Before Ontario Power Generation finalized its decision last month to construct North America’s first small modular reactors, the state-owned utility considered alternatives.
Building the first of GE Hitachi’s 300-megawatt BWRX-300 reactors could have cost the Canadian province more upfront — particularly for the first unit — than building nearly nine gigawatts of wind and solar generators backed up by batteries. But after weighing the volatility of the intermittent wind and sunlight, all the land it would require, and how much new transmission infrastructure would need to be built, the Ontario’s Independent Electricity System Operator ultimately determined the SMRs were the lower-risk option.
It’s a prime example of the limits to the metric that has long dominated debates over which generating sources to build: levelized cost of energy.
LCOE, as it’s known, is a financial measurement calculated by dividing the total cost of a project by the energy produced throughout its lifespan. It’s a crucial benchmark for deciding whether to invest in solar panels, gas turbines, or nuclear reactors. But, according to the Clean Air Task Force, it doesn’t tell the whole story.
On Tuesday, the Boston-based climate group published a 32-page report outlining the risks of relying too heavily on LCOE to determine what kind of energy sources to build.
The findings break down into four main points:
- Electricity bills are shaped by a combination of generation, storage, transmission, and distribution costs — and LCOE only captures a sliver of that total.
- Technologies with higher LCOEs, such as nuclear reactors or geothermal plants, can reduce total system costs because they require fewer new power lines than equivalent-sized solar or wind facilities.
- LCOE excludes qualities such as reliability and dispatchability.
- And, finally, better tools for regional planning already exist.
Just look at Ontario’s decision to invest in SMRs, said Kasparas Spokas, CATF’s electricity program director.
“Ultimately, the decision was made off a system-cost assessment,” Spokas told Latitude Media. “The SMR itself per megawatt-hour is a higher-cost resource than the alternative portfolio they were looking at, which had solar and wind. But when taking into consideration some of the other important factors… it led the government to go with a higher-levelized-cost resource.”
The report is hardly the first to raise concerns about what Spokas called the “misuse” of LCOE. Yet “some policymakers, advocates, and media — the ones who are less technical — have tended to use it as a fallback metric,” he said.
As an article Forbes magazine published in February illustrates, LCOE is typically misapplied to boost intermittent renewables such as wind and solar over firm resources such as nuclear and geothermal without considering the other costs associated with weather-dependent generators.
“It’s not that it shouldn’t be used,” Spokas said. “But grid operators and planners on the engineering side — they don’t use LCOE to make decisions.”
Alternatives to LCOE
Spokas compared LCOE to the upfront cost of a car. “There’s the sticker price versus the total cost of ownership of a car,” he said. “There’s the sticker price, then there’s maintenance, fuel, parking, and tolls.”
As an alternative, the International Energy Agency developed a metric called value-adjusted LCOE, which captures the shifting value of a generating asset during peak demand.
Another metric is levelized avoided cost of electricity, or LACE, “reflects the costs that would be incurred to provide the electricity displaced by a new generation project as an estimate of the revenue available to the plant,” according to the Energy Information Administration. This essentially represents the value of a generator to the grid.
“The relative difference between LCOE and LACE is a better indicator of economic competitiveness than either metric alone,” Manussawee Sukunta, an economist at EIA, wrote in a 2018 report. “A comparison of only LCOE across technology types fails to capture the differences in value provided by different types of generators to the grid.”
Then there’s the levelized full system cost of energy, which compares the cost to serving an entire market using just one generating source plus storage, according to a 2022 paper by economist Robert Idel.
A final option suggests modifying LCOE to add the cost of firming up intermittent resources. The consultancy Lazard, whose LCOE numbers are widely used in public discussions, last summer reworked its calculations to include firming costs into its estimates.
The broader recognition of LCOE’s limits show “policymakers need to start asking different questions,” said Malwina Qvist, the director of CATF’s nuclear energy program.
“Instead of asking what’s cheapest per megawatt-hour, they should ask what mix of resources will deliver reliable power at the lowest total system cost while transitioning to clean generation,” she said in a statement. “That’s the shift this report is designed to support.”
