Editor’s note: This article was originally published in July 2018.
For around 10 years, the conventional wisdom in the energy sector has been that natural gas is ascendant. Coal is dirty, and it’s getting expensive, but it’s too early to jump all the way to renewable energy. To get from the fossil fuel present to the renewable future, we will need ... a bridge.
Natural gas is meant to be that bridge, a way to reduce our emissions relative to coal while we work on scaling up renewables. (The shift from coal to gas is a big part of why US emissions have declined over the past few years.)
In its role as a bridge, natural gas seems to have a comfortable future. First, it will replace coal and nuclear “baseload” plants, and then, as renewables grow to supply the bulk of power, it will provide flexibility, filling in the gaps where variable renewables (wind and solar) fall short. By playing these multiple roles, natural gas will long outlive coal and prove useful well into the latter half of the 21st century. It will enjoy a long, slow exit.
Or so the story goes.
Around 2015, though, just five years into gas’s rise to power, complications for this narrative began to appear. First, wind and solar costs fell so far, so fast that they are now undercutting the cost of new gas in a growing number of regions. And then batteries — which can “firm up” variable renewables, diminishing the need for natural gas’s flexibility — also started getting cheap faster than anyone expected. It happened so fast that, in certain limited circumstances, solar+storage or wind+storage is already cheaper than new natural gas plants and able to play all the same roles (and more).
The cost of natural gas power is tethered to the commodity price of natural gas, which is inherently volatile. The price of controllable, storable renewable energy is tethered only to technology costs, which are going down, down, down. Recent forecasts suggest that it may be cheaper to build new renewables+storage than to continue operating existing natural gas plants by 2035.
That means natural gas plants built today could be rendered uncompetitive well before their rated lifespan. They could become “stranded assets,” saddling utility ratepayers and investors with the costs of premature decommissioning.
Meanwhile, gas’s environmental reputation has suffered from a series of reports, most recently a study in Science, showing that gas’s lifecycle methane emissions are much higher than previously estimated and could seriously dent gas’s climate advantage over coal. (See author and activist Bill McKibben for an extensive exploration of this point.)
Even if methane emissions are reduced, they can’t be reduced to nothing. And the US needs to completely decarbonize — get to net-zero carbon emissions — by mid-century. Natural gas simply isn’t compatible with a net-zero-carbon future unless a massive infrastructure is built to capture and bury its carbon emissions. Until and unless that happens, natural gas must eventually be eliminated.
Luckily, there is good news. While it is far too early to say that we’ve reached the end of the natural gas bridge, we can perhaps say that the end has come into sight — somewhat hazy, but you can see it if you squint just right.
Here’s how this post is going to go. We’ll take a look at recent prices of renewable energy and storage relative to natural gas. We’ll run through several recent examples of regulators or utilities either turning against natural gas or enduring political blowback for supporting it. Then we’ll check out a couple of recent reports that try to quantify the threat to natural gas. And then we will conclude with some big takeaways.
A quick note about natural gas plants
By way of background: New natural gas power plants come in one of two flavors. Combined-cycle gas turbines (CCGT) run two successive cycles — first burning natural gas in what is effectively a jet engine, then capturing the waste heat to run a secondary steam generator — and are more efficient, generally cheaper, and run more often. Open-cycle gas turbines (OCGT) are less efficient and produce more expensive power, but they are faster. They jump in during peaks in power demand — thus, “peakers.”
Beating CCGT is mainly a matter of cost; wind or solar can do that on their own, once they get cheap enough. Beating peakers doesn’t just require being cheap, it requires being fast, flexible, and dispatchable. For that, you need storage.
Renewables and storage are getting cheap
According to the consultancy Lazard, the all-in, “levelized cost of energy” (LCOE) from some renewables is already lower than the LCOE of a lot of fossil fuels in many cases, even without subsidies and without environmental benefits factored in. Wind is the cheapest energy of all, and utility-scale solar is competitive with the cheapest natural gas.
(These are 2017 prices; cost declines have continued since then.)
Meanwhile, storage is moving within striking range of gas peakers.
Peakers range in price from around $8.18/kW-month (according to consultancy Brattle Group) to as high as $15/kW-month (for a recently proposed Entergy plant). Compare that to a recent solicitation from utility NVEnergy, which elicited bids for 50 MW, four-hour storage projects at as low as $6.11/kW-month. (That’s with the 30 percent federal tax credit, so the “true” price is $8.72/kW-month, still right up against the cheapest peaker.)
It’s worth noting that batteries have advantages over peakers other than price. They’re faster to build. A natural gas plant takes three to five years, while Elon Musk promised South Australia he would build them the world’s largest battery bank in 100 days or it would be free — and he delivered.
Batteries can also provide a range of other services to the grid, like voltage support and frequency regulation, faster and more precisely than natural gas plants. They can ease grid congestion and help avoid new infrastructure investments.
“While current battery storage chemistries may not be able to do all of the same work as gas,” Kelly Speakes-Backman, CEO of the Energy Storage Association (ESA), told me, “the portfolio of storage technologies may be able to address the majority of the same functions, especially as longer-duration chemistries emerge.”
When seeking resources to meet anticipated peaking needs, utilities still often default to natural gas plants, but that’s changing as storage gets cheaper; many utilities are growing suspicious of natural gas and leaning toward technology-neutral solicitations. (ESA has a guide for utilities on how to integrate storage into their planning.)
Many states (including California, Massachusetts, New Jersey, and New York) has passed storage mandates, forcing utilities to procure storage. And the Federal Energy Regulatory Commission recently issued Order 841, which mandates that all major US energy markets allow storage to participate on equal footing.
The times are a’changing. Let’s take a look at how some regulators and utilities are responding to these shifting dynamics.
Some regulators and utilities are rethinking natural gas
All over the country, examples are popping up of regulators and utilities either rejecting new gas plants or fighting through a hail of resistance to get them built. Gas may still be the default choice for many utilities, but environmentalists aren’t letting them make it thoughtlessly — it is a struggle to build a gas plant, almost anywhere in the US.
- In March, Arizona regulators took the unprecedented step of rejecting the integrated resource plans (IRPs) of the state’s major utilities. They cited too much reliance on natural gas and the risk of stranded assets. Then they placed a nine-month moratorium on new natural gas plants larger than 150 MW and asked utilities to model scenarios with high penetrations of renewables and storage. (That’s from a utility commission composed of Republicans, in a state with no renewable energy mandate.) The utilities are now actively soliciting storage.
- In Minnesota, an administrative law judge recently recommended that the Public Utilities Commission reject a petition by Minnesota Power to build a giant gas plant (the Nemadji Trail Energy Center), because the utility had not given enough consideration to clean alternatives.
- In New Orleans, Entergy recently won approval from the city council to build two new natural gas plants in the city, but only after a furious campaign of resistance. (The company was subsequently busted for paying people to show up to public meetings and support the plants.)
- In Colorado, the state’s biggest utility, Xcel Energy, filed a plan to close two of the states coal plants and replace them with a package of clean energy resources — some of the cheapest on record — and zero gas.
- In June, Michigan’s second-largest utility, Consumers Energy, submitted an IRP that included 5,000 MW of new solar, 550 MW of new wind capacity, a range of investments in storage and demand management — and zero new natural gas plants. But the state’s biggest utility, DTE, recently won approval for a massive, $1 billion new CCGT, despite analysis from environmental groups showing that a clean-energy portfolio approach would be cheaper.
- In Nevada, NV Energy is pulling together an IRP that would involve 1,000 MW of new solar and 100 MW/400 MWh of energy storage. In the process, it’s getting record-breaking low bids. (The 300 MW Eagle Shadow Mountain Solar Farm is signed to a 25-year solar power purchase agreement at a flat rate of $23.76 per megawatt-hour. That’s the cheapest US solar power to date, but the record has probably been broken since I typed this.)
- In New Jersey, the state Department of Environmental Protection recently approved a new gas plant in Meadowlands, but environmentalists are rallying, calling on the governor to reject it.
And then there’s California, where the catastrophic natural gas leak at Aliso Canyon exposed the fragility of the natural gas distribution system and prompted a state of emergency. Within six months, 70 MW worth of battery systems were installed in the canyon and helping to avoid the feared electricity outages.
When Southern California’s natural resources group proposed a giant new natural gas plant, advocates pushed back hard, marshaling evidence that renewables and storage would be faster and cleaner; the natural resources group subsequently shelved the plan. (Though it’s not yet completely dead.)
California regulators forced utility Pacific Gas & Electric to solicit bids for energy storage rather than continue payments to three natural gas plants (one CCGT, two peakers). Pacific Gas & Electric has since opted to replace them with the world’s largest storage procurement to date: four projects totaling a whopping 2,270 MWh.
They also rejected a plan by the Southern California Edison electric supply company to refurbish its Ellwood Peaker Plant, pushing the utility to bring more storage and solar online instead. “At this time, absent very compelling circumstances, we should be directing all of our investments in infrastructure and energy to clean energy resources,” said commissioner Clifford Rechtschaffen.
In April, the Glendale City Council voted to table a proposal for a natural gas plant in their city, to explore cleaner options. And so on.
This (very partial) list shows that natural gas’s value proposition is no longer taken for granted. Gas’s purported dominance was based on necessity — it is cheaper and cleaner than coal but able to do things renewables can’t. That sense of necessity is crumbling. There are now emerging alternatives, packages of clean energy resources that can do all the same things, at competitive costs, with no greenhouse gas emissions.
When will natural gas begin feeling pressure in earnest? Some recent research suggests it may come far sooner than popularly expected.
New research shows that building new gas plants is getting risky
We know that it is possible for portfolios of clean energy resources — wind, solar, batteries, demand response — to generate reliable power, ramp up and down quickly to follow loads, and provide grid services, effectively doing everything gas plants can do with no carbon emissions. And we know that all the components of such portfolios are getting cheaper at a rapid clip.
So the only real question is when, not if, those costs will dip below the costs of gas plants.
The Rocky Mountain Institute (RMI) addressed that subject in a recent report. They begin by noting that as much as $1 trillion may be committed to new gas plants by 2030, with up to $100 billion already announced. But clean energy is coming up fast behind natural gas and may well catch up before new plants reach the end of their life cycle.
RMI researchers took a close look at four proposed gas plants in different parts of the country — “two announced combined-cycle gas turbine power plants, planned for high capacity-factor operation, and two announced combustion-turbine power plants, planned for peak-hour operation” — and compared their costs to regionally appropriate aggregations of solar and other distributed energy resources.
They find that “clean energy portfolios can often be procured at significant net cost savings” today. Specifically, in three of the four cases, clean energy portfolios are lower net cost (8 to 60 percent) today, and if costs continue coming down or a small price is put on carbon, clean energy will be cheaper in all four cases. (The problem, of course, is that all the costs of clean energy are up front, while fuel costs for power plants string out for years. Persuading utilities and regulators to compare net costs at all can be a challenge.)
More strikingly, “the new-build costs of clean energy portfolios are falling quickly, and likely to beat just the operating costs of efficient gas-fired power plants within the next two decades.” Depending on gas prices, within two decades — well under their rated lifespan — natural gas plants built today will be uncompetitive. That’s a lot of potential stranded assets.
RMI also ran a scenario that saw more than half the retiring fossil fuel plants between now and 2030 replaced by clean energy portfolios. Using conservative assumptions, they find this strategy would come at small net savings and “unlock a $350 billion market for renewables and distributed energy resources through 2030, while avoiding 3.5 billion tons of CO2 emissions over the same time period.”
Of course, if natural gas prices rise or a serious price is put on carbon anytime in the next two decades, natural gas’s reckoning could come even sooner. Earlier this year, Jim Robo, the CEO of energy giant NextEra Energy, predicted that he would be selling energy from solar farms with four hours of energy storage for 3.5 cents/kWh within a few years.
“That’s lower than the operating costs of existing coal and nuclear,” Robo said. “That’s a fact that most of the rest of the industry hasn’t come to grips with yet.”
“As energy storage gets deployed at scale in the coming years, gas plants will be left to compete for the remaining megawatts that storage isn’t a good fit for,” Ray Hohenstein, market applications director at energy storage company Fluence, told me. “With over 100,000 megawatts of peakers already built in the US today, it will become more and more difficult to justify new gas plants economically — with a high risk of those assets becoming stranded — and the least efficient of the existing plants will be competed out of the market.”
One interesting question is which kind of natural gas plant will be pushed out first. It might be different for the US than for the world broadly.
According to GTM Research, it is peakers that are going to fall first in the US. With the rate of decline in battery prices, they will be competing directly with peakers in a few years and beating them consistently by the mid-2020s.
Of the 20 GW of peakers slated for between now and 2030, GTM estimates, 10 GW or more could get beat out by batteries. “I can’t see a reason why we should ever build a gas peaker again in the US after, say, 2025,” said Shayle Kann, then a senior adviser at GTM.
Back in 2015, Robo said, “post-2020, there may never be another peaker built in the United States — very likely you’ll be just building energy storage instead.”
Notably, the global outlook for natural gas plants is somewhat different, at least according to Bloomberg New Energy Finance (BNEF). In its annual New Energy Outlook, BNEF sees CCGTs plateauing around 2030 as renewables eat into midload and then baseload. But peaker capacity will expand “almost 300 percent, from 268GW today to 1,002GW by 2050, as the need for flexibility increases with growing wind and solar penetration.”
The rise in peakers comes mostly in India and China, where BNEF believes that renewables+storage won’t be enough to meet “extreme peaks in demand.”
It may be that peakers will die out first in relatively sophisticated developed-world energy markets, where novel clean-energy portfolios take hold, while CCGTs will slow first in developing nations. Or it may be that neither dies out and we miss our collective carbon targets! Time will tell.
“Gas plant replacement with storage is happening as we speak, and is accelerating due to economic forces, broader energy production trends, and increased climate change awareness,” says Ryan Kladar, a senior analyst at consultancy Strategen. “There are plenty of things to slow it down, but I currently view natural gas as a bridge to nowhere.”
One thing we know: energy changes faster than we think it will
Of course, we shouldn’t forget that forecasts out to 2050, like BNEF’s, are best seen as a genre of science fiction. Nobody knows what the world is going to look like in 2050. The energy sector is already changing with vertiginous speed. In the past 10 years, developments in clean energy have come so fast and furious that forecasts have been revised again and again. Even 10-year forecasts have been rendered goofy.
There’s no reason to think the pace of change will slow any time soon. Quite the opposite.
So there’s a great deal we do not, and cannot, know.
We don’t know where the price of natural gas will go, what policymakers will do, or what kinds of economic or other disruptions might be in store. More to the point, though, we scarcely have any idea what clean energy is capable of.
We don’t know what’s possible once gigawatts worth of electric vehicles are connected to the grid and charged or discharged depending on needs. We’ve barely scratched the surface on demand response and have only the faintest glimmer of what we can do with millions of appliances hooked up to the grid for use as thermal storage or flexible demand.
We cannot predict what new industries or uses might arise around dirt-cheap renewable energy, or what kind of demand might swarm in to absorb the abundant energy underneath the duck curve, once markets are properly aligned. We talk a lot about smart grids that can support transactive power systems, but have barely begun to build and connect them.
We have no idea what’s going to happen.
But we do know a few things. We know the US needs to decarbonize as fast as possible (as all developed nations do), and that eventually the federal government will get its act together. We know that natural gas, at least without carbon capture and sequestration, is not compatible with a zero-carbon future and must eventually be eliminated.
We know that clean energy resources, in all their varied glory, can do all the things natural gas power plants can do. We know that the cost of natural gas power is tethered to the price of natural gas and has little room to fall, while the cost of clean energy is tethered only to technology, which has gotten and is continuing to get cheaper and cheaper.
And we know that clean energy has defied all our forecasts, maturing and falling in cost faster than even the most optimistic advocates predicted. We should have some confidence that will continue.
Natural gas still has enormous global momentum. But it has already gotten risky to build a new gas plant in the US. The UK is turning away from gas. So is South Australia. It will happen first in developed markets that already have adequate capacity and then, depending on how cheap clean alternatives get, growing markets next.
Clean energy is approaching in natural gas’s rearview mirror, fast, always faster than anticipated. Think about what today’s forecasts will look like in 10 years. Which side would you bet they err on?
My bet: The bridge is shorter than anyone now predicts.