Who needs 'baseload' power? (Or, let the markets do their job)
In April, U.S. Secretary of Energy Rick Perry announced a 60-day study on electricity market design and grid reliability, meant to assess to what extent current market designs fail to adequately compensate "baseload" (coal- and nuclear-fired) power plants.
The memo (PDF) commissioning the study presents as "fact" a curious claim: "baseload power is necessary to a well-functioning electric grid." This notion has been thoroughly disproven by a diverse community of utilities, system operators, economists and other experts that moved on from this topic years ago. To these practitioners, this premise seems as backward as if President Dwight Eisenhower, instead of launching the interstate highway system, had called for restudy of the virtues of horse-drawn carriages.
Today, the grid needs flexibility from diverse resources, not baseload power plants. Leveraging market forces to help us decide between options offers the best chance of avoiding the multitrillion-dollar mistake — and gigatons of carbon emissions — of blindly reinvesting in the past century’s technologies.
Modern grids don’t need baseload
Utilities in the United States have had at least a decade of comfortable experience operating grids with a declining share of baseload power relative to low-cost renewable energy. Meanwhile, across the Atlantic, both reliability and renewable energy adoption levels are higher than in the United States; notably, the lights failed to go out in England when the U.K. grid recently ran for a full day without any coal power for the first time since 1882, foreshadowing its planned phaseout by 2025.
Analytically, scientists working for the Department of Energy’s (DOE’s) own world-renowned national laboratories, among others, consistently have shown that grids with moderate-to-high (30–80 percent) shares of renewable energy, and commensurately lower shares of baseload capacity, work just as reliably and at least as resiliently as fossil fuel-based power systems, but with lower operating costs and risks.
Utility executives, too, increasingly see the writing on the wall that not only is baseload unnecessary for a reliable grid, but it is financially incompatible with a rapidly changing energy landscape. The CEO of National Grid said in 2015, "The idea of baseload power is already outdated," as consumers look to cheaper resources, closer to them, to meet their needs. An executive from PG&E, one of the nation’s largest utilities, said, "The idea of a large baseload generator that runs pretty much all the time … just doesn’t have as good a fit to the market conditions we expect to see" in the grid of the future.
Even in Tennessee, where the legislature recently passed a bill to freeze wind development in the state, the CFO of the Tennessee Valley Authority admitted that the changing grid landscape "makes all of us very concerned about making any long bets" on big, new, expensive power plants.
It is curious, then, that the stated premise of the DOE study neglects to take into account the empirical evidence, analytical rigor and financial reality that, together, have been dissolving the baseload argument for at least a decade. While the physics of electric power delivery hasn’t changed meaningfully since the days of Westinghouse and Edison, the comparative economics of options to meet these physical constraints have turned upside down.
Reliability is a system attribute, not a unit attribute
Nobody is denying that our economy needs a reliable grid, where supply of electricity must match demand at all hours, but that’s no reason to think that we need individual power plants that run all the time. That’s a good thing, because such plants simply do not exist.
Baseload plants have forced outages between 2 percent (nuclear) and 10 percent (coal) of the time, and are expensive to back up when they fail unexpectedly, as they do inconveniently during heat waves and cold snaps (PDF), when they are needed the most. This is true even if they have "fuel on hand," a puzzling term for a characteristic that applies much more readily to solar and wind power than coal plants, in that those renewables don’t depend on hauling fuel from afar — because they use no fuel.
Thus, for the entire history of the grid, a portfolio of diverse resources always has been the cheapest way to guarantee reliability at least cost. In the past, this diverse portfolio has combined baseload plants with cheaper-to-build, more agile plants (natural gas-fired turbines). This combination let the power grid adapt to the variability of load, which is often less predictable than wind and solar energy production.
What’s different today, compared to even five years ago, is that the default choices for low-cost energy are, in fact, wind and solar, with long-term fixed prices already outcompeting the costs of new nuclear, coal and gas plants — or even just the operating costs of old ones. Given those renewables’ incredibly low and still-falling costs, flexibility is even more important than it has been — and luckily, flexibility too is cheaper, cleaner and more plentiful than ever before.
Flexibility is the new coin of the realm
Flexibility to integrate clean energy resources reliably into grid operations comes in many forms, and at all timescales. There are three primary use cases to consider, corresponding to various operational needs of a reliable grid:
1. Seconds to minutes: At these short timescales, the grid needs "ancillary services" to maintain reliability and ensure that demand and supply always match. Traditionally, these services have been provided by fossil or hydroelectric generators with spare capacity, but increasingly, it’s becoming cost-effective to procure them from distributed energy resources and other new, clean energy alternatives.
For example, a utility-scale solar plant in California has shown (PDF) that many of these ancillary services can be provided as a valuable byproduct of the plant’s inverters — extremely reliable solid-state power-electronics devices already paid for by their primary task of conditioning renewable power to feed into the grid. Aggregators in the PJM grid are controlling fleets of water heaters to balance second-to-second mismatches between demand and supply. And importantly, these and other clean energy resources often provide better ancillary services than the fossil plants they are offsetting (faster and more precise response to signals), so they get paid more under the Federal Energy Regulatory Commission’s "pay for performance" rules. Other resources, including electric vehicles and dedicated batteries, are increasingly participating in these same markets, providing many valuable services to the grid using hardware already paid for by their primary use case (driving for vehicles or backup power for batteries).
2. Hours to days: At these medium timescales, the grid needs resources that can balance energy supply and demand across the hours of a day as both demands for end-use services and solar and wind energy production ebb and flow. (Other kinds of renewable power, from geothermal and small hydro to burning wastes, don’t vary much.) Clean energy resources already excel at providing these services; indeed, they are the bread and butter of the multibillion-dollar U.S. market for demand response.
For example, about 60 GW (PDF) of loads are enrolled in demand response programs operated by retail utilities and wholesale market operators. Demand response is often the least-cost resource in competitive auctions for peak capacity, handily beating new gas-fired generation. Increasingly, the same technologies can be used to provide intelligent load shifting, not just curtailment, to balance demand and supply across hours. Smart air-conditioners, water heaters and electric vehicles are among the most promising options, but at least 30 percent of U.S. electricity demand has at least some flexibility that can be harnessed to provide grid value. The potential is even larger with full use of thermal storage — ice-storage air-conditioning, for example, or just controlling buildings so they "coast" on heat or cold stored in their own materials, providing constant comfort from use of heating and cooling equipment artfully timed to match renewable supplies.
The potential is even larger with full use of thermal storage — ice-storage air-conditioning, for example, or just controlling buildings so they coast on heat or cold stored in their own materials, providing constant comfort from use of heating and cooling equipment artfully timed to match renewable supplies.
3. Weeks to months: Within a given geographic region, renewable energy production from wind and solar does not always closely match energy demand across seasons. It has been suggested that this poses an operational challenge for a low-carbon grid: how can we meet demand on calm, cloudy days? Because this is not an issue today, and won’t be for decades, market signals haven’t yet been used to incentivize innovative solutions. However, there is much evidence to suggest that, once the market signals the need, capable and ample technologies can step in.
One promising idea gains inspiration from then-Gen. Eisenhower’s advice: "If a problem can’t be solved, enlarge it." That is, expand its boundaries until they encompass what the solution requires.
For example, it’s easier to solve heat and electricity problems together than separately. Coupling electric heating devices (PDF) with combined heat and power plants allows "excess" renewable electricity on windy and sunny days to offset heating by gas, improving the economics of an otherwise "oversized" renewable energy system that provides enough energy on calm and cloudy days. Enlarging the problem also can mean expanding it geographically. Wind and solar power tend to reinforce each other by working best in different weather patterns and times, especially when interconnected over larger geographic areas. Indeed, long-distance high-voltage transmission lines, under development in the U.S. today but already commonplace in China and other countries, can move low-cost, high-capacity-factor wind from remote regions to cities, where it nicely complements the availability patterns of local renewables.
Or, even more elegantly, it is possible to move the consumption of data center load, via fiber-optic cable, to grids with excess renewable energy. This can allow the growing energy demand of the cloud to literally follow the sun and wind around the world, without any incremental spending on dedicated transmission.
If these technologies are not commercial today, that is not for a lack of technological prowess or some limitation of physics. Rather, it is because the issues of seasonal energy balancing are decades away, and we have yet to unleash the power of the market to address them.
From a shouting match to a symphony
The past few months of discourse on the future of the grid have been shrill and reminiscent of a shouting match between incumbent baseload plants (and their lobbyists), on one hand, and pro-market interests on the other. With falling demand and plummeting energy prices due to currently low (but inherently volatile) natural gas prices, the incumbents, without question, are losing. Even brand-new gas plants can’t scrape by, and are being resold for pennies on the dollar, according to a recent UBS analysis. Power plant owners are left to resort to appeals for bureaucratic interference and "studies" that find novel ways to help them shout louder, so that today’s winners can be forced to pay yesterday’s losers.
We need to cut short this shouting match and let the markets do their job. Markets, at their best, can act as the conductor of a symphony — coordinating when and where individual energy resources, like instruments in an orchestra, can provide the most value to end customers. The sweetest symphonies integrate many instruments, at the right time, at the right volume. No instrument plays all the time, but the ensemble continuously creates beautiful music.
Right now, the conductors coordinating the U.S. grid are sending unmistakable signals that we certainly don’t need any new baseload, and that the time has come for existing baseload plants to consider planning for their final bow. And, contrary to a common misconception, it’s not even renewables that are killing the economics of fossil-fired power plants — it’s cheap gas. If old baseload power plants can’t handle the vagaries of commodity fuel prices, how can they hope to survive the rising wave of cheap wind and solar?
The trillion-dollar question
Thus, we are faced with a choice, and the stakes are high. About 50 percent of operating thermal capacity in the U.S. will reach retirement age by 2030 — and that timeline might accelerate if demand remains flat or even shrinks, putting additional pressure on profitability. If we were to take at face value the DOE study’s outdated assumption that "baseload power is necessary to a well-functioning electric grid," we would reinvest, like-for-like, in the same kinds of power plants that have led us straight into this contentious shouting match in the first place.
In doing so, we would lock in massive costs to pay for new assets, many of which likely will become "stranded," in the utility parlance, as advances in efficiency, renewables and other clean energy alternatives rapidly erase the value of those investments. That would place a handicap on our economy’s ability to compete globally, particularly against China, India and Europe, which are already shifting more aggressively to renewables to lock in ultra-low electricity costs for decades to come. In addition to the stranded asset risk, this pathway would lock in gigatons of carbon emissions, either directly (if we actually run all those plants) or indirectly (by squandering capital on resources incompatible with climate stabilization).
Alternately, we can take to heart the lesson learned from the experience of leading U.S. states and other countries that a flexible grid, based largely on renewable energy, is an equally reliable, less costly and much less risky way to power our growing economy. We can choose to leverage the renewable energy industries’ incredible propensity for innovation in reducing the delivered cost of wind and solar energy. And we can unlock new innovation, from today’s leading American firms as well as from plucky upstarts, in both supply- and demand-side flexibility to best integrate this low-cost energy.
We only have limited opportunities to make the right choice. Today, we should seize the opportunity presented by the DOE study and the current debate on wholesale market reforms to ensure that tomorrow’s electricity markets work in service of, not contrary to, our society’s goals.