The new year could bring a new benchmark for pairing energy storage with wind and solar projects.
In a report on its solicitation process released late in December, Xcel Energy’s Colorado utility subsidiary said it received more than 400 proposals, including what may be record-low prices for renewable energy paired with energy storage.
The median bid price for wind-plus-storage was $21/MWh and for solar-plus-storage was $36/MWh, beating the $45/MWh price seen last year in Arizona in a power purchase agreement between Tucson Electric Power (TEP) and NextEra Energy, GTM Research’s Shayle Kann noted on Twitter.
But the winning bids likely won't be selected for at least another six months and the lack of details on the Xcel proposals makes any comparisons difficult.
The median price for a standalone wind project in the Xcel report is $18.10/MWh. For a standalone solar project the median price is $29.50/MWh.
Solar, wind prices at inflection
“The prices for bare wind and solar power in the solicitation are so low, they are really impressive,” James Lazar, senior advisor at the Regulatory Assistance Project, told Utility Dive.
“Colorado has always been a good wind state, but to see standalone storage, and all renewables-plus-storage bids come at such low prices, this sets a new benchmark for other markets on the mainland."
Director of Energy Storage, GTM Research
Prices for solar and wind power have reached “an inflection point,” Lazar said. They are now cheaper than a gas plant based on variable fuel and O&M costs, he said. “It is a tectonic shift.”
The Colorado prices are among the lowest reported to date, Ravi Manghani, director of energy storage at GTM Research, said in a report. It is possible the Xcel prices, especially the solar-plus-storage prices, will set a new benchmark for markets outside of Colorado, Manghani said in an email exchange with Utility Dive.
“Colorado has always been a good wind state, but to see standalone storage, and all renewables-plus-storage bids come at such low prices, this sets a new benchmark for other markets on the mainland,” he said.
“At a minimum, the Colorado results could provide other markets a futuristic view of what resource planning could look like, by including non-traditional, but cleaner and somewhat dispatchable resources,” Manghani said. He noted that the economics of a given project could change in states that do not have as robust a wind regime as Colorado, but noted that “capital costs are coming down rather uniformly, irrespective of the location.”
Cost comparisons difficult
The Colorado data are likely to invite comparisons between the cost of either renewables or renewables-plus-storage with conventional technologies. But that is difficult to do because the renewable resources and conventional resources, except for standalone storage, are reported in different metrics. The conventional resources are reported in $/kW-mo, which better captures the value of capacity, while the renewable resources are reported in $/MWh, which captures the value of energy as dispatched.
The reality, said Manghani, is the solicitation covers demand growth and the voluntary retirement of coal plants, which means Xcel will likely use a combination of dispatchable and energy-only resources to come up with a least cost and reliable portfolio.
The solicitation, in fact, encompasses four separate requests for proposals: for dispatchable PPAs, renewables PPAs, semi-dispatchable (renewables-plus-storage) PPAs, and for company owned contracts.
When Xcel selects the winning bidders, it will screen the bids for costs and run the selected portfolio of bids through a system planning model to assess overall system reliability.
It’s premature to predict which combination of bids will pass both the screening and the modeling tests, Manghani said. He also noted that Xcel has received bids representing 112 GW of resources, while the upper limit of the resources it needs is only 1.1 GW.
The solicitation is the result of a two-track process, Xcel spokesman Mark Stutz told Utility Dive. It began as Xcel Colorado’s Electric Resource Plan, filed in May 2016, that identified a need for 450 MW. In August, Xcel filed its Colorado Energy Plan that called for up to 1,000 MW of wind power, 700 MW of solar power and 700 MW of gas-fired generation.
In the current solicitation, Xcel invited all resources to respond. In about two months, the Colorado Public Utilities Commission will hold hearings on the Colorado Energy Plan and is expected to rule on it by the end of April. A final decision on the winning projects is not expected until the end of July. The upshot is that the report that has attracted so much attention is still very preliminary.
Until the regulatory process advances, it could be difficult to tease out more details from the median results in the report, which was released to comply with regulatory requirements.
For instance, one cannot reliably evaluate the value of the energy storage bids because there is no information on duration, Lazar said. “You don’t know if it is two drops or two gallons of storage.” That lack of detail also makes it hard to compare the median prices with the TEP-NextEra PPA.
Without knowing the capacity of the bids and MW and MWh metrics of the renewable-plus-storage projects, it is hard to compare them with the TEP-NextEra deal, Manghani said. The TEP-NextEra project calls for 100 MW of solar power and 30 MW, 120 MWh of storage.
Timeline drives price
“You’d expect the price will be higher for more MW storage per MW of solar, as well as more MWh/MW, but setting those unknowns aside, the biggest driver for the lower median price is the project timeline for this solicitation,” Manghani said.
The solicitation requires the projects to come online by May 2023, about four years after the TEP-NextEra project is due online.
“While I’m sure the industry is not simply straightlining the cost declines for the additional four years, there are enough reasons for the bidders to be confident that some of the cost drops on battery and non-battery components will continue into the future.”
If that is the case, it makes sense that the TEP-NextEra benchmark lasted only a year.