Study documents impacts of large-scale entry of rooftop solar panels on competition

Fossil-fuel plants are increasingly being forced to stop and start production in response to changes in output from renewables. In a new study, researchers developed a dynamic competitive benchmark that accounts for startup costs and other unit-level operating constraints. They apply their framework to Western Australia, a setting where rooftop solar capacity more than doubled between 2014 to 2018 to world-leading rooftop solar penetration rates.

The study found that the large-scale expansion of rooftop solar capacity can lead to increases in the collective profitability of fossil fuel plants because competition softens at sunset—plants displaced by solar during the day must incur startup costs to compete in the evening.

The study, by researchers at Carnegie Mellon University and Monash University, is published in the American Economic Review.

“We developed a framework to measure market power in wholesale electricity markets,” explains Akshaya Jha, associate professor of economics and public policy at Carnegie Mellon’s Heinz College, who co-authored the study. “This framework accounts for features of generating unit technology such as fixed startup costs and ramping constraints that are becoming increasingly relevant in light of the global transition to intermittent wind and solar technologies.”

Firms that incur the fixed costs required to start production expect to recover these costs by earning revenues in excess of their variable costs in subsequent periods. This presents two challenges for studying competition: First, market power is usually measured based on the markup in prices above short-run marginal cost, but economists have long recognized that setting prices equal to short-run marginal cost ignores the requirement that prices must be sufficient for firms to recover their fixed costs. Second, fixed costs are a barrier to competition: Decisions on the extensive margin to incur the fixed costs necessary to produce affect the intensity of later competition.

Researchers applied their framework to Western Australia, a world leader in rooftop solar penetration rates. They measured market power by comparing observed plant output and market prices to their benchmark—a counterfactual time series of plant output and market prices—that accounts for the recovery of the fixed costs required for plants to start up.

Specifically, they extended static production function approaches by using high-frequency data on input gas use and output electricity to estimate unit-level cost functions with three components: variable costs, startup costs, and the costs associated with running not tied to output levels. Then, their dynamic benchmark sets output levels to minimize the daily total costs of dispatching power plants to satisfy demand in each half hour of the day while setting prices that allow each plant to recover their fixed and variable costs.

Using this framework, they found that increases in rooftop solar penetration corresponded to sizable increases in the market power rents earned collectively by the fossil fuel fleet after the sun set. Since retail prices paid by electricity users were set via cost-of-service regulation, increases in market power rents largely constituted transfers from retail electricity consumers to producers.

Although rooftop solar penetration had a small effect on efficiency in the wholesale market, the external welfare gains associated with reductions in greenhouse gases are not captured within the wholesale electricity market. In Western Australia, increases in rooftop solar penetration corresponded to substantial declines in the carbon emissions that contribute to climate change, with sizable drops in gas-fired electricity output and thus daytime carbon emissions and only small increases in carbon emissions in the evening associated with solar-induced increases in starts by gas units.

“Our findings speak to the growing relevance of adopting several design features not present in most markets outside the United States,” says Gordon Leslie, senior lecturer in economics at Monash University, who coauthored the study.

Features include allowing suppliers to submit startup bids in addition to energy supply curves can allow for co-optimization across hours. This can improve market efficiency, especially as more units stop and start production in response to output from intermittent wind and solar resources.

Previous research has documented that the benefits of allowing financial participation in the day-ahead market are especially large in contexts where physical operating constraints are more likely to bind and market participants have significant market power. This study showed that increasing rooftop solar penetration rates can exacerbate unit-level start costs and physical operating constraints, resulting in suppliers having greater ability to exercise market power in the evening.

Worldwide, most electricity users face retail prices that do not vary contemporaneously with wholesale prices and often not even by hour of day. Allowing retail prices to reflect hourly variation in wholesale prices will likely shift some demand from evening to day. As a result, fewer fossil fuel units will need to start up in the late afternoon to compete effectively at sunset. This can reduce firms’ ability to exercise market power during evening peak demand hours, leading to lower wholesale prices and declines in the retail prices paid by consumers.