I’m currently most of the way through Alex Epstein’s Fossil Future, and I find the arguments within very convincing.
In (extreme) brevity, they are:
Fossil Fuel use and the cheap energy it enables is responsible for unprecedented human flourishing for billions. Billions more need access to ever-increasing amounts of cheap energy to flourish.
Any argument against fossil fuel use must argue that its side effects (CO2 warming the planet) overwhelm the good they do by providing cheap energy. These side effects must be so bad that it’s worth compromising the safety and flourishing of billions of humans to curtail their use. Such an argument must also prove that those negative side effects are beyond what humanity is capable of adapting to or overcoming, given cheap energy provided by fossil fuels.
No such argument is justifiable, given the current state of climate science.
Does anyone (preferably those who’ve read the book, although I don’t want to restrict answers to just those people) have an opposing view/opinion, and if so why?
I’d like to do my intellectual homework on this one, and actively seek disagreement, given how convincing I’ve found the argument so far.
Haven’t read the book, but I’m a cleantech consultant with a materials science and physics background. I spend my days helping companies find and adopt and strategize around new technologies.
It’s a good question and worth engaging with. I’ll start by saying that, first of all, we need to consider that climate change is not an existential risk. It’s a serious risk where not adapting fast enough can cause large amounts of suffering for billions of people, but it won’t wipe us out or prevent the continuance of technological civilization unless we do incredibly stupid things in response to said suffering. Which is not impossible. But Earth isn’t going to become Venus, or anything like that.
I’ll follow that up with an observation that in the excerpts of this and the author’s other books that I’ve found, he mostly accurately represents a number of misunderstandings and misrepresentations from the past and present of environmentalism, but he also misrepresents many aspects of the technological and economic state of competing energy sources, and of the current state of climate science. He makes some very isolated demands for rigor, like complaining about nuclear being expensive relative to fossil fuels without (AFAICT) noting that this is because it’s held to a uniquely high regulatory standard way out of proportion to it’s actual risks in ways that drive up construction times, financing costs, and plant engineering requirements, while preventing scale-up from letting industry grow enough to realize experience curve cost reductions and preventing plant design iterations over time. The sections I’ve been able to access read like someone who decided their bottom line before they decided what to write above it.
More concretely:
Today, in much (possibly most) of the world, kWh for kWh, solar (unsubsidized, without energy storage) is the cheapest source of electricity on the grid. This happened within the last <10 years depending on location, and I think it’s important to note that Alex Epstein does not seem to have adjusted his rhetoric in response to this shift. Yes, not every market can currently afford the capex/opex tradeoff, or has the land and weather available for large scale deployment, or has a grid that can remain stable under increased intermittency, but over time these problems have been getting steadily less severe as the equipment gets cheaper, grid management gets better, natural gas use grows and gets used as peaker plants, and existing dispatchable fossil fuel capacity gets used for load following more. Today nations can credibly do things like ban building new coal plants, and there’s no outcry from industry and no flight of data centers out of said country.
Energy storage is also getting cheaper rapidly. For lithium-ion batteries it’s been on the order of 10x in the past decade, down to around $150/kWh today. This, combined with improving power and energy densities, are why we’re starting to see more electric cars, and more plans by utilities to installing energy storage. I’ve done the math for a few models of electric and serial hybrid electric cars at current retail gasoline and electric prices, and it seems common to find that after 40-50k miles the electric car has lower total cost of ownership. For grid storage, currently most near-term plans I’ve seen talk about batteries on the order of 1-4 hrs of backup. That’s enough to stabilize the grid to short-term fluctuations of supply and demand, improve efficiency by reducing need for spinning reserves and load following, and just generally allow each type of power plant, including fossil fuels, to be operated in the most efficient way for that plant instead of varying due to the instantaneous needs of the grid. For other types of batteries it’s harder to say because they’re less commercially mature, but there’s been a lot of progress in areas like vanadium flow batteries, which are easier to scale to longer durations, and iron-air batteries, which are less efficient but should be much cheaper to build (making them potentially suitable for seasonal storage, where you slowly charge them whenever there’s excess generation to use a few times a year.
Fuels. Hype about batteries and hydrogen aside, it will likely be decades before there’s a truly viable alternative to hydrocarbons for many diesel and jet fuel use cases. Biofuels are pretty cost competitive these days but biomass availability is finite and probably can’t sustainably scale enough to meet a large fraction of current usage, let alone future demand. E-fuels—which use electricity to capture CO2, make hydrogen from water, and turn it to hydrocarbons—are very expensive ($15-40/gallon) due to high electricity requirements and high capex. Yet we see the EU (in one of it’s bold but less ridiculous moves) mandating use of sustainable aviation fuel and companies signing offtake agreements for more e-fuels than suppliers can provide. Why? Because it’s a very early stage technology, and we can expect capex to fall over time like it does for anything else not unavoidably dependent on large amounts of precious metals. Because in the near term people are building plants in places where they can get nearly free electricity from trapped assets (like hydro in Norway and wind in West Texas). And because by the time it does need to scale up, that electricity is going to be coming from renewables, and peple are going to be building plants wherever renewables are cheapest. That means (even with current solar power tech) <$0.02/kWh, compared to a global average of around $0.19 today for the grid as a whole. That’s why we see Australia announcing a 50GW solar power project to be built in the outback by 2030 to make hydrogen, ammonia, and methanol for export.
In the long run, that which is not sustainable will not be sustained. Global fossil fuel use will eventually end, and we’re lucky humans are in a position to decide how to do that on our own terms. An abrupt end would genuinely be terrible. A decision 30 years ago to phase fossil fuels out by today would genuinely have left humans with no viable replacement that could have been cost-effectively scaled up in time. That’s no longer the situation we’re in. We’ve solved many of the problems technologically, and some of them economically. We can see the shape of both the major remaining problems and of multiple solutions forming for each of them that can feasibly be ready by the time we need them.
I’d encourage you to read Fossil Future, because Alex addresses a lot of the points you make.
One of the points he emphasizes is that (quoted from my response below)
He talks extensively about how nuclear power is the best option, and has only failed to replace fossil fuels because the regulatory requirements have effectively criminalized it.
Epstein supports the eventual replacement of fossil fuels with better alternatives, but believes it should happen over time as those replacements become genuinely better in the market, without government interference or distortion.
Thanks for the correction on nuclear.
I understand the impulse to want to avoid government interference, but I don’t think it’s reasonable unless he is making detailed proposals for a lot of other policy changes to happen in advance or in parallel. I’ll set aside past subsidies and government investments in fossil fuel extraction, use, and infrastructure to support same except to note that the playing field did not start level. For anyone involved. Instead I’d focus on how existing policy choices have shaped and constrained the development of alternatives. 1) We have pipelines for oil and methane, not hydrogen and methanol and CO2. 2) Our grid (due to Edison-era tech constraints) uses AC, which is great for spinning turbines, but means solar and wind need more equipment to interface with the grid, and similarly makes charging batteries (for vehicle or grid-scale storage) less efficient. It also makes long-distance transmission less efficient. 3) For safety (to avoid errors), airports only have one fuel type on site, and it’s kerosene. Any alternative has to be refined and blended to be identical in function and properties to kerosene. 4) Electricity market structures vary extensively and are defined by regional and national government policy. How prices are set, how capital expenses get planned and approved, who gets to compete in the market, all heavily regulated in ways that, for historical reasons, favor fossil fuels. This situation has getting better for a while, but slowly and very unevenly.
He’s just plain wrong on wind and solar being useless for fuels and industrial energy use. It’s going to take decades to scale up and implement the relevant tech, because any shift that large and complicated does and because there just isn’t enough renewable energy production yet to make sense at scale. But:
1) Global shipping companies like Maersk are ordering dual-fueled ships that can burn methanol, and partnering with companies scaling up production of e-methanol (from CO2, water, and electricity).
2) Major steel companies like POSCO are spending billions to replace blast furnaces with hydrogen primary steelmaking facilities. There are also technologies for molten electrolytic and aqueous iron ore reduction that have been getting lots of investment and attention in recent years.
3) Ammonia production, for fertilizer, is one of the prime use cases for green hydrogen, and we’re seeing more MW and GW scale projects in development and planning, respectively, to make it from wind and solar. It’s also potentially a fuel that can be used in ships and fuel cells. Along with methanol, ammonia is one of the most plausible candidates for long distance hydrogen transport, the same way we transport fossil fuels today, because it’s stable/dense/liquid and has existing supply chains used to handling megatons per year.
4) Conversion of CO2+H2 to various hydrocarbons, for making diesel and jet fuel and chemical feedstocks, is still very early commercially and not yet price competitive, and the market is very distorted by near-term subsidies and long-term mandates. But it can be done, and is being scaled up starting in the past few years. And it’s not just startups, the airlines, major oil companies, catalyst developers, chemical companies, and EPCM firms are getting involved through investment, R&D, plant construction, and other ways. At this point it’s mostly a matter of normal engineering, iteration, scaling, and having enough wind and solar capacity available, to get to more reasonable costs. Just like with anything else.
5) I have serious doubts about battery-electric and hydrogen-powered trucks and planes in general, but they are possible options, especially for lighter duty trucks and shorter distance flight routes.
6) As a side note, the same infrastructure needed to make synthetic hydrocarbon fuels is also a way to solve many of our trash and recycling problems. Waste gasification, converting biomass and plastic and solvents etc. to syngas, can feed trash into the same plants to make whatever hydrocarbons we decide we need, while rendering hazardous compounds non-hazardous and separating out the remaining inorganic components for further sorting and recycling. The efficiency used to be pretty terrible, but it’s gotten much better and cleaner, and we’re starting to see scale-up for municipal, agricultural, hazardous, medical, and other wastes.
Also, I like to point out that some of these problems can help solve each other, in time, with scale. Put enough BEVs with V2G capability on the road and the grid suddenly has 1-2 days of stored electricity available, enabling much higher wind and solar penetration rates while also reducing need for hydrocarbon fuels and reducing BEV cost of ownership. Demand for synthetic fuels and hydrogen can act as a form of energy storage too, soaking up excess production on sunny and windy days to stabilize demand by adjusting how they run based on real time wholesale electricity prices. They also act as price arbitrage and long-distance energy transport, increasing price stability. Renewable and nuclear power, in general, have much more stable cost profiles over time relative to oil and gas prices, too.
If you really want to get fancy and long-term, there are wavelength-tuned LEDs now available for indoor farming that, with tandem solar cells, are efficient enough to potentially grow multiple acres of crops through indoor farming with one acre of panels, with less water, less fertilizer, less loss due to pests, less variability from weather, and higher yield than conventional farming. There are even transparent solar cells that can be installed on farmland to shade plants, but that only use the frequencies plants can’t; these can actually increase yields while reducing water consumption.
And ultimately if 5-10% of world GHG emissions ends up intractable, so what? Solving 90% of the problem gives us 10x as long to solve the rest. And carbon capture and storage or utilization, including direct air capture, become much more feasible when the problem is that much smaller, and you’re doing it in a world with so much more experience using the component technologies in economically viable ways.
I don’t think anyone with real decision-making power is arguing for switching all the fossil fuels off immediately, and they’re very good at tuning out the activists who do. I also think the needed timelines for competitiveness of alternatives range from “it already happened, we’re just catching up” to “decades, but still much faster than the premise of this book implies.” And achieving a distortion free market requires significant government interference, both to force companies to pay for negative externalities and to compensate them for positive ones. I’d estimate that getting close to to distortion-free would require vastly more intervention and subsidy than is being done or than will actually be needed to replace almost all fossil fuel use, because of how far we’ve already come technologically and because of the secondary benefits of switching.
Edit to add: historically, nuclear should have been the best option for baseload power. It could have gotten there for slower load-following use if capex had stayed low and gotten lower, in better regulatory regime. The higher capex/lower fuel cost would not have let nuclear replace faster load-following and peaker plants at any time in the past 50 years. Hydro and geothermal could have helped with that, but I’m not sure to what degree, Today energy storage technologies are at a point where they could fill in the gaps like the can for renewables, but that’s only true because of all the government policy kickstarting and supporting their development that Epstein doesn’t think should have happened.
Thank you for the long and detailed response! This was exactly the sort of stuff I was looking for.
I think the foundation of Epstein’s argument—that we should be prioritizing human flourishing (which requires large and increasing amounts of cheap energy) and carefully evaluating the costs and benefits of our choices (not just the costs) is largely accurate.
That being said, you’ve made me think that Epstein’s treatment of solar, wind, and other alternatives to fossil fuels is perhaps too short and/or not up-to-date.
Out of curiosity, how long would you expect it to take for a large percentage of the world (say, already-developed economies) to move to getting 90% of our energy needs from non-carbon-emitting sources? Based on my own understanding, what you’ve said, and the current state of permitting and environmental review, I’d guess no sooner than 2050 at the very earliest, with 2075 being more likely.
I think that’s a very reasonable range, with more developed countires likely getting there by the 2050s and others a bit later.
Most countries and companies that have stated net-zero goals have set them in the 2040-2050 time frame. And that’s not just because all the current leaders will be retired by then :-). All the ones that don’t have aggressive commitments will take longer.
I think 2075 is a bit conservative actually? Economically speaking I would doubt any developed country is still building coal plants by 2030 or gas plants by 2040, the ones that do get built are already increasingly using designs chosen to be retrofittable for other fuels (hydrogen, methanol, biofuels, etc.), and more of them are being used for peaker plants and not baseload (lower utilization, so they make up a smaller fraction of generation relative to nameplate capacity). Plus the existing stock will mostly all be retired by the early 2050s. By the early 2030s I’d expect renewables + storage to be cheaper to build than anything but natural gas plants even in developing countries.
Like you said, permitting and reviews are some of the big limiters here. We are plausibly talking about 10,000-200,000 km2 of solar panels, worldwide, for a complete transition combined with continued economic growth. The Australian Outback might be one of the best places in the world to make green hydrogen, for example, but it’s also one of the largest mostly-undeveloped wilderness regions remaining. Somehow arguments like “Yeah, but even more will be lost if we don’t replace fossil fuels, and all the other options are worse” seem to lack the power to overcome project-specific objections.
Last I heard, the big question was what positive-feedback “tipping points” exist, and at what CO2-level they become triggered. This would give quite wide error bars on what average heating is caused by a given quantity of cumulative emissions. If we can burn all the fossil fuels, turn the rainforests to desert, and vaporise all the methane clathrates, and still not end up like Venus, that’s… reassuring, I guess
Doing all that would be very bad. Not Venus, but enough to greatly decrease Earth’s carrying capacity for humans and everything else, for a long time. We really should want that not to happen. But methane clathrate release turns out not to be as rapidly self-reinforcing as once feared, and at this point there’s no longer an economic or technological reason to think we’ll need to keep using fossil fuels and cutting down rainforests long enough to get to that point. We’re already seeing slowdowns in net deforestation, in part because in some countries there is net reforestation, though rainforests in particular are being lost faster in part because these exist mostly in less developed countries, but even then it’s (2020 aside) slowing down.
If Epstein’s thesis is, broadly, “cheap energy from fossil fuels is awesome and climate change isn’t that bad”, weaknesses would be likely to fall somewhere under these, classified in increasing controversy:
Climate change might be worse than he’s positing. Particularly, climate is a global system we only partly understand, and our error bars for the effects of inadvertently perturbing it may be quite large
Cheap energy may be obtainable from non-fossil sources. Epstein is keen on nuclear energy (why?), but, as AnthonyC points out, solar & wind are getting surprisingly cheap (for example https://www.bloomberg.com/news/articles/2021-06-23/building-new-renewables-cheaper-than-running-fossil-fuel-plants). Obviously that only gives you a bonanza of cheap energy when the wind blows and the sun shines, but “the sometimes-free-energy bonanza will destabilise the grid!” feels a long way from common-sense
GDP may be possible to decouple from energy use by purely technical means: compare the computation-per-watt-hour of a Pentium with that of an iPad. Compare a Passive-House-standard building with an average building. How far are we from the optimal frontier? There will be, at some point, a limit, because we’re operating with optimal designs and can’t get any better, but we might be quite far from that point—countries with more energy-tight housing often have codes that require that, rather than the market doing it spontaneously? Why is that? Principal-agent problems? “Market for lemons” informational issues?
GDP may be possible to decouple even further from emissions use via selective lifestyle changes (higher-density cities, more bicycles, fewer automobiles)
GDP may be argued to be decouple-able from human welfare. Broadly, this is the “degrowth” argument. Critique here, for example (https://www.noahpinion.blog/p/people-are-realizing-that-degrowth)