Friday, July 01, 2016

EVs and The Service Station of the Future

Tesla Motors is apparently in talks with Sheetz, Inc. to install electric vehicle (EV) Superchargers in the latter's chain of gas stations. This caught my eye, because I was involved in a much earlier effort to install EV recharging facilities in service stations in the late 1990s. It wasn't just ahead of its time; it was stymied by some of the same economic challenges noted in the Washington Post article, as well as physical and regulatory issues that weren't mentioned.

The logic of an alliance between Tesla and gasoline retailers like Sheetz seems sound. Tesla embarked on its strategy to build a network of quick-rechargers in order to sell more cars. Its Superchargers are likely to be more effective in that role if they're installed in places that are both convenient to highways and offer a variety of other amenities for drivers, while they wait 15 minutes or more to top up their car's range. High-volume fuel retailers like Sheetz have already optimized their sites for convenience of location, and they have a wider range of food and beverage choices than the average gas station.

They also provide another essential feature: space. When Texaco was evaluating adding rechargers for GM's ground-breaking EV1 electric car to its Southern California retail network nearly 20 years ago, the fire marshals with whom we met insisted that high-voltage electricity and pumps dispensing volatile fuels like gasoline could not share the same pump island. They had to be widely separated for safety, and few of our L.A. locations had large enough footprints for that. Sheetz, by contrast, typically has large stations--many in rural or suburban locations--that could accommodate EV charging without endangering customers filling up with gas or diesel.

Another obstacle I encountered at Texaco was that EV rechargers are expensive, while electricity is cheap. Even if you're allowed to charge customers for it--we weren't, for regulatory reasons--it takes a lot of usage to pay back the substantial investment in equipment and installation. With EV sales still occupying a small niche in the market, that calculation hasn't changed much in the intervening decades. However, Tesla's primary motivation isn't to make money selling electricity, but to generate profits and support its stock price by selling more premium EVs. I would hate to see the standalone P&L for Tesla's growing Supercharger network, but that's beside the point.

This resolves a major hurdle for Sheetz and other fuel retailers that might want to add EV recharging to expand their customer base, or "green up" their image to enhance the loyalty of current customers, especially among Millennials. The profitability of such an investment would still be questionable, even if they sold EV owners lots of premium coffee and snacks while they wait. But if someone else is footing most of the bill for the added hardware, the extra revenue in the convenience store is all upside.

The service station of the future has been slower arriving than my colleagues and I envisioned when we developed Texaco's first global scenarios for the future of energy nearly twenty years ago. Sales of EVs and cars running on hydrogen have not grown as fast as we expected, while the improving performance of gasoline cars has raised the bar for alternative vehicles. However, current trends suggest that our vision of facilities offering a diverse mix of transportation energy was more premature than wrong. I will be very interested to see how Tesla and Sheetz or others move ahead with this idea.

Tuesday, June 21, 2016

Another Step Backward for Nuclear and the Environment

I don't normally do breaking news, but today's announcement by PG&E and a coalition of environmental groups on retiring the Diablo Canyon nuclear power plant in California within 8-9 years merits immediate comment.

Given the enormous social and political challenges PG&E faced in undertaking the re-licensing of the facility when its current operating licenses expire in 2024 and 2025, this action is understandable, though regrettable. I lived in California when Diablo Canyon was planned and built. It was sufficiently controversial in the 1970s, and the environment has only become more contentious. Extending the operating licenses of nuclear power plants to 60 years has become typical elsewhere, but the utility's board must have concluded that it was a non-starter in today's California.

However, we should not be misled by press-release language about replacing "power produced by two nuclear reactors...with a cost-effective, greenhouse gas free portfolio of energy efficiency, renewables and energy storage." Under California's extremely aggressive renewable energy and storage targets, the alternative energy mentioned here was coming, anyway, but it was intended to replace higher-emitting sources like out-of-state coal and in-state natural gas generation. Until there is an overall surplus of zero-emission energy--when?--the energy mix is a zero sum game.

This agreement--perhaps the best deal possible under the circumstances--thus represents the net loss of 18 billion kilowatt-hours (kWh) per year of zero-emission electricity. That's equivalent to 9% of all utility-scale electricity generated in California last year. The state went through a similar event in 2013 with the permanent shutdown of the San Onofre Nuclear Generation Station between L.A. and San Diego. As I noted at the time:

How much emissions will increase following the shutdown depends on the type of generation that replaces these units. If it all came from renewable sources like wind and solar, emissions wouldn’t go up at all, but that’s impractical for several reasons. Start with the inherent intermittency of these renewables, and then compound the challenge by its scale. Even in sunny California, replacing the annual energy contribution of the SONGS units would require around 7,200 MW of solar generating capacity, equivalent to nearly 2 million 4-kilowatt rooftop photovoltaic (PV) arrays. That’s over and above the state’s ambitious “Million Solar Roofs” target, which was already factored into the state’s emission-reduction plans.

Grid managers from the state’s Independent System Operator indicated that in the near term much of the replacement power for SONGS will be generated from natural gas. Even if it matched the mix of 71% gas and 29% renewables added from June 2012 to April 2013, based on “net qualifying capacity”, each megawatt-hour (MWh) of replacement power would emit at least 560 lb. more CO2 than from SONGS. That’s an extra 4 million metric tons of CO2 per year, or 8% of California’s 2010 emissions from its electric power sector and almost 1% of total state emissions. If gas filled the entire gap, or if the natural gas capacity used was not all high-efficiency combined cycle plants, the figure would be closer to 6 million metric tons, equivalent to the annual emissions from about 1.5 million cars.


So far, the state's environmental data supports this conclusion. Although offset by larger imports of low-emission power from out-of-state, there was a noticeable uptick in greenhouse gas emissions from in-state generation from 2013 to 2014. (See Figure 8 in the 2016 California GHG Inventory.) 

California will get more renewables either way, but shutting down Diablo Canyon when it still has decades of useful life left represents a net loss to California consumers, PG&E shareholders, and to the global environment. 


Thursday, June 16, 2016

Could the Hydrogen Economy Run on Ethanol?

  • Plans for a fuel cell car running on ethanol look like a clever way to circumvent the obstacles faced by other fuel cell vehicles.
  • However, it is not clear that ethanol's perceived logistical benefits or emissions profile would give Nissan an edge in the competitive market for green cars.

Japan's Nissan Motor Co., Ltd. made headlines this week when it announced plans to produce a fuel-cell car that would run on ethanol, instead of hard-to-find hydrogen. As reported by Scientific American, the company expects to commercialize this approach by 2020, even though competitors like Toyota already have fuel cell cars in their showrooms. It's an interesting choice. Ethanol seems to offer logistical advantages over hydrogen, but the technical challenges involved aren't trivial, nor is ethanol without drawbacks from an energy or environmental perspective.

Fuel cells have long promised a different and potentially superior path to electrifying automobiles, compared to battery-electric vehicles (EVs) with their limited range and relatively long recharging times. One of the biggest obstacles has always been the lack of infrastructure and supply--hydrogen must first be liberated from water, methane or other compounds--and the problems of storing sufficient quantities of it on board. I've driven prototype fuel-cell vehicles (FCVs) and found the experience pretty similar to driving a regular car, as long as you have a hydrogen filling station handy.

Nissan makes the case that ethanol (chemical formula C2H6O) is much easier to source and distribute than gaseous hydrogen, and the process for making it give up its hydrogen is routine, at least under laboratory conditions. However, as the alternative energy research subsidiary of my former employer, Texaco Inc., found in pursuing a similar concept with gasoline, it's one thing to do this in a bench-scale device and quite another to do it in a size and shape that will fit easily and safely in a car and run as reliably as an internal combustion engine. I suspect Nissan's engineers have their work cut out for them for the next four years.

The bigger questions about this approach are more basic: Does it make sense from an economic, energy and environmental perspective, and can it find a large enough market? Consumers already have a wide range of green alternatives from which to choose, ranging from Prius-type hybrids (gasoline only), plug-in hybrids (gasoline + electricity) and battery EVs, not to mention the continuous improvement of non-electric cars. 

Nissan didn't include many numbers in the documents accompanying its press release, but the chemistry and math involved are pretty simple. At 100% efficiency, a gallon of ethanol could produce just under 0.8 kilograms (Kg) of hydrogen (H2) using the standard steam-reforming process. The best efficiency I could find for this ethanol-to-hydrogen conversion  was around 90%, so in the real world that gallon of ethanol would yield around 0.7 Kg of H2--enough to take Toyota's Mirai FCV about 46 miles. That's pretty good, considering that same gallon in a Chrysler 200 equipped as a flexible fuel vehicle (FFV) would drive an average of just 21 miles. Fuel cells are much more efficient than internal combustion engines.

The economics of operation don't look bad, either. If we use today's average US price for E85 (85% ethanol + 15% gasoline) of $1.87/gal. as a proxy for an ethanol retail price, that equates to around 4 ¢/mile, using the Mirai's published fuel economy data. That's about 15% cheaper than a Prius on regular gasoline at this week's US average of $2.40/gal., but it's also around 10% more expensive than a Nissan Leaf using off-peak electricity in northern California.

Emissions are trickier to assess. There's a lively and growing controversy about whether biofuels produced from crops can truly be considered carbon-neutral, even in places like Brazil where the yields from sugar cane are so high. There's much less controversy that the production of most US ethanol from corn is anything but a net-zero-emission endeavor. Corn requires fertilizer sourced from natural gas, and ethanol refineries consume gas (or coal) and electricity in their production process. In any case, when Nissan characterizes their planned ethanol FCV as having "nearly no CO2 increase whatsoever", they are either oversimplifying a very complex discussion or taking a large leap of faith. 

We can count the CO2 coming out of the tailpipe of such a car, and it would need a tailpipe because the onboard ethanol converter would emit about 12.5 lb. of CO2 for every gallon of ethanol converted to pure H2, plus some CO2 from the ethanol burned to heat the unit. My back-of-the envelope calculation gives a figure of 135 grams of COper mile, or 20% lower than a Toyota Prius on gasoline. It would not be a Zero Emission Vehicle (ZEV), though of course an EV running on average grid electricity isn't really a ZEV, either, except in isolated regions or at specific times of day.

Even if there aren't any deal-killers here, I'm skeptical about Nissan's fundamental assumption that the ethanol infrastructure for their FCV would be that much easier to develop than the H2 infrastructure other FCVs require. That's because of the cost and ownership structure of the retail fuels business, which as I've argued previously helps explain why your corner gas station is unlikely to sell E15 (85% gasoline, 15% ethanol) any time soon, despite the EPA having approved it for newer cars

At least in the US, most gas stations are owned by small businesses, not by the oil companies whose brands they display. Margins are slim, and these folks don't have deep pockets, so adding a new fuel like pure ethanol or the ethanol-water mix that Nissan suggests, poses a difficult business decision: Do you take over an existing tank and stop selling diesel fuel, or premium gasoline with its high margins? Or do you rip up the forecourt to add a new tank, which entails being out of business for months--or even longer if you discover that one of your existing tanks is leaking? Either way, the investment costs and disruption to current customers are significant, in exchange for selling what at first would certainly be a low-volume product. When I was in the fuels supply & distribution business, we would have called that kind of decision a "no-brainer."

If Nissan can't encourage enough service stations to add ethanol or an ethanol/water blend--E85 would not work--to their product mix, do they start their own service station network? That seems unlikely. And if you buy one of these cars in a few years, should you carry a case of vodka in the trunk as an emergency range-extender? That's only half-facetious.  

I give Nissan credit for pursuing a novel option for making fuel cell cars more viable, as an alternative to today's range-limited EVs. Ethanol looks like a cost-competitive source of hydrogen, and it is at least easier to store than H2 gas or liquid H2. However, they face practical and marketing challenges that might well offset most of the advantages the company claims to see. The ethanol FCV could encounter the same chicken-and-egg dynamic as FCVs running on hydrogen, or indeed any new model requiring a fuel that is not distributed at scale today. It will be interesting to watch their progress.



Thursday, May 26, 2016

On Track for a Golden Age of Gas?

  • The global energy industry must overcome significant new challenges if natural gas development is to achieve the vision of a Golden Age of Gas.
  • Low energy prices and reduced investment are only half the battle as regulations complexify and organized opposition grows. 

Five years ago the International Energy Agency (IEA) issued a report entitled, "Are We Entering a Golden Age of Gas?" Gas development was booming, from both conventional resources and US shale deposits, and gas was widely seen as a vital tool for reducing greenhouse gas emissions. Much has happened since then, including a collapse in global oil prices, the signing of a new climate agreement in Paris, and a broadening of the anti-fossil-fuel focus of climate activists. If we're still on the path to a golden age of gas, the ride will be bumpy.

This is probably most evident across the pond, where Nick Butler, the Financial Times' respected energy analyst, observed this week, "Unless something changes radically, Europe has passed the point of peak gas consumption." He cited Germany's ongoing "Energiewende" (energy transition) which in order to maximize wind and solar and minimize nuclear power, ends up squeezing gas out between renewables and much higher-emitting coal.

Earlier this month France's Energy Minister announced she was pursuing a ban on imports of US shale gas--effectively any gas from the US--since France already bans domestic fracking. That strikes me as a textbook example of having to keep making bad decisions to be consistent with the first one, but it's their sovereign choice.

As the IEA defined it at the time, this Golden Age would entail faster growth in gas demand in every major sector, compared to the agency's main "New Policies" scenario in its then-current annual World Energy Outlook (WEO). They anticipated compound average growth of 1.8% per year, much faster than oil or coal, with gas consumption ending up 13% higher than the WEO's projection for 2035. That's like adding an extra Russia or Middle East to world gas demand within 20 years.

One gauge of whether that still seems realistic can be found in the US Energy Information Administration's (EIA) just-released 2016 International Energy Outlook. The EIA's long-term forecast actually has gas consumption growing slightly faster than IEA's Golden Age track in the developed countries of the OECD between now and 2035, but with a slower ramp-up to essentially the same end-point in the non-OECD countries.

Of course one forecast can't really validate another, so let's consider how some of the big uncertainties that the IEA identified in the 2011 report have shifted, starting with energy pricing. After oil's recent rebound, oil and gas have fallen by around half their 2011 US prices. That makes investments in oil and gas exploration and production considerably less attractive. Nearly $400 billion of projects have been canceled or deferred, globally, setting up slower growth in production from both gas fields and oil fields with associated gas in the near-to-medium term. This deceleration is evident in EIA's latest monthly Drilling Productivity Report for US shale.

With the contract price of liquefied natural gas (LNG) often tied to oil prices or competing with pipeline gas that has also fallen in price, large gas infrastructure projects like LNG plants look less attractive, too. We've already seen cancellations of new facilities in Australia and Canada. Fewer LNG export facilities are likely to be built in the US than previously planned. All this means less new gas reaching markets where it can be used.

Cheaper oil also reduces the attractiveness of gas as a transportation fuel. Although increasingly popular as a cleaner fuel for buses, natural gas hasn't made much headway in US passenger cars. However, this application has been growing in places like Italy and Iran, for different reasons.

Viewed in isolation, these price-related responses seem likelier to delay, rather than derail the expectations the IEA set out in 2011. The bigger challenges come from a set of issues the IEA identified a year later, in a follow-up report called "Golden Rules for a Golden Age of Gas." As Dr. Birol, now the Executive Director of IEA, indicated then, these boil down to the industry's "social license to operate."

Transparency, water consumption, emissions including methane leaks--all on IEA's list--are some of the key issues over which companies, regulators, NGOs and activists are sparring today. The UK is a prime example. Conventional energy production is declining rapidly and a large shale gas potential has been identified by the British Geological Survey, but every attempt to drill exploratory wells has encountered strong opposition.

A new factor the IEA did not anticipate is the emergence of political movements focused on fossil fuel divestiture and a "keep it in the ground" mantra. These may be based on unrealistic expectations of how quickly the world can transition to a zero-emission economy, but they illustrate the scale of a stakeholder engagement challenge the global oil and gas industry has so far failed to meet adequately. 

Just as social media are transforming politics, they are also altering the balance of power between organizations and their critics. The gaps that must be bridged if new gas development is to remain broadly acceptable to the public are growing in ways that will demand new approaches and new strategies to address. 

Considering the shifts in the global energy mix that will be necessary to reduce global emissions in line with the goals of last year's Paris Agreement, gas ought to have a future every bit as bright as the Golden Age the IEA described five years ago. Achieving that now likely depends less on the price of energy and the scale of available resource than on convincing regulators and the public that the trade-offs involved in obtaining its benefits are still reasonable.


Tuesday, May 10, 2016

A New Angle on Carbon Capture

In my last couple of posts I looked at the difficulty of meeting ambitious targets for cutting greenhouse gas emissions (GHG) without help from the lower-emitting portions of our current energy mix. Last week ExxonMobil announced that it is pursuing a new pathway for capturing carbon from power plant exhaust. That could help revive another important strategy for large-scale emissions reduction from our existing energy sources.

Carbon capture and sequestration (CCS) has fallen out of favor, lately, mainly due to the high cost and technical challenges of the early prototypes for large-scale implementation of the technology. Not only are the initial investment costs of today's CCS hardware still very high, but it is also inherently expensive to operate. That's because of the high energy consumption of the process, resulting in a "parasitic" load on the host power plant that reduces its net output by up to 20%, making the remaining output much more expensive. That creates a large deterrent in any market that doesn't provide either direct subsidies for carbon removal, or a high carbon tax or price for traded emissions offsets.

Another reason that CCS has received less attention recently is that the costs of renewable energy technologies like wind and solar power have kept falling. To some they now look cheap enough, especially with further cost improvements extrapolated, to enable us to reach our emissions goals mainly through wider deployment of solar modules and wind turbines.

Even if that were technically feasible, like most other energy industry experts I have met I am convinced that the deep emissions cuts desired for mid-century will require implementing or retro-fitting CCS onto the fleet of coal and gas-fired power plants that will likely still be in service decades from now. CCS underpins several of the emissions stabilization wedges pioneered by Princeton engineering professor Rob Socolow and his colleagues ten years ago.

What makes the approach that ExxonMobil and FuelCell Energy, Inc. have described so attractive is that, instead of being a drain on power generation, capturing CO2 via fuel cells would actually add significantly to a facility's reliable power output. It would increase revenue, rather than curtailing it.

The clever bit, and its potential advantage over current carbon-capture technology, is that CO2 capture in a carbonate fuel cell occurs as a byproduct of the power generation step. That means that it doesn't require a big, expensive, power-hungry process unit, the only function of which is to strip CO2 from flue gas and concentrate it for subsequent shipment and storage.

These fuel cells would still require natural gas for fuel, and they would produce CO2 emissions in the process of generating electricity, though at a lower rate than the coal or gas-fired plant with which they would be partnered. However, both their direct emissions and the CO2 extracted from the power plant exhaust would come out in a highly purified form suitable for geological sequestration and stay out of the atmosphere.

That brings up an important advantage of this approach over various schemes to capture CO2 directly from the atmosphere. Although the article on the Exxon/Fuel Cell Energy development in MIT Technology Review  described the CO2 concentration in power plant flue gas (5%-15%) as "low", that is still hundreds of times higher than its concentration in air.

400 parts per million of CO2 in the atmosphere may be worrying from a climate perspective, but it is still just 0.04% of air that remains mostly nitrogen and oxygen. And the lower the concentration, the harder--and normally more expensive--it is to extract. (Green plants can do this trick cheaply thanks to billions of years of evolution combined with cost-free sunlight.)

The press release makes it very clear that this new carbon-capture technology has so far only been demonstrated in the lab. Scaling it up will require additional work, and success is uncertain. Many other promising innovations, including a host of cellulosic biofuel technologies, have failed to scale. However, its potential applications are compelling enough to justify a lot of patience and persistence. I wish them luck.




Wednesday, April 20, 2016

Out of Reach Without Nuclear and Shale

  • US emissions reduction goals for 2025 could not be achieved without nuclear power and the fracking technology necessary to extract shale gas. 
  • Recent revisions by the EPA in its estimates of methane leaks from natural gas production and use do not negate the benefits of gas in reducing emissions.
In its lead editorial yesterday, the Washington Post took presidential candidate Bernie Sanders to task for his attacks on nuclear power and natural gas. The Post focused its critique on greenhouse gas emissions and the emissions trade-offs involved in substituting one form of energy for another. That speaks directly to one of the main reasons that Mr. Sanders' argument resonates with his supporters, but it ignores an even more basic problem. The energy contribution from shale and nuclear power is so large that if our goal is a reliable, low-emission energy mix that meets the future energy needs of the US economy, we simply cannot get there without them, at least not in any reasonable timeframe.

The pie chart below shows the current sources of US electricity in terms of the energy they generate, rather than their rated capacity. This is an important distinction, because the renewable electricity technologies that have been growing so rapidly--wind and solar--are variable and/or cyclical, generating only a fraction of their rated output over the course of any week, month, or year.


For example, replacing the output of a 2,000 megawatt (MW) nuclear power plant such as the Indian Point facility just north of New York City would require, not 2,000 MW of wind and solar power, but between 7,600 MW and 9,400 MW, based on the applicable capacity factors for such installations. Now scale that up to the whole country. With 99 nuclear reactors in operation, rated at a combined 98,700 MW, it would take at least 375,000 MW of new wind and solar power to displace them. As the Post's editorial points out, money spent replacing already zero-emission energy is money not spent replacing high-emitting sources.

At the rates at which wind and solar capacity were added last year, that build-out would require 24 years. That's in addition to the 36 years it would take to replace the current contribution of coal-fired power generation. It also ignores the fact that intermittent renewables require either expensive energy storage or fast-reacting backup generation to provide 24/7 reliability.

That brings us to natural gas, the main provider of back-up power for renewables, and the "fracking" (hydraulic fracturing) technology that accounts for half of US natural gas production. Fracking has transformed the US energy industry so dramatically that it is very hard to gauge the consequences of a national ban on it, even if such a policy could be enacted. Would natural gas production fall by a third to its level in 2005, when shale gas made up only around 5% of US supply, and would imports of LNG and pipeline gas from Canada ramp back up, correspondingly?

Or would production fall even farther? After all, one of the main factors behind the rapid growth of shale gas in the previous decade is that US conventional gas opportunities in places like the Gulf of Mexico were becoming scarcer and more expensive to develop than shale, which was higher-cost then than today. Either way, the constrained supply of affordable natural gas under a fracking ban would not support generating a third of US electricity from gas, vs. 20% in 2006. So we would either need even more renewables and storage--in addition to those displacing nuclear power--or, as Germany has found in pursuit of its phase-out of nuclear power, a substantial contribution from coal.

One of the primary reasons cited by Mr. Sanders and others for their opposition to shale gas, aside from overstated claims about water impacts, is the risk to the climate from associated methane leaks. Here he would seem to have some support from the US Environmental Protection Agency, which recently raised its estimates of methane leakage from natural gas systems.

Methane is a much more powerful greenhouse gas than carbon dioxide (CO2), so this is a source of serious concern. However, a detailed look at the updated EPA data does not support the contention of shale's critics that natural gas is ultimately as bad or worse for the climate than coal, a notion that has been strongly refuted by other studies.

The oil and gas industry has questioned the basis of the EPA's revisions, but for purposes of discussion let's assume that their new figures are more accurate than last year's EPA estimate, which showed US methane emissions from natural gas systems having fallen by 11% since 2005. On the new basis, the EPA estimates that in 2014 gas-related methane emissions were 20 million CO2-equivalent metric tons higher than their 2013 level on the old basis, for a year-on-year increase of more than 12%. This upward revision is nearly offset by the 15 million ton drop in methane emissions from coal mining since 2009, which was largely attributable to gas displacing coal in power generation.

In any case, the new data shows gas-related emissions essentially unchanged since 2005, despite the 44% increase in US natural gas production over that period. The key comparison is that the EPA's entire, updated estimate of methane emissions from natural gas in 2014, on a CO2-equivalent basis, is just 2.5% of total US greenhouse gas emission that year. In particular, it equates to less than half of the 360 million ton per year reduction in emissions from fossil fuel combustion in electric power generation since 2005--a reduction well over half of which the US Energy Information Administration attributed to the shift from gas to coal.

In other words, from the perspective of the greenhouse gas emissions of the entire US economy, our increased reliance on natural gas for power generation cannot be making matters worse, rather than better. That's a good thing, because as I've shown above, we simply can't install enough renewables, fast enough, to replace coal, nuclear power and shale gas at the same time.

What does all this tell us? Fundamentally, Mr. Sanders and others advocating that the US abandon both nuclear power and shale gas are mistaken or misinformed. We are many years away from being able to rely entirely on renewable energy sources and energy efficiency to run our economy. In the meantime, nuclear and shale are essential for the continuing decarbonization of US electricity, which is the linchpin of the plans behind the administration's pledge at last December's Paris Climate Conference to reduce US greenhouse gas emissions by 26-28% by 2025. That goal would be out of reach without them.

Thursday, April 14, 2016

Lessons from the Coal Bust

Yesterday's Chapter 11 filing by the largest US coal mining company is the latest in a series of coal bankruptcies. While factors such as regulations and poorly timed acquisitions have played a role, this trend reflects the parallel technology revolutions playing out across the energy sector. Here are a few key lessons from the ongoing coal bust:
  • There are many other ways to make electricity, and coal brings nothing unique to the party. In a growing number of markets it is no longer the cheapest form of generation, and it is certainly the one with the most environmental baggage, from source to combustion.
  • Coal-fired power generation is in competition with alternatives that are already producing at scale, like nuclear and natural gas generation, or growing rapidly from a smaller base, like renewables. It may not compete with all of these in every market, but few markets lack at least one of these challengers.
  • The costs of renewables and gas have fallen significantly in recent years, due to major technology gains. Coal has also benefited from some improvements in scale and end-use technology. Today's ultra supercritical coal plants are more efficient than coal plants of a generation ago, but they are more expensive to build, even without carbon capture (CCS). However, wind and solar power continue to grow cheaper and more efficient, while gas has benefited from resource-multiplying production technologies and advanced gas turbines that can exceed 60% efficiency and ramp up and down rapidly to accommodate the swings of intermittent renewables.
  • Despite all of these threats, coal is not on the verge of being forced out of power generation, even in developed countries where all the above factors are at work. Replacing its enormous contribution to primary energy supply and electricity generation will be a very heavy lift, particularly where another major energy source like nuclear power is being phased out. Germany is the prime example of that.
Consider what it would take to replace the remainder of coal in the US power sector. Last year coal generated 33% of US electricity, down from nearly 45% in 2010. Gas picked up 70% of the drop in coal's power output, but that still left coal's contribution at 1,356 Terawatt-hours (TWh) or about 6x the grid contribution of all US wind and solar power last year. (A Terawatt is a billion kilowatts.)

Displacing coal completely from US electricity would require doubling the 2015 output of US gas-fired power generation and a roughly 36% increase in US natural gas production. By comparison, the US nuclear power fleet would have to more than double. If coal were to be replaced entirely by renewables, which in practice probably means gas pushing coal out of baseload power and renewables reducing gas-fired peak generation, the hill looks steep.

Last year the US added 7.3 GW of new solar installations and 8.6 GW of new wind turbines. Assuming they were mostly sited in locations with reasonable solar or wind resources, their combined annual output should be around 35 TWh. At that pace it would take another 36 years to make up what coal now generates. It's true that net annual wind and solar additions continue to grow at double-digit rates, but keeping that up may get harder as the best sites become saturated and earlier wind turbines and PV arrays reach the end of their useful lives in the meantime.

In other words, driving coal from here to zero seems possible but very difficult, even with an all-of-the-above strategy in a market without demand growth. And if electricity demand continues to grow, as it is globally, or resumed growing in the US and other developed countries to enable a big shift to electric vehicles, the prospect of retiring coal entirely recedes into the future.