Low natural gas prices, not high renewable energy penetrations, are the main cause of low and negative electricity prices that are negatively impacting coal and nuclear operators, a new report from two DOE national labs finds. 

"There is little relationship between the location of recent (2010‐2016) coal, nuclear and other thermal retirements and [renewable] penetration levels," Impacts of Variable Renewable Energy on Bulk Power System Assets, Pricing, and Costs from Lawrence Berkeley National Laboratory (LBNL) and Argonne National Laboratory says.

The findings have important implications for the current federal proposal to provide cost recovery for generators with 90 days of fuel supply onsite, such as coal and nuclear plants. Advocates for these resources argue they are needed for reliability, but disadvantaged by state and federal supports for variable renewable energies (VREs).

But the numbers — displayed in the second and third graphs below — show that the market share loss for coal and nuclear is mainly due to their inability to compete against natural gas.

The debate over the merits of DOE's cost recovery proposal raises questions about how nuclear, coal, natural gas and VREs fit together and what their relative values are to power markets. Those are questions to which the new research provides detailed, data-based answers.

Blame it on the gas

There is widespread agreement that electricity prices are trending lower and there are three common answers to why, LBNL research scientist and report co-author Andrew Mills told Utility Dive. One is low natural gas prices, another is flattening load, and the third is the growth of renewables. "But the relative shares of those factors have not been quantified before."

When the researchers did that quantification, they found the impacts of flattened load and high renewables penetrations to be much less significant than the impact of low natural gas prices.

The researchers first isolated the "wide array" of factors that impact wholesale electricity prices. The factors that go into the price of electricity begin with the cost of fuel and other operating expenses. They also include environmental regulations, emissions allowance prices and the cost of accessing available flexibility, when needed.

Total system load, peak load growth and daily and seasonal demand also influence price. Decisions about transmission investment and costs along with plant investments and retirements that affect supply can impact the market. And regional and local market rules, local reserves and local scarcity pricing also cause prices to fluctuate.

The researchers modeled changes in each factor individually to calculate its specific contribution to the wholesale price, Mills said. They used the California (CAISO) and Texas (ERCOT) systems, which simplified the study to single states.

Graphs show the impact of VRE in both systems from 2008 to 2016. The impact of VRE on the electricity price drop was similar to the impact of many of the other factors, Mills said. But the graphs show the impact of the decline in natural gas prices was a "dramatically large portion" of the decline in the wholesale electricity price.

About negative pricing

The nuances of electricity pricing are revealed in the question of negative prices, the report says. Negative pricing occurs when supply exceeds demand and when generators bidding into the market have an economic incentive to pay load to take their output.

There are two common reasons for excess supply. The first is over-procurement. The other is local transmission congestion that prevents supply from being widely distributed.

There are two common economic incentives for generators to continue operating in a negative price environment. One is that it may be more costly for inflexible coal and nuclear generators to dial down operations than to sell at negative prices. Another is that it may be more profitable for renewables supported by production tax credits or state mandates to sell at negative prices than to be curtailed.

Nationally, "negative prices have concentrated in areas with significant VRE and/or nuclear generation along with limited transmission," the report says. It is also most common "during periods with lower system‐wide load." Negative prices have increased recently in California and showed up in Texas in the early 2000s, when a policy-driven wind boom overwhelmed the state's transmission capacity. New capacity has mostly eliminated the issue in Texas. 

Brattle Group Principal Hannes Pfeifenberger told Utility Dive negative pricing gets attention because "it is so counter-intuitive that a price can be negative." It seems to suggest "something is wrong with the market,” he said.

But when coal and nuclear plant operators like FirstEnergy and Exelon "complain about negative prices, they are really complaining about low power prices," Pfeifenberger said. "They argue subsidized wind and solar are depressing the power price, but the report shows that low and negative prices are almost always due to low natural gas prices."

At specific locations, a high penetration of VRE can drive prices negative, he said. Examples are Illinois, where there is a lot of installed wind capacity and a big nuclear plant, or Western Kansas, which has a large coal plant but is load-constrained and transmission-constrained. But the report shows those are exceptions, Pfeifenberger said.

LBNL's Mills said there are times when the impact of VRE is "pretty dramatic" but "we don't see that happening across the country."

In addition, both Mills and Pfeifenberger stressed that negative pricing is almost entirely in real-time markets.

"Negative pricing has very little impact on a coal plant, a nuclear plant, a combined cycle natural gas plant or even wind," Pfeifenberger said. "Most of their revenue, about 95%, comes from the day-ahead market. With the exception of California, there's almost no negative pricing in the day-ahead market."

What does this mean for nuclear and coal?

The researchers did not find evidence of "a widespread impact of VRE on power plant retirements."

While VRE penetration levels are not linked to coal and nuclear plant retirements, the report did identify factors that correlate with plant shutdowns, LBNL's Mills said. There are "relatively strong" correlations of baseload capacity retirements with falling load and with high reserve margins indicative of excess generation capacity. There is also a strong correlation between coal plants with high SO2 emissions and retirements, the report says.

There is stronger evidence that some higher penetrations of VRE will eventually have quantifiably higher impacts on retirements, Mills said. Higher levels of VRE will "tend" to cause system operators to build new capacity but to give a higher priority to new and existing flexible and low‐capital-cost generation, the report forecasts. This will likely shift pricing to support the new priorities.

The overall result would be a significant reduction in the capacity factors of less flexible baseload coal generators, LBNL's Mills said.

Pfeifenberger said this could make a significant difference in the economic viability of even somewhat flexible generators. "A combined cycle natural gas unit might run 70% of the year without renewables but only 30% of the year with a lot of renewables," he said.

Cost versus value

While increasing VRE may not be the main drivers of coal and nuclear retirements, what are the implications of more renewables?

As VRE penetrations grow and the contributions and costs of energy mix components shift, a resource’s value to the system will become more important than its levelized cost of energy (LCOE), the report says.

“LCOE is an imperfect measure of the relative economics of generation resources,” it argues. Generation resources “have widely varying technical and economic characteristics” and “deliver different services” at different locations and times, it adds. Making procurement decisions is more complicated than simply picking the lowest LCOE.

Planners, investors, and forecasters “use complex computer models” to arrive at the least-cost portfolio of resources, the report says. Decisions are based not only on LCOE but also on “the specific values (or costs) that each resource brings to the grid — energy, capacity, balancing, and transmission."

The researchers used existing studies and original empirical data to conclude higher VRE penetration levels lead to “a declining system value for renewables,” Mills said.

System value includes energy, capacity, balancing, and transmission considerations,” the report says. Beyond avoidable direct cost, it can also include “other societal values” like resource diversity and environmental benefits.

The report details the decline in the system value of solar as its penetration has grown in California and the decline in wind’s system value in Texas before new transmission became available.

But the rebound of wind’s value in Texas when transmission congestion was alleviated shows “there are actions, what we call flexibility solutions, that can help mitigate the value decline,” Mills noted.