The Trump administration absolutely united around one goal: supporting the US coal industry.
Back in April, Energy Secretary Rick Perry commissioned a study on the resilience and reliability of the US power grid, with a special focus on whether “fuel-ready” power plants like coal and nuclear were being unfairly driven out of business in wholesale electricity markets.
In July, a nearly complete version of the study leaked, and to everyone’s surprise, it was ... good. The analysis was professionally done, and the conclusions (which were not at all friendly to the administration’s position) mostly mirrored those of similar studies by outside groups. To wit: The loss of these big plants is not hurting reliability, and keeping them open wouldn’t help it.
When the final version was released, in August, there was some political spin at the top and bottom, and some huffing and puffing about the wonders of baseload. But the analysis survived mostly unmolested.
Then earlier this month, Perry released his plan to save coal and nuclear plants, and it was, in a word, bonkers (read this post for more). It recommended to the Federal Energy Regulatory Commission (FERC) that it intervene in markets in the most crude and ham-handed way, dumping windfall profits on the owners of big, uncompetitive power plants.
It also raised a big question: What would Perry have recommended if he had taken his own department’s study seriously, instead of following Trump’s orders to save coal at all costs?
We don’t have to guess. Though the details are incredibly devilish, there is consensus on at least the broad strokes of a plan to ensure grid resilience and reliability, much of which was represented in the DOE study itself.
In this post, I’m going sketch out what a smart reliability plan would look like, in terms that even a grid layperson can understand. Electricity policy is notorious for being opaque and riddled with acronyms, but the basic elements are not actually that complicated.
For help, I’m going to draw on Alison Silverstein, an energy analyst who headed up the technical work behind DOE’s study and was one of its lead authors. In an editorial and in a recent presentation to the Committee on Regional Electric Power Cooperation (CREPC), she outlines what she would have recommended based on the findings of the study.
Wholesale energy markets are working, and reliability is fine
It is true that lots of coal and nuclear plants (especially coal) have retired since 2000, and it’s true that they retired most and fastest in areas with wholesale energy markets. But that’s because markets are working.
Market competition did what it was supposed to do, which is drive older, less efficient plants out of business. Almost all the coal plants that retired were old — most over 35 years, some as old as 70 — and, crucially, were no longer operating as baseload plants.
As Silverstein says, “baseload” describes a mode of operation, not a particular fuel source. It refers to plants that are always (or at least most of the time) running, satisfying the base level of around-the-clock demand.
These older plants were no longer serving that function. They were too clunky and expensive, mostly kept around for peaks in demand. And with more variable renewables on the grid, there is more need for flexibility, rapid ramping up and down, and coal plants aren’t good at that either. With demand no longer rising, natural gas prices stubbornly low, and pollution standards biting, they started dropping off the grid.
That’s what was supposed to happen. And reliability is fine. As the DOE study found, reliability has been improving lately, “due to better planning, market discipline, and better operating rules and standards.”
Does that mean everything’s hunky-dory? No. There’s evidence that wholesale markets are not properly valuing some services and may be vulnerable in the long term if they keep limping along in their current form. So here are some recommendations from Silverstein, our alternate universe non-hack Perry.
1) Protect and improve wholesale markets
Market competition is good. But markets must be well-designed to produce socially beneficial outcomes.
Wholesale markets, like all US electricity regulation, were designed around two attributes: reliability and low cost. They are every utility’s primary obligation, and what they are regulated to produce.
Problem is, over the years, we’ve begun to want other things out of our electricity — most notably low carbon, but also jobs and economic development, resilience, perceived social benefits, and sometimes just politically preferred sources.
So wholesale markets end up operating in the context of all sorts of extra-market subsidies, mandates, and regulations. The more of those there are, the less room there is for markets to work. And Perry wants to add a doozy.
The answer is not to abandon markets, or to abandon the pursuit of those other attributes (especially low carbon!). The answer is to figure out how to integrate the two — how to value those attributes within markets, so that they benefit from market competition.
That’s the core issue.
The way Silverstein puts it is, “What would market design look like if we started fresh, but informed by 20 years of experience?”
2) Figure out what the grid needs and put a value on it
Right now wholesale markets value — i.e., pay for — electricity. In some cases, there are capacity markets alongside electricity markets, which pay plants to stay open even if they aren’t running, just in case. And in some cases, there are “ancillary services” markets alongside those, which pay for things like frequency regulation and voltage control.
Insofar as resilience and reliability are undervalued attributes in those markets, the key is to figure out a way to better value the attributes, not to subsidize specific technologies. In other words, figure out what resilience and reliability services the grid needs, and then pay for the services. Let the market figure out the best way to provide the services.
Silverstein summarizes thusly: “Identify, define, productize, and compensate essential reliability and resilience services to meet multi-hazard threats and scenarios.”
We’re not very good at that yet. There’s a lot of discussion and experimentation to be done to get markets properly valuing those services. But the goal, the North Star, is to design a market in which the services we want are valued in a way that guarantees their provision.
That’s not what Perry did. He picked one particular attribute — “fuel assurance,” which means having big piles of coal around — and proposed to exempt plants with that attribute from competition entirely. That’s dumb on both ends. First, it defines resilience and reliability far too narrowly, in terms of just one (only marginally relevant) attribute. And second, it removes the service from markets rather than using markets to better provide the service.
Here’s one quick example of a reform grid operators could make. Grids need “frequency response,” which is a fancy way of saying they need sources of electricity that can ramp up and down quickly in response to minute-by-minute fluctuations of demand and supply on the grid. Traditionally that role has been played by “spinning reserves,” plants that are already running (have inertia) and can ramp quickly.
Recently, there have been successful cases of distributed renewables and energy storage providing frequency response, through “inverter-based synthetic inertia.” So DOE should investigate other sources of frequency response, and markets for frequency response should be made, insofar as possible, technology-neutral.
Define the services so that markets can provide them. That’s the key.
3) Resilience and reliability come from diversity
Perry’s core mistake is to ascribe the attributes of resilience and reliability to a particular kind of power plant. That’s a category error. What matters to customers is the resilience and reliability of the grid, which are emergent properties of the interactions of a whole range supply- and demand-side resources.
“How many big power plants with big piles of fuel do you have?” is a caveman approach to resilience. As many critics have already noted, big piles of fuel would have done nothing to help in the wake of the polar vortex, Hurricane Harvey, or the other recent US disasters. There are many kinds of disruptions and disasters that can be planned for, but many can’t (and this will become even more true as climate change progresses). The key to resilience is a system that holds up under a wide variety of possible conditions.
To build that kind of system, grid operators will need to draw together a diverse array of resources on both the supply and demand side. In his order, Perry purported to be addressing resilience and reliability but completely passed over demand-side tools — the increasing ability to shrink or shift consumer demand, in real time — which are some of the most promising.
“Market operators are trying to incorporate societal needs and state preferences into reliability and market rules,” Silverstein says, “but they need help (as do state and federal regulators) to design cost-effective, risk-moderating portfolios of supply- and demand-side resources that will deliver lasting value under diverse, uncertain future paths.”
4) Subsidies should be purposeful and temporary
Here is where Silverstein differs from Perry. She notes that every source of energy receives subsidies of some kind:
Oil and gas get depletion allowances, renewables get production tax credits, investment tax credits and R&D, nuclear generation gets insurance, R&D and construction work in progress, natural gas gets depletion allowances and R&D, and so on.
(I will just note here that by even conservative estimates, fossil fuels are far more heavily subsidized that renewables.)
But insofar as there are a bunch of legacy subsidies and regulations in need of reform, the answer is not to do what Perry did, which is stomp in with a big new subsidy for the least competitive plants. “New subsidies for coal and nuclear plants won’t level the playing field relative to renewables nor undo the impact of old subsidies,” Silverstein writes, “they’ll just make the playing field even bumpier.”
The renewable energy tax credits are set to phase out. Congress should phase out fossil fuel subsidies as well. If we want resilience, we should define resiliency services in a tech-neutral way and let market actors compete to provide them. If we want carbon-free power, we should define it in a tech-neutral way and let market actors compete to provide it. And so on.
Silverstein says: “Any new subsidies — including direct state payments, out-of-market uplift payments, or potential cost-of-service payments for non-competitive resources — should have a specific purpose and a limited duration.”
That ... does not describe Perry’s proposal.
The discussion is just beginning, and Perry is not helping
Electricity system nerds (they exist) will read what’s above with wry amusement. It’s easy enough to describe needed reforms in lofty, abstract terms: “Be more market-y!” But there are many contentious fights and boring regulatory hearings to come, zillions of details and devils in every one. Reform in the electricity sector is difficult even in the best of circumstances, and as you may have noticed, we’re on the darkest timeline.
We don’t yet know exactly how to value resiliency services in a competitive market. But we know enough about the basic shape of needed reforms to know that what Perry proposes is a step backward, a grievous and distorting blow to markets that would attempt to undo their greatest success.
The problem Perry and Trump face is that no sensible, market-friendly reform will help coal. The more coal is exposed to competitive markets, the more it loses. Renewables are storming in, and the more they do, the less the system needs (or can accommodate) “baseload” power services.
As the chart in Michael Liebriech’s post illustrates, the increase in variable renewables will occasion an increase in the need for flexible resources — supply- and demand-side resources that can ramp up and down quickly, balancing out variability.
Coal plants are not good at that. They are big and slow, expensive to ramp up and down. (Nuclear plants are a special case. There’s some evidence that current plants can do “load following” — ramping up and down — and there’s hope for smaller, faster reactors in the future.)
As for resilience, it will be better served by good information, planning, and management of diverse, distributed resources than by having big piles of coal around.
Electricity markets need to get smarter about valuing important grid services, including resilience (and carbon reduction!). But the smarter they get, they less they will need coal, because coal is dumb.