Berkeley Lab uses empirical PV+battery data to evaluate how business models influence dispatch

October 18, 2021
  • Berkeley Lab analysis of empirical PV+battery data from a sample of 11 plants operating in 2020 finds that dispatch and impacts to the grid depend on the owner’s business model.
  • In addition, Berkeley Lab updates the Solar-to-Grid report with new data for 2020, detailing trends in system impacts, reliability, and market value of stand-alone solar in the United States.
  • A free webinar summarizing key findings from the report and journal article will be held on October 26 at 10 AM Pacific/1 PM Eastern. Register for the webinar here:

Increasing solar generation is expected to progressively impact the bulk power system—changing prices for energy and other grid services. Solar generation is driven by sunshine and thus often highly correlated over the course of a day within a region. Without the deployment of storage or an increase in price-responsive load, growth in solar capacity drives wholesale prices lower during sunny hours, leading to a decline in market value.

Berkeley Lab’s Solar-to-Grid report, updated with data through 2020, confirms these expectations in California, where solar penetration is the highest in the country and its effects on net load, wholesale prices, and solar’s market value can readily be seen. The effects are small in most other organized markets where solar penetration is below 5%, though becoming evident in New England with 5.2% solar penetration.

Applications to interconnect with the bulk power grid show that commercial developers plan to couple almost all of their new solar projects with batteries in California, and over one-third of all projects across the U.S.

While adding energy storage to solar photovoltaics (PV) plants is expected to increase the value to the grid, there has been little empirical data to confirm that to be the case.

Building on last year’s Solar-to-Grid report, Berkeley Lab researchers collected empirical data from a sample of 11 PV-battery hybrid plants operating in 2020 to assess the impacts of adding storage. These plants all operate within organized wholesale power market regions, but rely on a variety of business models to justify the investments.

The research confirms that adding battery storage to PV plants increases the market value across all 11 plants in the sample. The increase in value relative to the value of standalone solar -- the “storage value premium” -- varies from $1 to 48/MWhsolar, depending on the size of the storage, the location, and the way that the storage is used.

To account for the differences in location and storage size, the researchers compare the empirical storage value premium to a calculated “baseline” value for a plant in the same location and storage size, but with simulated storage dispatch that maximizes the energy market revenue. The remaining difference between the empirical and baseline storage value premium is primarily driven by the underlying business model.

The research identified four different business models for PV+batteries, with widely varying value propositions for their owners, and for the grid. Only a few plants in our sample rely solely on selling power directly into wholesale electricity markets (“merchant”). The rest employ business models where the revenue depends less on energy market signals.

Instead, these plants dispatch the storage to reduce peak load to lower ISO/RTO-administered transmission or peak demand charges (“peak load reducer”), to comply with incentive program rules (“incentive participant”), or to offset customer bills or increase resilience (“large energy consumer”).

Merchants: Merchants fall short of our modeled baseline because they do not have perfect foresight of market prices and PV generation. Their storage value premium could be increased with experience and improved forecasting.

Peak reducers: Peak-load reducers in New England dispatch storage to avoid costly transmission and capacity demand charges that are assessed during monthly and annual demand peaks (bottom right panel in figure below). Although the operators may not follow wholesale energy price signals closely in other hours, they can still earn a multiple of what they would get via an optimized merchant dispatch (top right panel).

Incentive participants: Participating in incentive programs requires operators to comply with program requirements, for example administratively set seasonal discharge windows in Massachusetts’s SMART incentive program, which may have a dynamic peak-load component like Massachusetts’s Clean Peak Standard (bottom left panel). Following these dispatch signals can yield much higher private revenue than what would be available via merchant dispatch (top left panel). Incentives can vary widely by state.

Large energy consumers: Industrial retail electricity tariffs can come with hefty non-coincident peak-demand charges ($25,000/MW for the consumer in the researcher’s sample). Reducing apparent load during these peak demand hours can be much more profitable than engaging in energy arbitrage, even if the battery dispatch does not coincide with high electricity prices.

Additional value can come from using the battery to provide ancillary services. Still, in a number of cases, the plants employing these business models underperform the baseline that maximizes energy revenue. Absent larger reforms to tariff and incentive program structures, deviations in optimal dispatch decisions between highest system and private value will likely persist, driving a wedge between the potential market value of PV-battery hybrids and what they contribute in practice.

The research shows that private profit-maximizing dispatch signals are often stronger than the price signals conveyed by wholesale energy markets, resulting in much higher realized private revenue than that offered by wholesale markets. Understanding those private dispatch signals will be key for grid system operators seeking to understand market behavior, especially as the hybrid projects will represent a more sizeable share of the overall generator portfolio.

A free webinar summarizing key findings from the research, along with highlights from the 2020 update to the Solar-to-Grid report, will be held on October 26 at 10 AM Pacific/1 PM Eastern. Register for the webinar here:

For questions, please contact Andrew Mills ( or Joachim Seel ( at Lawrence Berkeley National Laboratory.

We appreciate the funding support of the U.S. Department of Energy Solar Energy Technologies Office in making this work possible.


Signup for our Mailing List