The Biden administration’s goal to decarbonize the electricity sector by 2035 requires policy coordination and careful planning to align the retirement of around 250 GW of coal-fired power plants with new generation, much of it inevitably natural gas infrastructure.

“Certain locations, especially those that are going to see rapid coal retirements, may need to add new natural gas capacity in the short run to manage reliability needs,” Jesse Jenkins, assistant professor at Princeton University, said at a webinar on Tuesday. “But that means other regions across the country may actually be able to retire older natural gas plants and not replace them, so on net, across the country we’re kind of flat.”

Columbia University’s Center on Global Energy Policy (CGEP) hosted the webinar, where Jenkins outlined Princeton University’s Net-Zero America Project mapping pathways for the U.S. to reach net-zero greenhouse gas emissions by 2050.

Jenkins co-authored an interim study in December estimating that achieving net-zero emissions economy-wide will require at least $2.5 trillion in additional capital investment into energy supply, industry, buildings and vehicles over the coming decade. “Don’t expect any major changes” in the final report to be issued next month, he said. (See Net Zero Price Tag: $2.5 Trillion.)

Emily Grubert, assistant professor at Georgia Tech, brought the perspective from her December article in Science magazine, “Fossil electricity retirement deadlines for a just transition.”

“It may not be necessary right now to commit entirely to one of the many available pathways, but trying to shut down the coal fleet as fast as we can makes sense,” she said.

Average Lifespans

CGEP founder Jason Bordoff noted that the timeline Grubert presented for retirements of fossil fuel-fired power plants shows 80-year lifetimes for nuclear plants, which would not have been the case several years ago.

“So how do lifetimes play out in real life?” Bordoff asked. “Do they often get extended? What keeps them online past their lifetime versus early retirement? Does that have implications for how we think about the sorts of policies we’ll need in addition to some of the incentives to use different types of energy?”

Grubert said there are some important resilience implications for understanding actual plant lifetimes. One thing that underpins her recent research, she said, is to assume only the average age on retirement that has been observed for power generators with the same tools and the same technology, in her article’s case between 2002 and 2018.

“This often goes beyond what we call book life, and a lot of these plants do last a long time,” Grubert said. “Interesting for ratepayer policy is that, because it’s an average, when we try to set retirement deadlines as policy, it’s often easy to miss how big the standard deviations are on the historical shutdown basis.”

Not only are a lot of plants well beyond their typical lifespan, but there are big units that could theoretically close down well in advance of what planners expect, partially because they are physical infrastructure, she said.

“They have parts that are moving, and stuff happens, and sometimes you can’t get the part that you need and it doesn’t make sense to try to keep the plant online,” Grubert said. “It’s a combination of what do we want to keep online and this background risk of, if we are assuming typical lifespan, we actually have to acknowledge that there may be some that just break.”

Planners must ensure that the new infrastructure is matching up temporally with the old infrastructure closing, she said.

“We can decide when we want these plants to close, but that’s not necessarily always going to align with when it comes time to close,” Grubert said. “Some of this stuff has been driven by policy, but occasionally we’ve seen nukes in this situation go offline forever when we weren’t planning on it.”

Payback Planning

Without proper planning and policy signals, it’s likely that the large-scale retirement of about 250 GW of coal capacity could lead to an equal scale deployment of new natural gas in the short run because there aren’t other viable clean technologies ready for widespread use now, Jenkins said.

“We do need some policy coordination and signals, both to pull forward the deployment of clean, firm capacity and drive its innovation, and also to say, ‘we’ve taken it as far as we can go with the existing gas fleet,’ which is already built and partially amortized,” he said.

If anyone is going to build any new gas capacity, it should be under clear conditions, he added.

He said those conditions include:

• helping retire a coal-fired power plant and reduce emissions;
• amortizing it over an appropriately short timeframe, so if a power plant comes online in 2030, it isn’t rate-based for 40 years; and
• it’s needed for reliability, where necessary capacity can’t come from a combination of demand flexibility and storage.

Planners need to think deeply about how to provide energy services at a lower energy intensity, and there might be good arguments for deep building retrofits, Grubert said.

“We are going to reach the point where we have poor ore grade, essentially, for a lot of these projects, either in the sense that the resource isn’t that good, or people don’t want it there,” she said. “How do you avoid those marginal and more difficult projects later? Some of that is increasing resilience of people to be in their houses, decreasing their costs for all these other benefits that we tend to get as ancillary with some of these building retrofits.”