate installations that offer economies of scale can come at a premium.
“Our results apply widely, but they should be calibrated to each system’s circumstances,” Fox-Penner said. Like solar-suitable roofs, he said, land availability and costs vary. Another factor affecting community solar economics is the cost and accessibility of the interconnection.
“Two 2.5 MW community solar projects might be cheaper than one 5 MW project because the bigger one might connect to one feeder and the smaller projects might connect to two feeders. Depending on the cost of the interconnections, the value proposition might favor the smaller projects.”
The upside of rooftop solar
There are some areas where rooftop solar offers more benefits, Fox-Penner said.
Rooftop solar offers the potential of greater resilience for the homeowner against outages, he said. It is only a potential advantage because most rooftop solar systems, as currently installed, automatically shut down when the grid goes out “to protect workers and to meet fire protection requirements.”
Distributed solar can also be configured into micro-grids or home systems that island.
“It requires different, protective connections with the grid that require some intelligence and switching capabilities," Fox-Penner said. "It can be done but it adds cost.”
Over time, islanding installs will become more affordable and more common, he said, as costs come down and solar installers’ capabilities ramp up.
“Resilience is not an advantage today, but it can be harvested from rooftop or community solar in the future.”
Brattle found the jobs created by rooftop solar to be “roughly equivalent” to those created by utility-scale solar.
“We did not verify that quantitatively but it is our qualitative opinion they are roughly equivalent,” Fox-Penner said.
The definitive annual statistical assessment of solar industry jobs comes fromThe Solar Foundation.
“The Brattle study’s jobs findings are counterintuitive and require more evidence,” Executive Director Andrea Luecke observed. Her organization’s upcoming National Solar Jobs Census 2015 will, for the first time, break out residential, commercial, and utility scale solar jobs.
That may not affect the Brattle analysis of jobs numbers.
“We expect more jobs to be created in the construction of rooftop solar systems because it costs more to do the installation,” Fox-Penner explained. That theoretically leaves electric customers who get the same number of megawatts from utility-scale solar, in aggregate, with disposable income.
“It is a universally accepted assumption that consumers with disposable income spend. It is called the re-spending effect and it creates jobs,” he said. “We don’t question that more solar industry jobs are created with rooftop solar. We are talking about total jobs.”
One other advantage rooftop solar might have, though it is difficult to quantify, is the engagement effect, he said.
“When people see solar, they think about solar and energy efficiency, and it has more impact when they see it in their neighborhoods than when they drive by a project on the interstate."
"It isn’t something I know how to quantify," Fox-Penner said. "I don’t think it’s big but I don’t think it’s zero.”
Net energy metering and other policy questions
Brattle’s conclusions give quantitative substance to many utilities’ assertion that large-scale central station solar provides more value to their customers than distributed solar. It is an argument they commonly deploy in calling for changes to retail rate reimbursement for net energy metering.
“The reimbursement rate should be changed from the retail rate to the rate the utility pays for wholesale solar generation from utility-scale projects connected to its distribution system,” argued Tucson Electric Power Energy Supply Senior Director Carmine Tilghman recently.
“I can buy solar energy from a larger scale project on my distribution system, paid for by my ratepayers, at $0.058 per kWh," he said, "half the price I am forced to credit people who push energy back on the grid from net metered systems.”
“Net metering is an important policy issue and there has been quite a lot of work done on it," Fox-Penner said, "but the purpose of this report is to look at the aggregate total costs of two solar options."
Brattle intentionally did not look at the policy question, he explained.
“We would expect the costs to be allocated differently, depending on how the tariffs are set up but how they should be set up is far afield from this.”
Social benefits are typically dealt with qualitatively rather than quantitatively, Brattle argues.
Because the study compares the costs of two energy sources that offer comparable social benefits, it did not include such considerations. Policy considerations may justify departing from a least-cost approach but “costs nonetheless are an appropriate starting point.”
However, “the magnitude of most non-monetized benefits achieved is generally proportionate to the higher solar output associated with scale," the study adds. "Displacing carbon emissions or water consumption is roughly 50% greater for 300 MW of PV capacity deployed as utility-scale than it is for 300 MW of PV capacity deployed as residential-scale.”
“We could purchase our own health care coverage or we could purchase it through a single network,” Fox-Penner suggested as an analogy. “In both cases we could get excellent health care coverage, but purchasing through a network, we can probably get it cheaper.”
NO QUICK NEWS