The transition to a zero-carbon economy has many elements. The most obvious is the need for clean electricity, but there are hundreds or thousands of other sectors to be addressed, from energy storage to steelmaking to the manufacturing of fertilizer.

Often, there are many potential solutions under development. How do we pick a winner?

150 Direct Air Capture Startups

I found myself pondering this question after listening to a recent episode of the excellent climate podcast, Catalyst. Andy Lubershane of the climate VC fund Energy Impact Partners notes that startups are exploring an incredible number of approaches to pulling carbon out of the atmosphere:

I think we found a list somewhere and it's a pretty credible list of over 150 direct air capture startups. … It turns out that if you want to suck carbon dioxide out of the atmosphere and then dump it somewhere, there's a ton of different ways to do it theoretically.

Carbon capture may be an extreme example, but we also see a wide range of concepts being tried for energy storage and next-generation nuclear power (fission and fusion), and multiple approaches in hundreds of other sectors. Who decides which solution should win out?

If “The Market Decides”, Who is The Market?

I think the textbook answer is that, at least in the US, “the market” decides which technology should win out, meaning something like:

1. Inventors notice an opportunity – a new market opening up, or the potential to improve on solutions in an existing market – and propose solutions.

2. Investors decide which ideas look promising enough to deserve early funding.

3. Customers decide which offering best meets their needs at the lowest price.

But there’s a lot more to the story. Inventors, investors, and customers are fallible individuals and organizations, working from limited information, analyzing it imperfectly, and subject to ulterior motives. For a market like direct air capture – with many solutions on offer, none of them mature – it is difficult for participants to choose.

The podcast notes the valuable role played by sophisticated early buyers. Tech giants like Google, Microsoft, and Facebook built internal teams to evaluate sources of clean electricity for data centers. These teams had the resources to select promising solutions, and the buying power to help those solutions succeed. Something similar is now playing out in the carbon capture market, with Microsoft and Stripe as important early buyers.

Government agencies also play an important role, both through funding programs, and by writing rules which may explicitly or implicitly favor particular solutions.

The Many Ways This Can Go Wrong

Lubershane argues that the plethora of solutions in markets like energy storage or carbon capture leads to analysis paralysis. Investors and customers are hesitant to act; many different solutions struggle along at small scale, and it takes too long for a winner to emerge and begin climbing the scale and efficiency curves.

Governments or sophisticated buyers can help this along by channeling resources to the most promising solutions, but especially when the government is involved, this is called “picking winners” and has a bad name. In the worst case, the process may be influenced by vested interests, allowing flawed ideas (like corn ethanol in the US) to win out. But even without vested interests, a centralized decision making can lead to a monoculture, where some promising ideas don’t get a fair shake.

Famously, “the market” doesn’t account for externalities, such as the climate impact of carbon-burning technologies. Mission-driven inventors may choose to pursue clean technologies, but most investors and customers won’t adequately prioritize climate impact unless forced through government intervention, which is uneven and not always well-designed.

The Difficulty of Challenging an Early Winner

Once a solution has time to scale, it can be hard for another solution to come along and displace it. For example, lithium-ion batteries now benefit from massive economies of scale and highly optimized manufacturing processes. Any new energy storage technology, in its early days on the market, will have a hard time competing. This is not always bad; one reason lithium-ion has achieved such massive scale is that these batteries really do work quite well for a wide variety of applications, and there are benefits in using similar technology in applications ranging from electric vehicles to grid-scale energy storage.

But suppose an alternative technology were to come along that is fundamentally “better” than lithium-ion, in the sense that if it were developed to the same level of maturity, it would have large advantages in cost, weight, availability of raw materials, or other important considerations. That new solution might never be developed, because it would take an enormous amount of capital to catch up with lithium-ion batteries, and there might not be an investor with sufficiently deep pockets and enough risk appetite to make the investment.

This is known as “path dependence” – a technology may win out because we happened to scale it first, not because it is inherently superior. If history had taken a different path, a different (possibly better) technology might have won instead.

(Sometimes this problem can be overcome by finding niche applications where the new technology can find early success – such as Tesla starting out by making exotic sports cars, or lithium-ion batteries achieving early scale in laptops and cell phones before anyone considered them for powering vehicle. This allows the new solution to begin scaling.)

How Should We Solve The Explore / Exploit Tradeoff?

The challenge of allocating development funds and deciding which solution(s) to scale is an example of a classic problem, the explore-exploit tradeoff. We can devote resources to “exploring” new solutions, or we can use whatever information we’ve already gathered to pick a winner and then “exploit” it (put all of our resources into scaling up that solution).

The podcast host, Shayle Kann, refers to this as a choice to “develop” versus “deploy”, and notes that some people argue that the climate movement should shift toward deploying various promising solutions that have already emerged, rather than waiting to develop new solutions.

I’m sympathetic to the argument that we are exploring too many different solutions in areas like direct-air capture or next-generation nuclear, and that the best thing for emissions reduction would be to pick a few winners and focus on rapidly optimizing and scaling them. I’m also sympathetic to the idea that if we do this too early, we’ll wind up with another corn ethanol disaster.

I don’t know the best way to navigate this tradeoff. But I’m certain that our current hodgepodge of short-time-horizon investors, unsophisticated buyers, and markets that ignore externalities – even when supplemented by a handful of mission-driven investors, sophisticated early buyers, and sometimes well-designed government action – is not the best we could do. It would be nice to see the more explicit discussion of the develop-vs-deploy tradeoff with regard to clean technologies.

Share

If you’re interested in this kind of analysis of the central questions underlying an important topic, check out my AI blog, where I talk about where AI is going, what it will mean for the world, and what we can do about it.

I’ve been pretty quiet here, as I focus on my project to make sense of AI and make sure it goes well for the world, but I still follow developments in climate change mitigation. Here’s a quick summary of the ideas I took away. Think of this as a recommendation to either listen to the interview (the first third is free; the rest is behind a $7 paywall) or buy the book.

The rest of this post is my quick attempt to distill some of the key ideas from this terrific interview.

Focus On Change, Not Emissions Reductions

Our ultimate goal is to eliminate greenhouse gas emissions. But to reduce emissions by 100%, you shouldn’t begin by thinking “how can I reduce emissions by 10%”. That would be like a preteen athlete, wanting to set themselves on a path to success in the NBA, thinking “to begin with, how can I win my first NBA game?”. Rather than focusing on the end result, you should focus on the changes that are needed to enable that result.

A related point: incentivize the solution; don't penalize the problem. It's useless to try to convince fossil fuel producers (whether companies or countries) to reduce their production, because that's a fight to the death directly against their business model. Instead, look downstream – convince car buyers, electric electricity buyers, or other consumers to switch to clean alternatives, in part by promoting the development of those alternatives. Then you’re not trying to wrestle any stakeholder into acting directly against their core interests.

One implication is that “climate footprint” is the wrong way to think about emissions reductions. Instead of thinking about what you can do to reduce your climate footprint, think about your best leverage for reducing emissions. Spending $10,000 to put solar panels on your rooftop reduces your personal footprint; donating $5,000 to the Giving Green Fund doesn’t. But the latter will have much more impact.

The same applies to businesses, or countries. A battery manufacturer could insulate their factory, to reduce the need for heating and A/C. Or they could build a second factory and increase their battery production. One plan lowers climate footprint, the other doubles it. But increasing battery production is by far more beneficial to the overall course of emissions reduction.

Think Globally, Act Narrowly

Narrowly targeted policies often work better than attempts to tackle the entire emissions challenge in a single bill. For instance, an economy-wide cap-and-trade program will make enemies of every carbon-emitting industry, but will have relatively few supporters. A narrow bill that focuses on fostering a new technology rather than penalizing an old/dirty technology will have support from an array of interests that benefit from the new technology, but isn’t a direct attack on the old one and so won’t garner as much resistance. This helps explains why early subsidies and promotion of EVs, solar and wind power, and other technologies have achieved a lot.

Also, when we foster a new technology, we kick off a virtuous cycle – the cost of that technology comes down, suppliers emerge who have a vested interest in promoting that technology, and so forth. Attempts to strangle dirty solutions don’t trigger this effect. I wrote about a similar idea a couple of years back, in Should We Demand Less Supply, Or Supply Less Demand?

Similarly, Sharpe argues that when it comes to diplomacy, attempts to bring every country together to agree on broad policies, you wind up with a low-impact muddle. “Often the media reporting is all on these headline negotiations, like whether countries can agree exactly which form of words they're going to use to describe burning slightly fewer fossils.”

Instead, he argues that international efforts should be more narrowly targeted. For any given sector, such as steelmaking, there are generally at most ten countries who account for the bulk of production (or, in some cases, consumption). Getting a handful of countries to agree to serious action on a single sector is much more feasible. Specifically, Sharpe argues for limiting “scope, participation, and timeline” – each initiative should focus on a single sector (scope), among a small group of countries with leverage over that sector (participation), and near-term targets and actions (timeline).

The Battle Is Between Interests, Not Countries

Some people describe emissions reduction as a prisoner’s dilemma between countries. The idea is that each country would prefer to neglect reductions, while benefiting from reductions made by the rest of the world. Sharpe suggests that the important struggle is often between opposing interests within each country. For instance, many countries are struggling to take advantage of the low prices now offered by solar and wind power. Typically, within each country, there will be some interests in favor of moving to solar and wind power, and some interests opposed. Rather than pressuring a country to commit to emissions reduction targets, it can be more productive to support the clean energy interests within each country.

Enforced Change Stimulates Innovation

Sharpe argues that when regulations force an industry to change, this often stimulates a wave of innovation. The conventional wisdom is that regulations increase costs, but Steven Chu (Obama’s energy secretary) looked into this and found that historically, energy efficiency regulations for appliances have “actually accelerated price decline of those appliances”. His explanation is that tighter standards create evolutionary pressure, forcing companies to invest in product development instead of advertising.

In the past, I’ve found myself wondering why we keep finding clean solutions that seem straightforwardly superior to what came before. Solar power, EVs, and other zero-emissions technologies often seem to be not only greener, but cleaner, cheaper, and/or better performing than the dirty technologies they replace. That’s unintuitive; when we add a constraint (no carbon emissions), you’d expect the space of possible solutions to shrink, resulting in worse tradeoffs. But that’s often not how things seem to be turning out in practice. Perhaps Sharpe’s idea that constraints stimulate innovation can explain this.

You Should Get This From The Source

This is just my attempt to summarize a lengthy, idea-packed 88-minute conversation. If you found these ideas interesting, I highly recommend that you listen to the interview or buy the book. Again, it’s the best high-level take I’ve heard in quite some time on how to strategically approach emissions reductions.

Meanwhile, most of my writing these days takes place in my AI blog. If you’re interested in the big picture of where AI is going and what it will mean for the world, check it out.

Share

People ask me about this. I tell them that both ideas strike me as overblown. For the foreseeable future, I don’t see AI significantly changing the trajectory of climate change, in either direction. On the one hand, the amount of energy used for AI is just not that much on a global scale. On the other, the pace of emissions reduction is primarily limited by our ability to execute on things we already know how to do, rather than our ability to invent clever new things.

Using AI To Reduce Emissions

Speculations as to how AI might help reduce greenhouse gas emissions mostly boil down to “AI could make [thing] more efficient, thus reducing that thing’s emissions”. The thing in question might be supply chains, the electrical grid, forest management, heating and cooling, etc.

There is also talk of AI helping make communities more resilient to climate change, e.g. by using weather models to better plan for shifts in rainfall patterns and other impacts.

These proposals tend to have two things in common. First, they’re vague. When someone is writing about a new trend, vagueness is a bad sign. It’s a tell that there’s nothing really happening here – at least, not yet. If AI was having an actual impact on emissions reduction, or there were specific startups or projects promising an impact, all the articles talking about AI and climate change would cite examples1. The lack of such examples tells us that this is a made-up trend: people aren’t writing about it because they’ve seen it happening, they’re writing about it because they assume it will happen. I mean, obviously AI will have an impact on climate change, because – well, gosh, it’s going to impact everything, right? Spoiler alert: no, in the near term, AI will impact some things but not others and it won’t often be revolutionary, especially with regard to the sorts of physical-world activities that are relevant to greenhouse emissions.

What’s the second common element in proposals for AI-based projects to reduce emissions? To the extent that there’s actual potential in the proposal, that potential generally doesn’t have anything to do with AI. For instance, consider the idea of improving management of the electrical grid. There is absolutely room to better manage the grid, as I touched on in The Electricity Transmission Challenge. But the barriers to improved management are mostly things like “utilities are not incentivized to do it”, “management of the nation’s grid is fragmented across too many separate agencies and operators”, and “it’s a conservative industry where it’s hard to make changes”. Those aren’t problems that you solve with AI.

To put it another way, I strongly suspect that any proposal to improve grid management using AI will look like this:

1. Come up with an efficient plan for using grid resources.

2. Carry out that plan.

AI helps with step 1, but we don’t really need help there; it’s step 2 that has been holding us back.

Of course, AI is hot right now, and so everyone will be looking for excuses to slap an “AI” label on whatever they were doing anyway. Who knows; sometimes this might even succeed, as a matter of salesmanship, in attracting buy-in to a project. As a result, we may see some successful emissions reduction projects that are hyped as incorporating AI; that won’t necessarily mean that AI really was important to the outcome. Conversely, we might see projects which fail because everyone focuses on the AI part and then the project founders on the barriers to action – siting, regulatory approval, etc. – that make so many mitigation projects difficult.

Over time, there will be instances where newer AI techniques allow us to shave off a few percentage points here and there. I just don’t expect it to be a game changer anytime soon.

AI’s Use Of Electricity

From The New York Times, in October:

OpenAI’s ChatGPT exploded onto the scene nearly a year ago, reaching an estimated 100 million users in two months and setting off an A.I. boom. Behind the scenes, the technology relies on thousands of specialized computer chips. And in the coming years, they could consume immense amounts of electricity.

A peer-reviewed analysis published Tuesday lays out some early estimates. In a middle-ground scenario, by 2027 A.I. servers could use between 85 to 134 terawatt hours (Twh) annually. That’s similar to what Argentina, the Netherlands and Sweden each use in a year, and is about 0.5 percent of the world's current electricity use.

It’s worth noting that there is a lot of spin going on here. The first paragraph primes us to think big: ChatGPT “exploded”, there’s an A.I. “boom”, which could consume “immense amounts” of electricity. There’s also an implication that AI uses enormous numbers of chips (“behind the scenes”, no less – how suspicious!), but in fact “thousands of specialized computer chips” is quite a small number compared to the scale of modern data centers.

The second paragraph may leave you with the impression that AI is projected to use as much electricity as three entire countries – Argentina, the Netherlands, and Sweden. But the word “each” means that actually, the projection is only for AI to match one of those countries, not all three. So, why not just say “similar to what Sweden uses in a year”?

Spin aside, let’s dig into the numbers. Weirdly, I couldn’t easily find a figure for global emissions from electricity usage, but the “electricity and heat sector” generated 14.6 Gt of emissions in 2022. Let’s conservatively assume that electricity accounts for most of that, call it 12 Gt. If A.I. in 2027 is using “about 0.5 percent of the world's current electricity use”, that would be 0.06 Gt / year in CO₂ emissions2.

By comparison, livestock production generates around 6 Gt of emissions per year. So, if you’re worried about emissions from AI, here are some things you could do, each of which would have the same impact:

• Shut down AI entirely by 2027

• Reduce livestock consumption by 1%

• Make livestock management 1% more efficient

• Reduce methane emissions (e.g. leaks) from the energy sector by 0.5%3

• Use AI to achieve 0.1% in efficiency improvements across the global economy4

Even this paints AI in an unfairly negative light. A lot of AI chips go into data centers run by tech giants like Microsoft and Google, which are aggressively pursuing clean sources of power. Not only are these companies signing contracts to purchase electricity from solar, wind, and other clean sources, they’re often paying a premium to be early adopters of technologies such as advanced geothermal power or small modular nuclear reactors, thus helping those technologies achieve commercial viability. So data center electricity is probably cleaner than the overall electricity sector, and the tech giants are actively helping it to become cleaner still. (I’m not sure whether Amazon is contributing to the clean power transition as aggressively as Microsoft and Google.)

Also, to a substantial degree, electricity usage for AI is displacing electricity usage for other purposes. The world has only so much capacity for producing the kinds of advanced chips used for AI training and inference, and it’s running flat out. Every chip used for AI is a chip not used for, say, crypto.

In The Long Run, All Bets Are Off

I started this article by saying:

For the foreseeable future, I don’t see AI significantly changing the trajectory of climate change, in either direction. [emphasis added]

When it comes to AI, it’s not clear how far the “foreseeable future” extends. In the next five years, for sure, there won’t be much impact. We won’t be able to build chips quickly enough to amount to much net additional electricity usage, and neither will we have time to deploy any AI-assisted efficiency improvements at scale.

After 5 years, I still don’t see electricity usage from AI becoming a major concern. If we’re using AI so heavily that it starts to become a significant portion of global energy demand, then we’ve entered the “grander visions of AI” phase (see next paragraph). However, it does seem possible that we will eventually start to see substantial efficiency improvements due to AI. For instance, AIs might help develop new catalysts, enzymes, or genetic enhancements to improve efficiency in food production, chemical processing, and industrial processes. They might reduce soft costs for rooftop solar and other renewable energy projects, by helping developers fill out paperwork and even helping government agencies process it. This could sometimes be game-changing within a specific sector. But it will continue to be less important than finding ways to cut through red tape, get utilities and other incumbents to embrace change, and generally plan and execute better as a society.

If and when the grander visions of AI start to become reality, all bets are off. Maybe we’ll develop hyper-efficient chemical engines that use solar power to 3d-print anything you want using CO₂ captured from the air. Maybe we’ll spend our lives in virtual environments that don’t require us to ever get in a car, let alone an airplane, nor to use more living space than the size of your VR pod. Maybe robots will make mass-produced junk so cheap that we all buy a lot more of it, driving emissions ever higher. Maybe an AI-driven economic boom will leave everyone feeling prosperous enough to be comfortable paying a green premium to drive emissions to zero; or perhaps AI will exacerbate income inequality and political tensions to the point where advancing climate-friendly policies becomes impossible. In my AI blog, I’ve written about how a future with powerful AI will be utterly different from the present day. I don’t think we can predict how that will unfold; at best, we can focus on keeping up with developments, and being prepared to act as needed. “Chance favors the prepared mind”, as the saying goes.

Thanks to Michelle Deatrick for inspiring this post.

I’ve recently started another blog, on the topic of AI, and in particular the potential risks associated with advanced AIs. The goal of the AI blog is the same as for Climateer: to dive into a complex subject and attempt to explain it in a jargon-free fashion. ... Posts here on Climateer will be sporadic for the foreseeable future, though I have at least one more in the pipeline right now.

Readers of this blog might be interested in a post where I make an analogy between climate change and AI risks: To Address AI Risks, Draw Lessons From Climate Change.

While I’ve continued to focus on AI, progress on climate change mitigation has certainly not been standing still, and I thought it might be worthwhile for me to occasionally post links to good articles. I recently read a post on the Renewable Revolution blog, The Eight Deadly Sins of Analyzing the Energy Transition, that Climateer readers might enjoy. It reviews the reasons that projections of how quickly we will transition to clean technologies so often fall short of actual progress.

And as long as I’m posting links, I’ll share my favorite sources of climate news. For some reason, all of them are podcasts:

• Volts: if you listen to one climate podcast, make it this one. Host David Roberts is a highly engaging interviewer, covers a wide variety of topics, finds great guests, and asks them all the questions you’d like to hear asked.

• Catalyst: climatetech investor Shayle Kann covers a range of topics. Episodes are a bit shorter than the other podcasts listed here, and do a good job at teasing out the key concepts on topics ranging from rare earth minerals to electricity interconnection queues, without getting bogged down in technical details.

• Cleaning Up: British host Michael Leibreich gets good guests and isn’t afraid to go for challenging questions, making the show fun to listen to.

• The Energy Transition Show: the only non-free podcast on this list, and far and away the geekiest most technical. Very deep dives on everything relating to the transition away from fossil fuels. To try it out, you can listen to the first part of each episode for free.

• My Climate Journey: softball, but informative, interviews. Emphasis on interviewing startup founders, as well as people involved in climate finance (e.g. VCs) and workers in climate-relevant trades.

For a quick introduction to two of these podcasts, you could try the recent Volts / Catalyst crossover episode.

In the comments, please let me know if you’d like to see me send occasional linkposts of this sort (and what topics you’d be most interested in).

Solar panels, wind turbines, pumped hydro, battery storage, advanced geothermal, biomass, advanced nuclear: none of these electricity sources do much good unless they can be connected to electricity users. This turns out to be a difficult problem: so difficult, in fact, that it may be the single largest hurdle remaining on our path to phasing out fossil fuels.

Moving electricity around is inherently challenging, because we need enough capacity at every location in the country, at every moment of the day. It’s similar to the problem of moving cars around a city. When the traffic system falls short, we call that “rush hour”. When the electricity grid falls short, we call it a blackout.

In the coming years, we’re going to be using a lot more electricity, and it’s going to be coming from a wider variety of sources. This will greatly increase transmission requirements. Unfortunately, our system for planning new transmission projects in the US is terrible. (In this post, I focus on the US, but many of the issues apply across the globe.) The good news is that there are plenty of solutions available, if we can get organized to take advantage of them. The more transmission capacity we can add, the faster and cheaper the energy transition will be.

Big Demands Are Coming For The Transmission Grid

Transmission refers to the big high-voltage lines that carry electricity from generation sources to the areas where the power is needed – the equivalent of the freeway system. (As opposed to the last-mile “distribution” lines, the equivalent of local roads, which bring lower-voltage electricity to end users.)

In recent decades, improvements in energy efficiency have kept US electricity usage fairly flat, and so there hasn’t been much need to build new transmission lines.

However, demand for electricity is about to skyrocket, as we “electrify everything” in order to get away from fossil fuels. Ground transportation is shifting from gasoline to electric vehicles, heating is shifting from furnaces to heat pumps, and so forth. All the energy that is currently being moved around the country in coal trains, natural gas pipelines, and gasoline trucks will instead be flowing through the electrical grid.

At the same time, the mix of electricity sources is changing. Historically, most electricity came from coal, natural gas, nuclear, or hydropower. With the exception of hydropower, these can generally be located near the cities where their electricity is used, leaving relatively little need to transmit power over long distances. Solar and wind power use a lot of land, and are best sited in areas that get a lot of sun or wind. As a result, they can’t always be placed close to cities, and rely on long-distance transmission – sometimes very long distance. Wind power, in particular, is heavily concentrated in the middle of the country, and of course solar power is more plentiful in the south.

Solar and wind power are also much more variable than traditional energy sources, and so they get less usage out of the transmission lines they’re connected to. Imagine you have a city that uses 1 GW of power, and you build a 1 GW transmission line to either a coal plant or a solar farm. Loosely speaking, the coal plant can make constant use of the transmission line, while the solar farm only uses it during the day. With solar, the city will need an alternative power source at night, requiring a separate transmission line.

In summary, the transmission grid is going to need to move a lot more electricity, over longer distances, using a lot of intermittent sources instead of a few steady sources. The net result is that we’ll need to build an enormous amount of new lines, in a relatively short period of time, after a period where we’ve built very little and have lost the knack for it1. Studies show a need for US transmission capacity to increase by anywhere from 2x to 5x by 2050. More transmission gives us more ability to leverage cheap renewable energy, so the closer we can get to the high end of that range, the cheaper the energy transition will be.

Transmission Is Just Plain Hard

Electricity transmission is, in some ways, a uniquely difficult problem.

We don’t get much choice as to where to locate transmission lines. They have to run from the places where electricity is generated to the places where it is used. US auto manufacturing is clustered in a few regions, according to the convenience of the manufacturers, and we ship the finished cars to where they’re needed. Transmission lines have to be where they have to be, no matter how inconvenient.

Transmission lines are long, tall, and ugly. They take up a lot of space, they’re visible from a distance, no one likes having them in their backyard, and building a long line means passing through a lot of backyards.

Capacity needs to be just right in every location. If we have a shortage of wires running into Houston, extra capacity in Chicago won’t help.

We can’t “store” transmission. If a line is under-utilized at night, we can’t bank the extra capacity for use during the day2.

In other words, transmission is inherently local in time and place, to a degree that exceeds just about every other industry associated with the energy transition. If Ford underestimates demand for the Mustang Mach-E in Cincinnati, they can shift a few over from Indianapolis. If Mitsubishi makes too many heat pumps this month, they can warehouse them until next month. But if Con Edison runs out of capacity to supply New York City’s air conditioners on a hot afternoon, that’s a bad afternoon.

Summing up: transmission lines are expensive, unwanted, take years to build, and we need to have exactly the right amount of them connecting to every region of electricity generation or usage in the country, just as both generation and usage are going through the largest changes in half a century or more. That’s hard!

Not Only Is It Hard; We’re Bad At It

In the early days of electrification in the US, we had thousands of freestanding local utilities. Over time, smaller grids were merged together, and interconnections added on an ad-hoc basis. Very little of the resulting patchwork was designed with the efficiency of the overall national energy system in mind. It is a disorganized mess, resulting from individual utilities optimizing for their own local interests, often under circumstances very different from the present day.

When it comes to transmission, utility interests are rarely aligned with a coherent national grid, or even basic economic efficiency. Utilities operate in a regulated environment, with artificial profit incentives. In general, they receive a mandated return for capital expenses, such as building a power plant. They don’t receive any return for operating expenses, such as importing electricity from a neighboring utility. So from a utility’s perspective, it is always preferable to generate their own power, even if they could import electricity for half the cost.

Even if everyone involved were trying to do the best thing for the overall system, we make it hard for them to do so. Building an interstate transmission line typically requires approval from multiple utilities, the utility commission in each state, and various federal agencies, resulting in far too many cooks in the kitchen.

While coordinating among so many different organizations would be difficult under the best of circumstances, the teams in question are often under-resourced. As we’ve discussed, in recent years the rate of increase in electricity usage has been low. And renewable energy sources are often much smaller, individually, than a fossil fuel or nuclear power project, requiring a larger number of smaller grid connections. For these reasons, the various utilities and agencies are not staffed to handle the number of new grid connections that will be needed.

And of course any major project creates both winners and losers, so there are always participants who will try to stall or kill any major project. For instance, owners of legacy fossil-fuel power plants will fight attempts to bring cheap renewable power into the area. In general, the benefits of a new transmission line are diffuse – everyone in the region gets slightly cheaper or more reliable power, everyone on Earth gets slightly cleaner air. Because the benefits are diffuse, no one in particular is motivated to fight very hard for the project. Meanwhile, the downsides are concentrated – a coal plant owner loses customers, specific landowners get transmission towers as their unsightly new neighbor – and the losers are motivated to push back. As one source put it, “why should we build a line through Arizona when California gets all the benefits?”. It’s the classic NIMBY problem, but amplified, because of the large scope of a transmission project.

As if all of that weren’t bad enough, we don’t have a good system for deciding who should pay for new transmission lines. In some cases, if a new solar or wind farm can’t be accommodated on the existing grid, the project developer will be asked to pay for an entire new transmission line. This is the equivalent of going to the store for some soda, and being told that because they have to break open a new palette, your six-pack will cost $300. The result is that developers jockey for position, trying to free-ride on recently added lines that someone else paid for, and cancelling projects that would trigger a new line.

This Is Already A Really Big Problem

Given the inherent difficulty of building an efficient transmission system, and the design flaws of the US approach to grid planning, we might expect problems to arise in practice. And we’d be right. For years now, transmission bottlenecks have been a major impediment to renewable energy:

In 2005, for instance, the largest power company in Arizona proposed to build a transmission line to carry electricity to its customers from a new wind farm in Wyoming. Last month, after 18 years of legal battles and hearings and revisions, the TransWest Express project was finally approved. [NYTimes]

When a developer wants to build a new energy project, they must first wait in the “interconnection queue”, to receive permission for their project to be connected to the grid. Wind, solar, and energy storage projects currently in the queue total 2000 gigawatts of generation capacity. By comparison, “the total capacity of all existing power plants on the U.S. electric grid is 1,250 gigawatts”3. Many of these projects in the queue won’t actually be built; because the backlog is so long, developers submit applications for an array of potential projects, not all of which they expect to follow through on. This is a sign of a dysfunctional system, somewhat akin to the way early-Covid toilet paper shortages were exacerbated by panic buying when people saw store shelves emptying out.

Another sign of transmission bottlenecks is the fact that existing renewables are often “curtailed”, meaning that some of the electricity they produce can’t be accommodated on the grid, and needs to be thrown away. For instance, one source I found (but unfortunately can’t cite) states:

Total wind curtailment in SPP [Southwest Power Pool] in 2021 was 6,354 gigawatt hours, equivalent to 7.3% of total wind generation.

Yet another signal is the phenomenon of negative electricity prices. Normally, when local electricity supply exceeds demand in a particular location, the excess will be transmitted to other regions, and/or some generators will shut down. However, some power plants (e.g. coal plants) can’t shut down quickly, and others might keep running due to perverse incentives (e.g. a tax credit based on the amount of power they generate, even if the power isn’t needed). Sometimes, the amount of such “can’t stop won’t stop” power generation exceeds the capacity of the transmission grid to carry the power away, and the result is negative power prices, a sure sign of insufficient transmission capacity. From the same source:

As recently as 2015, negative wholesale prices occurred in less than 2% of all hours and locations. Since 2015, the frequency of negative prices has increased steadily, reaching almost 6% in 2021. There are hundreds of locations, mostly in the middle of the country, which now experience negative electricity prices during more than 20% of all hours.

Transmission Saves Money

The most efficient paths to decarbonizing the grid rely on a dramatic increase in transmission capacity. For instance, a 2022 report by the US NREL (National Renewable Energy Laboratory) outlines several paths for achieving a net-zero-greenhouse-emissions US electricity grid by 2035, under the assumption that electrification will lead to a substantial increase in energy demand during that time period. Here are some key numbers from that that report:

• Between 2020 and 2035, annual electricity usage increases by well over 50%.

• Over the same period, transmission capacity increases by anywhere from 27% to 187%, depending on the scenario. “Beyond already planned additions, these total transmission builds would require 1,400–10,100 miles of new high-capacity lines per year, assuming new construction began in 2026.” (Note that these figures are for 2035; earlier, when I mentioned an increase in transmission capacity of 2-5x, that was for 2050.)

• In the scenario that assumes the most construction of transmission lines, the net cost to decarbonize the grid is $330 billion. In the scenario that assumes the least construction, the net cost is $740 billion, over twice as much. (These figures do not take into account the health benefits from reduced pollution, nor the myriad benefits of reduced climate impact. When taking those factors into account, all scenarios show benefits greatly exceeding costs. The high-transmission scenario has benefits exceeding costs even without considering the impact on global warming, as the direct health benefits alone are projected at $390 billion.)

So improvements in our ability to add transmission capacity will yield substantial cost savings, as well as accelerating the climate and health benefits of the energy transition.

(For a deeper dive into the NREL report, I highly recommend the interview with one of the authors on episode 188 of the Energy Transition Show. Note that only part of the interview is available for free; it’s $7 to listen to the whole thing.)

What options do we have for adding transmission capacity? In the next few sections, I’ll describe several different approaches.

Get More From The Wires We Already Have

It’s possible to move more electricity through the existing distribution system. Some techniques:

• Dynamic Line Rating. The maximum amount of current which can be sent through a transmission line is based on the line’s ability to tolerate heat. More current produces more heat, which causes the metal cable to sag, and eventually it will sag too much. Traditionally, utilities determine capacity based on a worst-case scenario: continuous high usage on a very hot, sunny, windless day. Under “dynamic line rating”, the utility continuously monitors the actual temperature of the line, and adjusts the cap accordingly. An International Renewable Energy Agency report cites one example: “Oncor Electric Delivery, a US utility, implemented DLR [dynamic line rating] and observed ampacity [capacity] increases of 6–14% for 84–91% of the time.” A DLR project in Belgium achieved an average 30 percent increase in transmission grid capacity. DLR may be especially well suited for wind energy, for the stupidly simply reason that when wind turbines are producing their maximum output, that means the wind is blowing, which helps keep the wires cool.

• Topology optimization. Grid operators need to continuously adjust the flow of electricity through each line on the grid, in response to changes in supply and demand. This turns out to be quite difficult, and the software used to do it today doesn’t always manage to get the most possible use out of the overall system – just as a traffic jam can sometimes develop on one highway even when a parallel road is underused. A new generation of advanced algorithms promises to squeeze more electricity through the existing wires.

• Using batteries to time-shift transmission. Suppose you have a solar farm that can produce up to 500 MW of power during the day, but the local grid connection can only accommodate 300 MW. You could dump the excess 200 MW of power into a battery, and feed it into the grid at night. (If the power isn’t needed at night, it could be fed into a second battery, located at the other end of the power line.) In this way, batteries can be used to “time-shift” use of the transmission lines away from peak hours – the equivalent of staying late at work to avoid rush hour traffic.

• Dynamic pricing would raise electricity prices during time periods when the transmission grid is maxed out. The idea would be to convince some electricity users to defer their usage until a less-congested time. For instance, electric school buses don’t need to recharge the moment they arrive back at the depot; they could wait until the nighttime lull in usage. This is also known as “demand response”, and can also help alleviate other issues with renewable power, for instance by matching demand to times of day when solar and/or wind power are plentiful.

It’s hard to get a clear picture of the amount of additional capacity that could be squeezed out using these approaches, so I don’t know how much of the problem can be solved this way. But every little bit counts, and these techniques, where applicable, should be very cost-effective.

One challenge is the fact that, under the regulations in many US states, these approaches may not be very profitable for utility companies (as compared to alternatives such as building another gas plant). Per Canary Media:

Simply put, most U.S. transmission-owning utilities make money by convincing regulators to allow them to invest in new power lines and make other capital expenditures for equipment — not by making the power lines they already have work more efficiently.

Another way to avoid over-burdening the existing grid is to bypass it entirely, by locating electricity-intensive operations (including new / growth industries such as hydrogen production and water desalination) near renewable energy sources. I imagine we will see a certain amount of this. Rooftop solar of course is another way of co-locating supply and demand; but the supply of rooftops is not infinite, and the cost is substantially higher than for grid-scale solar.

Get More From The Right-of-Way We Already Have

Some of the challenges in building new transmission lines relate to finding a place to put them. Things get a bit easier if we can place higher-capacity wires on the existing right-of-way, and indeed there are a variety of techniques for this:

• Stronger wires. Because current is limited by the tendency of wires to sag when they get hot, we could just build wires out of stronger material, and indeed there are a number of approaches for this.

• Higher voltage. Heat is driven by the amount of current (amps) flowing through the wire. Increasing voltage allows more power to be moved for the same number of amps. For instance, Norway's grid operator found that “the capacity of existing 300 kV lines can be increased by almost 30% by upgrading them to 420 kV”.

• HVDC, or High-Voltage Direct Current, is a relatively new technology that uses direct current, instead of the traditional alternating current. This can require sophisticated conversion equipment at each end of the line, but is more efficient when moving large amounts of power over long distances.

• Transportation corridors – in some cases, it may be possible to add transmission lines alongside highways, railroads, and other existing right-of-way.

Again, I don’t have a clear idea of exactly how much added capacity could be added in this fashion. These sorts of upgrades can be expensive, but presumably less expensive and problematic than acquiring and clearing new right-of-way.

Get More Right-Of-Way

If all else fails, we could expand the transmission grid by, you know, expanding the transmission grid: acquire land, and build power lines on it. We did a lot of that in the 1970s, and in theory we can do it again. To make that happen, we probably need to revisit the processes for planning, permitting, executing, and paying for such projects.

The most promising approach would be to centralize authority under a single federal agency, probably FERC (the Federal Energy Regulatory Commission). FERC already plays this role for natural gas pipelines; extending that to cover major transmission lines could significantly reduce the cost and increase the speed of the energy transition in the US.

It’s The Wires, Stupid

In the next few decades, we’re going to need a lot more electricity, and the best places to get it from are going to be a lot more spread out. That means we need to get much better at the surprisingly difficult job of moving electrons around. It’s already a serious bottleneck, and the pace of renewables deployment is only going to accelerate, so we need progress urgently. Unfortunately, the current system for adding transmission capacity is almost a perfect storm of bad incentives, lack of coordination, understaffed agencies, inappropriate funding models, and excessive approval processes.

(There are a few instances of progress. For instance, California just approved a $7.3 billion plan to build thousands of miles of new high-voltage transmission. There are still many hurdles to clear, such as permitting all of the new lines, but it’s a good step, notably including a coordinated plan for the entire state grid.)

If we fail to add enough transmission capacity, the energy transition will still happen, but it will be slower and more expensive. We’ll need more nuclear power, more spare generation capacity, and it will take longer to phase out fossil fuels.

If you want to find a bright side, I suppose it’s the fact that there’s plenty of room for improvement. We have a broad array of technical solutions at hand: dynamic line rating, using batteries to time-shift transmission, higher-voltage wires, new transmission corridors, and many more. To make good and timely use of these options, we desperately need to address the institutional problems in the way transmission projects are planned today.

(For further reading, check out a recent Canary Media article, A backed-up grid threatens clean energy growth from Virginia to Illinois, that goes into more detail about the practical challenges facing renewable development in the US.)

Biomass is frequently touted as an important source of renewable energy. In some published scenarios for net zero emissions, it accounts for 25% or more of total world energy supply – roughly comparable to coal’s share today1. And yet, despite this major projected role, biomass receives comparatively little discussion. In this post, I’m going to present a general overview of the potential of biomass as a clean energy source, explain why it is not necessarily as clean as the prefix “bio” might suggest, and why models probably exaggerate its role in the energy transition.

A side note: subscribers may notice that it has been over a month since my last post. I’ve recently started another blog, on the topic of AI, and in particular the potential risks associated with advanced AIs. The goal of the AI blog is the same as for Climateer: to dive into a complex subject and attempt to explain it in a jargon-free fashion. In the first post, What GPT-4 Does Is Less Like “Figuring Out” and More Like “Already Knowing”, I review some of the key strengths and weaknesses of ChatGPT. Posts here on Climateer will be sporadic for the foreseeable future, though I have at least one more in the pipeline right now.

What Is Biomass?

In the context of energy production, “biomass” simply means any biological material that we can use to produce energy, usually by burning it. It’s a broad category, encompassing a range of materials, processing methods, and end uses, much as the term “fossil fuels” covers a broad range of materials and applications. The following are all examples of biomass energy:

• Turning trees into wood pellets to burn in an industrial power plant

• Burning municipal waste in an incinerator to generate electricity

• Processing cow manure in a “digester” (ugh) to produce methane (natural gas)

• Converting farm-grown corn into ethanol to be mixed into the automotive gasoline supply

• Cooking s’mores over a campfire

Because the term “biomass” covers so much ground, it’s important not to over-generalize. Each source and application of biomass has different pros, cons, costs, and environmental impacts.

Sources of Biomass

Sources of biomass break down into two categories. Dedicated biomass comes from plants which are grown specifically for this purpose. Trees, non-food crops such as bamboo or switchgrass, and food crops such as corn or soybeans can all be used.

(Readers of this blog may recall that I’ve touched on the subject once before, criticizing the idea of using dedicated biomass to produce electricity, primarily because it requires an exorbitant amount of land.)

Waste biomass encompasses all other sources of biological material used to produce energy. Some significant categories:

• Forestry residues (parts of the tree that are left in the forest after logging, such as branches and leaves, as well as entire trees that are too small, diseased, or otherwise unusable for lumber).

• Wood processing residues (material that makes it as far as a sawmill or other processing facility, but doesn’t find its way into the final product; for instance, bark, chips, and sawdust). Sometimes referred to as “secondary feedstocks”, as opposed to forestry residues which are “primary feedstocks”.

• Agricultural residues (any part of a crop plant that is left on the field after harvesting, such as stalks and leaves).

• Animal manure.

• Municipal solid waste (garbage, yard trimmings, some industrial waste, scrap wood, etc.).

These materials of course vary widely in availability, geographic distribution, processing requirements, and cost. Even within a particular category, there can be significant variation in size, water content, and other variables.

Applications for Biomass

Proposed uses for biomass primarily break down into four categories:

1. Burn it to produce electricity (competing with solar, wind, nuclear, and other sources of electricity).

2. Burn it to produce heat, often for industrial purposes (competing with heat pumps, hydrogen, and other sources of heat).

3. Convert it into some other fuel, to power ships, airplanes, long-haul trucks, or other vehicles (competing with electric vehicles, and other unconventional fuel sources and types, such as air-to-fuel, hydrogen, or ammonia).

4. Convert it into other chemicals, for instance to replace petroleum in the manufacture of plastic.

The proposed arguments in favor of biomass for a given use case generally boil down to an argument that solar power, wind power, and other comparatively straightforward sources of clean electricity aren’t applicable because:

• They are intermittent (unavailable when the sun isn’t shining and the wind isn’t blowing)

• The application can’t easily be made to run on electricity (e.g. high-temperature industrial heat for steel and cement manufacturing; regions with no reliable electricity supply)

• Batteries can’t provide sufficient range (e.g. long-haul aviation)

Biomass is not the only solution to these challenges, but it is one of the contenders.

Separate from energy production, biomass can be used as a source of carbon for sequestration (negative emissions); for instance, Charm Industrial is converting corn stalks into “a stable, carbon-rich liquid” and storing it underground.

And of course some forms of biomass contemplated for energy supply can also be used as food, especially animal feed.

How Much Waste Biomass Is Available?

Dedicated biomass has the unfortunate property of requiring significant amounts of land (the theme of my previous biomass post). Waste biomass doesn’t impose additional land requirements, but there is only so much of it available. How much?

It’s hard to find a clear answer. I imagine this is due to the fact that there are so many different types of biomass, each with local variations. Most of the sources we found report on just one part of the picture, such as certain categories of waste biomass or a certain region, and the assumptions are not always clear. Furthermore, the question is not really how much waste biomass exists, it’s how much we might use in practice, which depends on the location, cost and impact of extraction, and how aggressively you believe the source can safely be harvested (I’ll say more about this later).

For what it’s worth, here are some data points we found. All figures are for 2050, unless otherwise mentioned, and are in units of exajoules (EJ) per year.

• A recent report from the IEA projects 63 EJ of energy supply from “organic waste streams” and “forest and wood residues”, and states that this is “well below estimates of global sustainable bioenergy potential”.

• A different IEA report projects 51 EJ of bioenergy supply from “wastes and residues” in 2030; this roughly agrees with the previous source (which projected 46 EJ in 2030).

• A 2011 paper in Biomass and Bioenergy projects 28 EJ from “residues on cropland”, in a “business-as-usual” scenario, whatever that meant in 2011. This seems broadly consistent with the previous figures, since cropland is just one source of waste biomass.

• A 2007 paper in Progress in Energy and Combustion Science projects 76-96 EJ from crop and wood residues and wastes (doesn’t appear to consider manure or municipal solid waste).

• A 2018 paper in International Journal of Greenhouse Gas Control states that in 2100, incineration of municipal solid waste could generate 32 EJ.

For perspective, the first data point of 63 EJ per year is equivalent to roughly 10% of global energy usage in 20202. So at least in principle, waste biomass could be a significant source of energy.

Price vs. Supply

As for any commodity, the supply of waste biomass will change depending on the price it fetches. The cost of collecting waste biomass varies wildly according to circumstance; expensive sources won’t be tapped unless the price is high.

The research here was difficult to interpret, but just to convey the general idea, here is a figure from a 2015 paper in GCB Bioenergy. This graph projects availability of agricultural and forest biomass residues in 2050 (in EJ again), as a function of price:

As expected, price and supply increase together. A limit emerges at around 45 EJ. At that point, the article explains, “costs increase rapidly due to increases in transport and harvest costs as the marginal supply is increasingly spread out over larger areas”. The article also notes that transportation is the biggest cost component of forest biomass.

This graph chops off at a price of $10 per gigajoule. In terms of electricity, assuming 40% conversion efficiency, that would translate to $90 per megawatt-hour. For comparison, one source puts the fuel cost for a natural gas plant in the US at around $25 to $30 per megawatt-hour. In both cases, that’s just the price for the raw fuel. Building and operating the biomass power plant, and processing the fuel, would further increase the wholesale cost of the resulting electricity.

Carbon Impact of Biomass Fuels

I’d always assumed that biomass is carbon neutral, basically by definition. The carbon atoms came from the atmosphere, because the carbon in a plant originates in the atmosphere; photosynthesis allows plants to break down CO₂ and keep the carbon atom. So when we burn that biomass, we’re just putting the CO₂ back where it started. (There is also animal-based biomass, such as manure, but this ultimately originates in a plant.)

It turns out that there are a few flaws in this idea, as discussed in an article in Scientific American. First of all, by collecting biomass, we might be interfering with a carbon sink. For instance, suppose we collect and burn some forest waste. Without human interference, some of that material might have wound up sequestered in the soil. By collecting it, we’re negating the carbon sink, which is equivalent to introducing new emissions.

Secondly, even if a particular piece of wood was destined to eventually burn or rot, that might have been years down the road. By burning it today, we put that CO₂ into the atmosphere sooner, giving it more time to contribute to the greenhouse effect. Those short-term effects are important, affecting the peak temperature that we’ll reach on the way to net zero emissions, which in turn determines how much harm will result. The impact is especially bad when we log a forest for fuel, as opposed to collecting waste from trees that were being cut down anyway; it may take the better part of a century for the forest to regrow. From the Scientific American article:

“We call it ‘slow in,’ as in it takes a long time for the carbon to accumulate in the forest, and ‘fast out’—you’re burning it so it goes into the atmosphere rapidly,” said Beverly Law, an expert in forest science and management from Oregon State University.

A Natural Resources Defense Council report notes that logging old-growth forest for fuel is especially harmful, as it will be a long time (if ever) before the land returns to a state containing as much carbon as before it was logged. At this point, every remaining acre of old-growth forest on the planet should be protected from any sort of use that disturbs the trees.

The carbon impact can be reduced using BECCS, or BioEnergy with Carbon Capture and Storage. In this process, biomass is burned for energy, and the carbon dioxide exhaust is captured from the smokestack and stored somehow. BECCS is often touted as a renewable energy source with negative emissions, because the carbon starts out in the atmosphere (as discussed earlier) and ends up underground. In practice, it’s not always that simple. As we just saw, “the carbon starts out in the atmosphere” can be an overly simplistic analysis, depending on the source of the biomass. Smokestack CO₂ capture is not 100% effective. Finally, gathering, processing, and transporting the biomass all consume energy and have the potential to generate emissions. So BECCS can provide energy with reduced or even negative CO₂ emissions, but it’s a complex process, involving a lot of overhead and potential side effects.

Other Impacts of Biomass fuels

Carbon emissions aside, biomass usage can have a variety of impacts, depending on the source and how the biomass is used.

Assuming the biomass is eventually burned – whether directly, or after conversion to some sort of fuel – the combustion will likely emit pollutants: not just carbon dioxide, but potentially other greenhouse gases (such as nitrogen oxides and carbon monoxide), as well as particulate emissions.

When we collect forest or agricultural waste material, we may be robbing the soil of important nutrients such as phosphorus and potassium, reducing soil health and productivity, and/or requiring more use of fertilizer. The material that’s removed may also have played a role in protecting soil from erosion. We were unable to come to any clear understanding of how much biomass can safely be harvested, and recommendations seem to be all over the map. For instance, this chart from a paper in the journal “Energy, Sustainability and Society” lists various jurisdictions as requiring anywhere from 10% to 100% of “coarse woody debris” to be left behind when harvesting biomass from forests. Some of this is due to differences in local conditions, but I suspect much of the variation also stems from from differing opinions or out-of-date regulations. Titus et al. write, “most research is site-specific or relatively short-term and cannot necessarily predict how harvesting affects productivity or biodiversity on operational scales, or elsewhere or over the long-term.”

That said, harvesting waste biomass can also have positive effects. One publication notes that there can be positive impacts, such as enhanced seedling growth, or avoiding excessive nitrate accumulation in nearby bodies of water (from “nitrogen-rich logging residues”). Overall, the impacts of biomass harvesting appear to vary wildly according to the location, circumstances, and harvesting practices. From Energy, Sustainability, and Society:

Furthermore, forest biomass can be co-managed with conventional timber to create bioenergy feedstock from material that might otherwise be burned on-site or left to decompose, or removed to reduce fuel loading in fire-prone areas or facilitate replanting of harvested sites. Where there are markets, forest harvest residues can increase the economic value of products from managed forests, and may also stabilize carbon (C) stocks in fire-prone forests and thus prevent conversion of some forest land to other land uses because of the value of these C stocks.

When it comes to dedicated biomass, the overwhelming problem is land usage; especially when arable land is used for biomass crops such as corn or soybeans. As a 2022 article in Canary Media explains, using arable land to grow biomass crops almost certainly results in land being cleared somewhere else to grow more food for human consumption, with the result that biofuels are sometimes even worse for the environment than fossil fuels. If that wasn’t enough, dedicated biomass can also increase use of water (for irrigation) and fertilized. The article also notes that, unfortunately, this type of biofuel consumption often has strong political support from the farming sector.

Problems In Practice

Some sources of renewable energy are comparatively tidy. For instance, rooftop solar panels, once installed, tend to just sit there producing power for years; there is not a lot of complication to the story. It’s true that some mess is involved in procuring the raw materials and producing the panels, but the overall impact over the lifetime of the installation is relatively small.

Biomass energy, on the other hand, is inherently complex and messy. Enormous amounts of material need to be collected, which often involves a certain amount of in-the-field processing (e.g. selecting which forest materials to harvest for lumber, which to collect as waste biomass, and which to leave behind). It then must be transported, often from remote locations, to a facility for further processing, often involving many steps – potentially including sorting and grading, grinding or other preparation, conversion to fuel form, combustion, CO₂ capture and storage, possibly with additional transportation steps along the way.

Biomass is often low-quality fuel, containing substantial amounts of water, soil, or other foreign materials, thus reducing its energy content, increasing transportation and processing costs, and leaving behind wastes that must be disposed of.

This comment from a recent Volts podcast episode on biofuels illustrates the challenges nicely:

We have enough woody biomass to produce all the fuels we need at a cost no one will accept (the tyranny of trucking range is real). So it gets stacked and burned, hauled to the dump and buried, or re-arranged in the forest and the inevitable future fire burns with high intensity despite the work meant to mitigate that intensity. You can't get the feed-stock to the factory, nor can you bring the factory to the feed-stock. It is a frustrating situation, and the inability to get sensible biofuels operating in easy to get to locations like farmlands make the wilderness and forestry story just so much more unlikely.

The upshot of all this is that biomass energy, especially BECCS, looks better on paper than in practice. Depending on the assumptions used, it may be anticipated to be reasonably priced, reliable, and carbon-neutral or -negative. As a result, when a computer model is used to compute an optimal path to net zero emissions, the model will often call for a heavy reliance on BECCS. However, these analyses often suffer from a number of flawed assumptions:

• Neglecting that the supply of cheap, safe-to-harvest biomass is limited.

• Failure to account for negative secondary effects of biomass use (particulate emissions from combustion, ecological impacts, displacing other uses of land which could ultimately lead to deforestation elsewhere, etc.).

• Assuming that BECCS costs will fall at the same rate as wind, solar, and battery power. As we saw last time in Which Technologies Have Learning Curves?, BECCS is the type of technology which typically sees slow price reductions3, while wind, solar, and batteries progress rapidly.

• Failing to take into account innovations that reduce the need for chemical fuels, such as alternative steel and cement manufacturing processes; or the potential to use cheap electricity to produce hydrogen or other fuels.

Biomass solutions also provide ample opportunities for greenwashing or outright fraud. For instance, there are documented cases of wood pellets produced from whole oak trees being passed off as “waste”.

Conclusions

The best way to speed the path to zero emissions is to keep as many options open as possible. Different use cases require different solutions, and adding biomass to the menu gives us more options.

That said, I get the strong sense that model scenarios tend to strongly over-favor biomass energy and BECCS. Models are based on quantitative analysis, and BECCS looks better on paper than it is likely to work out in practice, especially if the models ignore secondary effects such as particulate emissions.

I am especially suspicious of dedicated biomass. It takes enormous quantities of material to produce energy, so growing the necessary amount of crops (or trees) requires large amounts of land (often contributing to deforestation), and harvesting, transporting, and processing the material requires large amounts of energy. There may also be significant need for water and fertilizer. When combined with the secondary impacts of biomass use in general – particulate pollution and so forth – it seems very difficult to square the account for dedicated biomass. I’ll finish with a few more data points:

• Per a recent episode of the always-excellent Volts podcast, 300 acres of corn can generate enough ethanol to drive a car the same distance as one acre of solar panels powering an EV.

• The same episode also notes that devoting one acre of land to corn ethanol yields much less carbon benefit than devoting that same acre to forest.

• Another analysis claims that producing sustainable jet fuel from corn or soybeans will use 16 or 70 times as much land (respectively) as would be needed to produce the same amount of fuel from solar power.

• A 2011 publication from the RAND Corporation finds that “on an equivalent energy basis, most biomass in most regions is more expensive than coal”, and coal is already being outcompeted in energy markets, even without taking into account its negative externalities (many of which are shared by biomass).

You’re probably tired of hearing about the astonishing cost declines for clean solar, wind, and battery power. And that’s not all; for more and more products, the “green premium” is vanishing, as non-greenhouse-gas-emitting solutions become cheaper, or otherwise superior, to their polluting counterparts. This is probably the single largest reason for climate optimism.

However, not all products come down rapidly in price. For instance, the cost of nuclear power has actually increased in recent decades, at least in the US. What gives?

A recent paper, Accelerating Low-Carbon Innovation, helps explain which technologies are likely to experience rapid cost decline. This can help us understand which technologies are good bets for future success. It also shows that government actions, such as subsidies, won’t always be able to make a technology become successful.

The authors of this paper were recently interviewed on the (always recommended) Volts podcast, during which they shared a lot of additional context. In today’s post, I’m drawing on both the paper and the interview.

Some Technologies Progress Faster Than Others

The basic idea is simple: over time, as we manufacture more solar panels or nuclear power plants or whatever, we naturally find opportunities to reduce cost. A variety of factors influence the rate of progress:

Standardization. Solar cells are pretty much the same whether they’re on a rooftop in the US or a utility-scale installation in India, so an innovation developed anywhere in the vast solar cell manufacturing industry can quickly be applied everywhere. However, the design of a nuclear power plant varies according to size, site, and local regulations, so ideas discovered during one project may only be applicable to that project. To the extent that a product needs to be adapted to customer preferences, regulation, or the physical environment, that reduces potential for improvements to propagate from one market niche to another.

Product complexity. Simpler products, like solar panels, are easier to optimize. It’s also relatively easy to optimize a product that has many components if they are only loosely coupled, such as a laptop or smartphone, where updating one component doesn’t affect others – you can update the camera without having to redesign the battery. In a nuclear plant, by contrast, modifying one component will likely require adjusting the entire design. I imagine there are a couple of reasons that product complexity affects cost reductions:

1. Solar panels have only a few components, so if you find a way to improve one component, you’ve made a substantial contribution. A nuclear plant has many components, so an improvement to any one component has less impact.

2. The tight coupling of components in a nuclear reactor makes it hard to improve anything – you need to consider the impact on the entire system. From the interview: “complex products have a lot of nonlinear interactions between subcomponents”, making it harder to get anything done.

Manufacturing complexity. “A certain level of complexity in the manufacturing process is needed to see high learning rates.” I imagine the underlying principle here is that if the manufacturing process is simple, then you quickly run out of things to optimize. Lumber doesn’t get cheaper, because it’s already as cheap as it’s going to get. But in the case of solar cells, while the end product is simple, the manufacturing process is not, and there is room for optimization – for instance, by finding ways to slice the expensive silicon ingots more thinly.

Barriers to entry. If a product is very complex, and/or comes in large, expensive units (a single conventional nuclear plant costs billions of dollars), that creates a barrier to competition: it’s very hard for a new competitor to find the money to enter the market. This reduces competitive pressure to find cost savings.

Project timeline. Large, complex projects take a long time to design, implement, and deploy, so learning happens at a slower pace. In the case of a nuclear power plant, the time from the initial feasibility study through design, permitting, site preparation, construction, and startup can be ten years. That means it takes ten years to test a new idea. The time to manufacture a solar panel, or even to build an entirely new factory, is much shorter, so improvements can happen on a faster timeline.

Market growth. As costs for a given product come down, it becomes harder and harder to find the next improvement. If the market is growing, then the increased size of the industry drives both economies of scale and higher R&D budgets, allowing cost reductions to continue. This is why booming industries such as solar panels or computer chips continue to wring out efficiency gains, decade after decade after decade.

The authors also note that we don’t generally see learning curves for products which are based on digging stuff out of the ground, i.e. raw materials.

The Power of Standards

By identifying the factors that cause technologies to come down in price, the paper shines a light on ways to accelerate the process. The first approach I’ll discuss is standardization.

Governmental policy at the local, national, and international level can all influence the level of standardization. This doesn’t necessarily mean enforcing a specific standard, though it can – by the end of 2024, all mobile phones sold in the EU will have to use the same USB-C charging plug. That’s being done for the convenience of consumers, but such standards also reduce the need for manufacturers to create a different product for each market.

In many cases, rather than requiring standardization, governments actually forbid it. Variations in building and transportation codes force products to be customized to each market. By harmonizing standards across regions and countries, we can promote standardization without relying on slow-moving government agencies to actually mandate a specific standard.

Streamlining regulatory processes also helps, by reducing project timelines and thus speeding up the innovation cycle. And if a market lacks for competition, governments may be able to reduce barriers to entry by removing legal hurdles and subsidizing early investment.

If regulations change over time, that can be even more disruptive than differences between countries. One example is nuclear safety rules in the US, which have long been a moving target; this impedes cost optimization from one project to the next.

Of course, there are tradeoffs to all of this. If we enforce a standard too early, that can block innovation. For instance, EV manufacturers have been finding better and better ways to integrate battery packs into the structure of a car. If the government had imposed an early standard on battery pack design, those improvements wouldn’t have been possible. This is why it may be safer to focus on eliminating code variations, without taking the further step of mandating a specific standard. But sometimes that further step is called for.

In the early stage of large, complex technologies, where individual countries can’t afford to run their own demonstration projects, international cooperation becomes important. In these cases, we need shared projects where the learning can be widely disseminated. Of course this is in tension with the desire for proprietary advantage, which has been an issue, for example, in CCS (Carbon Capture and Storage) technology, and will also likely be an issue for advanced nuclear designs. For instance, France and South Korea have historically managed to achieve modest cost designs on nuclear power, through the use of standardized designs, but this benefit has not propagated to other countries. Quoting from the paper:

…[The] Paris Agreement’s reliance on national green industrial policies is suited for promoting innovation only for some technologies. For highly complex and/or customized technologies, such as nuclear power, biomass power, and CCS, national policies need to be augmented with top-down measures to coordinate research, development, demonstration, deployment, and regulations across countries to facilitate learning-by-doing and inter-project and inter-context spillovers at a regional or global scale.

Subsidies Only Work When There’s a Learning Curve

If a technology is amenable to a steep learning curve – if it is standardized, simple, has short project timelines, and so forth – then costs will come down almost automatically as a side effect of increased usage. This is where subsidies, purchasing requirements, and other incentives can have an impact. Famously, early subsidies for solar panels – in Germany, for instance – carried the industry past its early period of high costs and allowed solar power to become competitive without subsidies.

Conversely, technologies that are not amenable to learning curves may not be good candidates for subsidy. The cost of nuclear power has not been decreasing with each new plant, so it’s unlikely that subsidizing more plants would bring costs down.

Which Specific Technologies Will Do Well

The paper provides a nice chart to organize different technologies. “Type 1” technologies in the lower-left corner show the steepest cost declines; “type 3” technologies along the top and right sides benefit the least.

Note that BECCS and biomass are in the “Customized” column in part because the feedstock varies from place to place. Biomass varies from place to place – by type, moisture content, etc. – and so it’s hard to design a standardized, mass-produced processing facility.

Implications for Climate Change Mitigation

Ongoing price declines have the potential to accelerate the transition to clean energy sources. Current models from the IPCC, IEA, and other organizations often either fail to take learning curves into account, or assume that all technologies will improve at the same rate. In the real world, technologies with steep learning curves – such as batteries and solar panels – will play a larger role than the models show, and may help the transition move faster than expected. In particular, once these simple, standardized “type 1” technologies become economically viable without subsidies, adoption accelerates rapidly, as a virtuous cycle emerges: increased sales → lower cost → increased sales. (In hindsight, it’s remarkable that subsidies for solar power were sustained across the very long time period required for the technology to finally become cost competitive.)

By the same token, we should be wary when models envision widespread use of complex / customized Type 3 technologies, such as BECCS or nuclear power. If those technologies aren’t economical today, they’re going to have a hard time getting there in the future.

Across all technologies, we can accelerate progress by harmonizing regulations and removing other barriers to standardization. For Type 3 technologies, international cooperation on technology development, demonstration projects, and innovation sharing are critical. And we should not expect subsidies to drive cost reduction for nuclear power, BECCS, or other type 3 technologies.

Prior to human intervention, the Earth’s climate was relatively stable from century to century. On longer time scales, you’d find the occasional ice age and whatnot. But things didn’t change at the speed we see today. The world’s population has grown remarkably, as have our technological capabilities. But even so, it’s a big planet, and we are still fleas on the back of an elephant. How have we managed to actually heat up an object as large as the world? And can we reverse course?

We spend a lot of time in the details of climate change, its myriad causes and effects. Occasionally, for perspective, it’s worth reviewing the big picture.

When something is stable, like the pre-industrial climate, we tend to think of it as static: “lacking in movement, action, or change”. However, nothing about the atmosphere is ever lacking in movement. Instead, stability was due to the system having achieved balance. Over the eons, critical parameters such as temperature and gas mix had wandered until they reached an equilibrium, where motion in one direction was matched by motion in the opposite direction.

Why do we burn fossil fuels? For heat, which gives us both warmth and motion. But the heat of those fires is not what’s warming the Earth. It’s the side effects, such as carbon dioxide, that are tipping the balance of forces that control our climate. That’s good news, because it means that halting climate change doesn’t require giving up the things we want. There are many ways of obtaining warmth and motion without cooking the planet; solar, wind, nuclear, geothermal, hydro, and wave energy are all climate-safe. Digging stuff up and burning it is the only energy source that triggers warming. In the long run, preserving the climate will be easier than destroying it.

I have a big project underway: a concise, comprehensive guide to climate change mitigation. Today’s post, adapted from that project, is a review of the basic science of climate change and the greenhouse effect. I’m sure this is not new to you, but give it a go anyway. Rather than recapping the dry facts, I’ll try to put them in perspective, to give you an intuitive feel for why human activities are having such a large impact, and why we’ll be able to stop.

Exposed to Extremes

Every time you go outside, you’re standing under a frigid, yawning abyss that extends to the edge of the universe. The temperature in outer space is -270°C, or -455°F. It’s cold out there.

Nestled in the cozy embrace of Earth’s atmosphere, it’s easy to forget that. But without the atmosphere, temperatures would plunge to lunatic, science-fictional levels every night. Then the sun comes up, and a colossal nuclear furnace, almost a million miles across, blasts us with 173 quadrillion watts – as if the sky were filled with one hundred trillion space heaters, spaced just a few feet apart.

On the moon, right next door and exposed to the same conditions, temperatures swing from insane cold (-130°C / -208°F) to literally boiling hot (120°C / 250°F) with every sunrise and sunset. On Earth, the tenuous layer of atmosphere insulates and buffers us from these wild swings1.

In other words, the temperature we experience at the Earth’s surface is a delicate balance of energy flooding in from the sun, and energy leaking out to space.

Energy Flows

When you have a mix of temperatures, they tend to even out: heat flows from hot to cold. Our experience of this is based on heat conduction (between objects that are touching one another, as when you pick up a hot dish straight out of the microwave) or convection (riding along with a moving fluid, as when you open the door on a cold winter day and freezing air pours in). In space, nothing is touching and there’s no fluid, so these mechanisms don’t apply.

Instead, the energy flows between the Sun, Earth, and outer space consist of radiation – the same effect that makes you feel toasty when standing in front of a fire. This radiation is a ray that shoots instantly from one place to another, like a beam of light. In fact, exactly like a beam of light, because visible light is simply a form of radiation. The warmth that you feel when stepping into direct sunlight, and the warmth that you feel under a heat lamp, are both caused by radiation landing on your skin. In the case of the sun, it’s white light; from the heat lamp, it’s “infrared” light – basically, a color that’s even redder than what we call red. Our eyes don’t pick it up, but it’s the same sort of stuff as visible light.

The sun is glowing white-hot; that’s why it emits light. A heat lamp glows infrared-hot (edging into actual red-hot, which is why the the heating coils look red). The Earth also glows infrared-hot, though it’s farther down the spectrum than a heat lamp. The infrared glow from the Earth radiates into space, taking heat with it.

Because the Earth isn’t nearly as hot as a heat lamp, its infrared glow is weak. However, that radiation emanates from every inch of land and ocean, 24 hours a day, 365 days a year. This is enough to balance the blazing white solar radiation, which is much more intense at its source, but less so when spread across the Earth’s full surface. The sunlight absorbed by the Earth carries about the same energy as the infrared glow shining from the Earth into outer space. That’s not a coincidence: over time, temperatures adjusted until these flows were in equilibrium.

The Greenhouse Effect

One fact about radiation – at least, the “electromagnetic” kind we’re talking about – is that it’s easily blocked. That’s why a simple parasol, or a canopy of leaves, can shade you from the hot sun.

Glass, of course, does not block visible light. (Duh; that’s like the whole point of glass windows.) But it does block some infrared light. So if you build a big glass box and set it out in the sun – in other words, if you build a greenhouse – then the sun’s rays can shine in, but infrared from the warm interior can’t shine out. With energy flowing in but not out, the inside will get hot. This is also why cars get so hot when left out in the sun.

Some gases, such as carbon dioxide, have this same property of blocking infrared light. And so just as glass walls trap heat inside a greenhouse, “greenhouse gases” trap heat in the atmosphere; hence, the “greenhouse effect”.

Clouds block both visible and infrared light. The net effect might be to warm the Earth, or to cool it, depending on the cloud’s altitude. Chemicals which affect cloud formation, therefore, can also affect the balance of energy flow.

Over Time, A Small Imbalance Has a Big Effect

The atmosphere itself is not static. Natural processes such as photosynthesis, plant respiration, and exchanges of gas between the air and ocean move far more carbon than the best efforts of the fossil fuel industry. However, fossil fuels are a new factor, not part of the equilibrium, and they push in just one direction: up.

In a day, or a year, the amount we burn is not really significant. But we’ve kept burning, and burning, and over decades and centuries it’s added up. Even now, the amount of excess CO₂ in the atmosphere isn’t much on a planetary scale. Suppose that, for visualization purposes, we were to freeze that excess; the resulting layer of dry ice would be just 1.35 millimeters thick. That’s 1/20th of an inch. This gossamer shell, along with even more tenuous wisps of other greenhouse gases, blocks only a tiny fraction of the Earth’s infrared glow; tipping the energy balance only slightly.

On average, allowing for day/night and equator/pole variations, the amount of sunlight reaching the top of the atmosphere is 340 watts per square meter. The infrared radiation escaping from the Earth back into space is just slightly less, about 339.2. In other words, the flows of energy in and energy out are almost identical; if we compare the size of the flow to the land area of the United States, the imbalance is proportional to Hawaii.

This slight imbalance in energy flow, caused by the historical accumulation of a slight imbalance in flows of greenhouse gases, is entirely responsible for global climate change. Think of it as the ultimate heat pump: our total energy consumption is “only” about 20 terawatts, but the resulting imbalance in energy flow between the Earth and outer space is 500 terawatts2. Inadvertently, we stumbled onto the worst possible way of generating power. Fortunately, to stop, all we need to do is to adopt literally any other approach.

Pick any article about climate change; it will be awash in statistics. Gigatons of CO₂, dollars per megawatt-hour, market share of EVs. It’s easy to get lost in the numbers and lose any sense of perspective.

For today’s topic, I thought I’d look for ways to actually visualize one of the most basic facts about climate change. This carbon dioxide that’s causing so much trouble: how much of it is there, exactly?

(I'm going disregard other greenhouse gases, such as methane and nitrous oxide. They have an important impact on climate, but exist in the atmosphere in quantities much smaller than CO₂.)

In total, the amount of extra CO₂ in the atmosphere, compared to the start of the Industrial Revolution, is about 1100 billion metric tons1. If we were to spread that evenly across the Earth’s surface – which, in fact, is exactly what we have done – how thick a layer would it form?

It’s difficult to visualize a gas. So let’s imagine we were to freeze the carbon dioxide into solid form. You’ve probably seen frozen CO₂ before: that’s what “dry ice” is. Distributed across the Earth’s surface, the extra CO₂ would form a layer of dry ice about 1.35 millimeters thick, or just under one twentieth of an inch2.

One twentieth of an inch doesn’t seem like much, but that’s actually about the thickness of a blanket, if you squeeze it flat. (I tried to Google “how thick is a blanket”, and got nowhere. So I went into the living room, grabbed my favorite fleece from the couch, and folded it over four times to make 16 layers. If I squeeze the resulting sexdecablanket3 down, it’s a bit less than an inch thick. So 1/20th of an inch is in the ballpark for one layer.)

Doesn’t it seem a bit odd that just one blanket-thickness of CO₂, almost literally a “blanket of CO₂”, can cause so much trouble? But we know that wrapping a blanket around something warms it up4. That applies to large objects as much as small ones5. Fortunately, all that CO₂ isn’t a very good blanket, as it’s only warming the planet by a degree or two. If we’d spent the last few centuries spewing fleece into the air, we’d really be in trouble.

Now Let’s Do Per-Person

Suppose we divide the extra CO₂ in the atmosphere by the population of the Earth. We just passed the 8-billion-person milestone. 1100 Gt / 8 billion people = 137.5 metric tons of CO₂ per person. How can we visualize 137.5 metric tons?

That seems like it might be in the right range for your bigger sort of tree. Let’s see if we can figure out how big a tree would need to be to contain that much carbon. It’s an obscure sort of question, but Eliana managed to dig up a worksheet on the NOAA website that relates a tree’s diameter to the amount of carbon it contains. For a red oak, it works out that if the tree’s trunk is 292 inches in circumference, then the above-ground portion of the tree contains about 137.5 metric tons of CO₂6.

The largest red oak on record appears to have a 405 inch circumference, so 292 inches is entirely feasible. I tried to find a picture of a tree that size; the best I could come up with, courtesy of The Gladwin County Record & Beaverton Clarion, is this photo of a 218-incher:

Our 292 inch tree would be substantially larger than this, but you get the general idea. (For scale, note the red building in the lower-left corner, which appears to be slightly behind the tree and hence looks a bit smaller than it should.)

(If we could somehow plant just one red oak tree per person, give them all the space, water, and other resources they need to grow, and wait until they achieve this size, that would make a huge dent in global warming – assuming we’d managed to do all that without disturbing land that already contains significant amounts of carbon. Unfortunately, as we saw last time, that’s difficult.)

Emissions vs. Atmospheric CO₂

So far, we’ve been talking about 1100 Gt of CO₂: the amount by which atmospheric levels have increased in the last few hundred years. However, our total historical emissions are much larger, around 2000 Gt or so (sources seem to disagree). Quite a bit of CO₂ emissions are absorbed by the ocean (causing acidification and other problems) or, in various ways, on land.

If we plug 2000 Gt into the formula, we get a red oak tree with a circumference of 374 inches. I am very disappointed in the Internet for not providing me with a photo of a red oak tree this size, but they do exist.

Ongoing CO₂ emissions run at about 37 Gt per year, or 4.63 metric tons per person per year. That corresponds to a red oak with a circumference of just 72 inches; you might just be able to wrap your arms around it, if you’re tall and your bones are made of rubber. Using the park bench for scale, this one seems like it might be a big large but in the right ballpark:

Twinkies

Any fan of Ghostbusters is familiar with another yardstick for measuring alarming situations: the Twinkie. If a Twinkie had the same carbon content as one person’s share of excess atmospheric CO₂, how large would it be?

Based on the nutrition facts (hah) for a Twinkie, a pair weighs 77g: 43g carbohydrate, 8g fat, and 26g unaccounted for. Using a mathematical simplification technique known as “who cares, it’s a damn Twinkie”, I estimate that this includes 29.7g of carbon7, or 14.85g per Twinkie.

Sometimes the Internet really comes through: of course there’s a web page checking Egon’s math on the size of the Ghostbusters Twinkie. It states, and I am inclined to accept, that modern Twinkies are 9.9cm long.

So: there’s currently 137.5 metric tons per person of excess CO₂ in the atmosphere, or 37.5 metric tons of carbon, equivalent to the carbon content of 2,525,253 International Standard Twinkies. A single Twinkie containing that much carbon would be 136.2 times larger in each dimension, or just under 13.5 meters (44 feet) long. In Ghostbusters, Egon says “According to this morning's sample, it would be a Twinkie 35 feet long, weighing approximately 600 pounds.” Our Twinkie is only moderately longer than that; though Egon was way off on the weight.

Wait, Did I Have a Point Here?

Surprisingly, yes.

We tend to talk about climate change in terms that are too large or abstract to grasp intuitively. By translating CO₂ emissions into a tangible object, we can get some grasp of the scale of the problem. We saw that the amount of carbon we emit per person, per year, is equivalent to a medium-sized tree. That’s not small; you’re not going to dent it by purchasing a reusable drinking straw. But it’s not astronomical either. It’s within the scale of things that we can reasonably influence, especially at the scale of industrial policy. This shouldn’t really be surprising, as it was mostly industrial activity that created the problem in the first place. But it’s nice to confirm it.

Of course, the idea of forestation is appealing. “I think that I shall never see / A poem lovely as a tree.” Trees are natural, nice to look at, and provide many benefits beyond absorbing carbon dioxide. But precisely because forestation is so appealing, it’s easy to become over-optimistic about its potential to mitigate climate change.

Last time, in Can't see the Forest For the Tree Plantations, I explored the many factors that affect the climate impact of tree-planting projects. In this post, I’ll dig into how much of the climate problem we can solve this way. Briefly: we could pull down a lot of carbon by planting trees, and the potential side benefits are enormous. But to fully realize this potential would take more land than we’re likely to commit, and would require using it in ways we’re not on track for – fewer plantations, more natural forest. Done improperly, forestation can cause more harm than good.

The primary focus of climate change mitigation must always be to eliminate emissions: replace furnaces with heat pumps, coal plants with wind farms, and so forth. When it comes to trees, reforestation is great but protecting our existing forests (”conservation” or “proforestation”) is more urgent. Climate change calls for an all-of-the-above approach, so we should be pushing to plant more trees, but only alongside emissions reduction and forest preservation.

Come For The CO₂ Sequestration, Stay for the Biodiversity

Before we dive into the CO₂ numbers, remember that forestation has many effects beyond pulling carbon out of the atmosphere. We’re talking about redecorating a significant portion of the planet, so we should expect a few side effects.

It’s important that projects be handled wisely. Ideally, the goal should be to recreate a natural forest, using a mix of species native to the region and appropriate to the terrain, and in consultation with the people in the affected area. Otherwise, a variety of negative effects can occur, and the project is less likely to achieve its basic carbon goals: a poorly-placed forest may destroy important habitat or release soil carbon, an inappropriate forest may not thrive, and an unwanted forest is liable to be cut down.

The benefits can be profound: “soil erosion prevention, recreation, local evaporative cooling, and possibly increased precipitation”1; “water filtration, flood buffering, soil health, biodiversity habitat, and enhanced climate resilience”2. Forests affect the local water cycle, in some cases promoting increased rainfall which can in turn support the health of the forest. Thus, forest restoration creates a positive feedback loop (making it easier for forests to thrive), while deforestation can lead to “desertification”: a negative feedback loop where rainfall decreases, making it harder for the remaining patches of forest to survive.

The downsides primarily occur when attempting to impose a forest on land not well suited to it, or when using inappropriate species, especially monoculture tree plantations. Impacts can include displacement of native species and loss of biodiversity, increase in invasive species, and loss of pollinating insects3, among others. Tree-planting projects may also displace other uses of the land, leading to forests being cleared elsewhere: if you plant trees on land that’s used for cattle grazing, the rancher is liable to raze the next forest over.

In Principle, We Really Could Plant a Lot of Trees

(If you’re not interested in a detailed analysis, feel free to skip to the last paragraph of this section. This entire post is fairly technical; if that doesn’t sound fun, skip directly to the last section to see my conclusions.)

Here are a few sources that model a “maximum” scenario, where we use every acre of land that is suitable for tree cover and is not currently farmed. This is not a realistic prospect, but it puts the potential of forestation into perspective.

I’ll do my best to translate each study into the same units: billions of tons of CO₂ removed per year (Gt CO₂/y)4. We can directly compare these numbers against emissions sources; for instance, passenger cars emit about 3 Gt CO₂/y.

Some models adjust for albedo effects: forests are dark, and so absorb the sun’s heat, thus warming the planet. If a reforestation project takes in 100,000 tons of CO₂ per year, but absorbs enough sunlight to warm the planet as much as 30,000 tons of CO₂ per year would cause, then we say the adjusted effect is to remove 70,000 tons per year. I’ll indicate which models adjusted for this, and which did not. It’s mostly an issue in northern regions that are otherwise covered in (shiny!) snow.

For each model, I’ll report the amount of land proposed for tree planting, in millions of hectares (Mha)5.

• Limited climate change mitigation potential through forestation of the vast dryland regions (Science, 2022) projects removal of 5.21 Gt CO₂/y, adjusted for albedo, using 1804 Mha of land. On careful review, this is only for “drylands” (see previous discussion). Adding in tropical rainforest and other non-”dryland” regions would yield higher figures.

• The Global Tree Restoration Potential (Science, 2019) projects 205 billion tons of carbon in total, using 1700 to 1800 Mha. I interpreted this as taking place over 80 years (a reasonable time for forest to reach maturity), so 9.39 Gt CO₂/y, not adjusted for albedo effects.

• Mapping carbon accumulation potential from global natural forest regrowth (Nature, 2020) presents two figures – a “national commitments” scenario based on what countries have promised to do, and a “maximum” scenario which attempts to model the maximum reasonable amount of land use6. For the maximum scenario, they project 8.91 Gt CO₂/y over the next 30 years, using 678 Mha. This figure is for “natural forest regrowth” (which they define to mean something similar to reforestation), and does not include albedo effects, but they exclude the “boreal” (northern) regions where albedo effects are most pronounced.

So, we have two estimates for annual CO₂ drawdown based on aggressive forestation: 9.39 Gt not allowing for albedo effects, and 8.91 Gt sort-of allowing for albedo effects (but only including reforestation, not “afforestation” on land that was not historically forested). And then we have 5.21 Gt just for “drylands”, and allowing for albedo effects. Worldwide annual greenhouse gas emissions are the equivalent7 of around 50 Gt of CO₂, so removing 5 to 9 Gt would make quite a difference – roughly equivalent to all of the emissions from steel and cement manufacturing, aviation, and shipping8!

Cattle Ranchers Would Like a Word

It’s unrealistic to expect that we’ll actually plant – and maintain! – trees on every possible acre. Determining what we can expect comes down to two questions: what would all of this cost, and what are the competing uses for the land?

I had a hard time understanding the nature of the land under discussion. It’s all suitable for growing trees, but not currently in use for agriculture. What sort of land is that, where is it located, what does it look like, how is it used today? Mapping carbon accumulation potential from global natural forest regrowth notes that the maximum scenario “would require dietary shifts towards a plant-based diet, which could release large areas of current grazing lands back to forest, as well as croplands that are used to produce fodder for livestock”. So it sounds like much of the proposed land is currently used for grazing.

Here are some attempts to model more achievable scenarios, as opposed to the maximum scenarios reviewed above. (Feel free to skip this bullet list.)

• Mapping carbon accumulation potential gives 5.86 Gt CO₂/y, using 349 Mha of land, for their “national commitments” scenario9.

• Regenerate natural forests to store carbon (Nature, 2019), presents figures based on the Bonn Challenge10: working out to about about 1.91 Gt CO₂/y through 2100, using 350 Mha. I don’t think this accounts for albedo effects, but they do seem to exclude the northern regions where that would make a big difference. They also note that an (older) IPCC report models the potential of reforestation and afforestation at about 2 Gt CO₂/y through the end of the century11.

• Natural climate solutions (Proceedings of the National Academy of Sciences, 2017) states that if we limit ourselves to projects that cost less than $100 per ton of CO₂ removed, forestry can draw down 7.3 Gt CO₂/y. If we restrict to $10 / ton, this drops to 2.3 Gt CO₂/y. This is helpful, because it gives us some idea of how much all this might cost. (For perspective, industrial methods for removing CO₂ from the atmosphere are anticipated to cost something like $100 / ton.) These figures are not adjusted for albedo effects, but do exclude areas where the albedo effects would be larger than the CO₂ benefit. I’m not sure how much land would be used in these scenarios; the article mentions figures ranging from 345 to 1,779 Mha.

• Negative emissions—Part 2: Costs, potentials and side effects (Environmental Research Letters, 2018) states that “based on a systematic review of the literature, our best estimates for sustainable global NET potentials in 2050 are 0.5–3.6 Gt CO₂/y for afforestation and reforestation”, and mentions a wide range of land areas used, as high as 2580 Mha. They also note that “There is high agreement on the maximal costs” of afforestation and reforestation being around $100 per ton of CO₂.

• Project Drawdown projects that 3.2 to 4.96 Gt CO₂/y could be sequestered through 2050 using a combination of tropical forest restoration, temperate forest restoration, and tree plantations on degraded land, using a total of 366 – 533 Mha.

This is a wide range of figures, from 0.5 to 7.3 Gt CO₂/y. However, progress to date suggests that we might fall far short of even these “achievable” scenarios:

• Mapping carbon accumulation potential notes that this scenario “will be difficult to achieve, with some countries committing to the restoration of more forest area than is available and relying on approaches other than natural forest regrowth to restore forests.”

• Regenerate natural forests to store carbon notes that their figures would require all 350 Mha to be allowed to become natural forest; under “current plans”, which include plantations and other less-than-natural forms of tree planting, they project only 0.73 Gt CO₂/y.

• Regenerate natural forests to store carbon notes that just a third of Bonn Challenge pledges are for land to be returned to natural forest; the remainder is slated for tree plantations and agroforestry (mixing trees with other crops), which may sequester much less CO₂, depending on how the wood is used. Also, if such a large increase in commercial tree plantings actually came to pass, that would presumably depress the price of wood products, calling into question whether these plans would actually be sustained.

• The global tree restoration potential notes that 10% of countries in the Bonn Challenge “have committed to restoring an area of land that considerably exceeds the total area that is available for restoration”, another sign that these commitments may not be well thought out.

Might we do better? In climate action, the pattern has been for national plans and commitments to start out highly inadequate, but improve over time. Plans for deploying clean power, electrifying transportation, and so forth keep accelerating. Perhaps we can hope for forestation plans to similarly increase. However, I’m not terribly optimistic about this. In many areas, climate change mitigation benefits from both increased awareness of the urgency of the problem, and the falling cost of clean technologies. When it comes to forestation, the latter doesn’t seem to apply; there’s no technological advance that brings down the cost of “please stop grazing your cattle here” the way we’ve brought down the cost of solar panels and EV batteries. It’s gotten to the point where the profit motive often pushes utilities to shut down coal plants, but there’s not much profit motive for natural forest restoration12. So it’ll be a tougher climb.

Perhaps most telling is the fact that deforestation continues unabated in many parts of the world. How feasible are global forest restoration commitments?13 examines this, noting that “tropical forests are being lost at a rate of 15.8 million hectares a year”, a high rate. If we haven’t even managed to stop cutting forests down, how do we expect to undertake the difficult work of restoring them? The same source also notes that many “restoration” commitments may not reflect true reforestation; for instance, “… although the United States officially reported restoring nearly 17 million hectares, only 4% … consisted of newly planted forests, agroforestry, or natural regeneration, with the remainder in silvicultural treatments like thinning and prescribed burning that restore existing forests.”

Forestation: Part Of a Balanced Diet

At the end of the day, what can we realistically hope to accomplish by planting trees? I think we can take the Project Drawdown range of 3.2 to 4.96 Gt CO₂/y as an optimistic upper bound; the reality is likely to be less. Those figures would require devoting 366 to 533 Mha of land. For comparison, I estimated an upper bound of 122 Mha of new land needed for renewable energy in 2050, and Wikipedia says that 4,911 Mha is used for agriculture. So a serious forestation effort will use much more land than renewable energy, but much less than agriculture.

Clearly, trees alone can’t counterbalance worldwide emissions, which are around 50 Gt of CO₂e/y. Over the centuries, we’ve cut down forests and burned vast quantities of fossil fuel. Reversing deforestation alone won’t undo fossil fuel emissions. And the agricultural needs of 8 billion people mean that complete reforestation is not in the cards.

We must find ways to eliminate almost all of the things we do that emit greenhouse gases – including the ongoing destruction of natural forests. We should only look for counterbalances to those few emissions sources that we can’t eliminate; this is where forestation can play a role. We need to find ways to promote “good” forestation, without excessive reliance on plantations and other projects that are potentially harmful, or at least not as helpful as natural forest. Even one or two gigatons per year would be a huge contribution, so forestry is a critically important piece of the climate puzzle. It’s just not a silver bullet – nothing ever is.

The topic turns out to be even more complex than I’d expected. The carbon impact of “one tree” is like the health impact of “one food”: you can’t say anything intelligent without specifying what kind, in what quantity, under what circumstances.

As a buildup to our review of the potential climate benefits of forestation, I’m going to spend today just walking through some of the things that influence a tree’s impact on the climate.

There’s More To a Tree Than What You Can See

The basic idea is, of course, very straightforward:

1. Trees are big.

2. Big things have a lot of stuff in them.

3. For living things, a lot of that stuff is carbon.

4. So, trees contain a lot of carbon.

5. They get that carbon from the air (splitting CO₂ via photosynthesis).

6. So, if you plant a tree, it will pull a lot of CO₂ from the air and store the carbon in its body.

However, it turns out that tree trunks and branches are not the only places where forests store carbon. There are at least five important categories:

• “Aboveground plant biomass” – the tree trunk, branches, and leaves.

• “Belowground plant biomass” – right, can’t forget the roots.

• Dead wood. This may eventually decay, but until then, there’s carbon in it.

• Litter. I haven’t found a straightforward definition, but I think this is basically fallen leaves, twigs, partially decomposed wood, and anything else with carbon in it that’s above ground and not part of the previous categories1.

• Soil carbon: everything below ground, except for the living tree roots. This includes any sort of dead-and-buried material, which may come from trees, but also microbes and other organisms.

There are a couple of important implications. First, it’s really hard to measure the amount of carbon in a forest. Only one of the five categories shows up as a big obvious thing that you can hope to measure by just walking around and taking photos – let alone satellite photography. There are scientific models for analyzing and predicting the other four categories, but much work is still needed.

Second, a plot of land might already be storing a lot of carbon before you plant any trees on it. A forest contains a lot more obvious aboveground plant biomass than grassland, but by the time you’ve taken things like soil carbon into account, the results may surprise you. One source states, “North American grasslands can store as much carbon in soil as tropical forests store as biomass”. The soil carbon may have been accumulating for thousands of years!

The upshot is that planting a forest may not absorb as much carbon as you’d think. If the act of clearing, planting and managing the forest disturbs the existing soil carbon, it might not yield much benefit at all2.

Diamonds Last Forever, Trees Maybe Not

The Earth’s carbon circulates in many different forms, from CO₂ in the air, to carbon-based chemicals in living organisms, to carbon-bearing minerals like limestone; not to mention carbon in soil or oceans. Some of these repositories hold onto their carbon for millions of years, others have short lives.

Diamonds are pure carbon, and very stable under normal conditions, which is why you’ve never heard of a diamond rotting. There’s a company that is literally pulling carbon out of the air and using it to make diamonds. Unfortunately, this is not in any way a solution to climate change; it’s far too expensive to be anything more than a gimmick.

People debate how long a carbon storage technique needs to last in order to be a constructive part of the climate solution. Typical suggestions range from 100 to 1000 years. Personally, I lean toward the idea that approaches which last less than 100 years can still be useful. In any case, some degree of durability is necessary.

Individual trees can live for over 1000 years; I’ve stood inside one. And forests can last for a very long time. But only under the right conditions. Fire, drought, disease, logging, and clearing all can release a forest’s carbon back into the air.

Unfortunately, human activity has been exacerbating every one of these threats to forest longevity. Climate change increases drought; hot, dry weather promotes fire; heat- and drought-stressed trees are more vulnerable to disease; international trade helps spread pests; population growth and increased demand for meat encourage logging and clearing; and so forth. Trees have never come with a guarantee of permanence, and statistics based on historical forests may not apply to newly planted forests in the decades to come.

Not All Trees Are Planted Equal

Tree-based climate solutions can be divided into four categories, some of which have more benefit than others:

• Proforestation, aka conservation: the preservation of existing forests, e.g. “Save the Rainforests” campaigns

• Reforestation: planting trees in an area which was historically forested, but lost its trees to logging, fire, etc.

• Afforestation: planting trees in an area which was not historically forested.

• Densification: “fill-in” planting in an area which already has some tree cover.

Each of these has different characteristics. Proforestation is generally “best”: undisturbed forests generally contain trees in a robust variety of ages, sizes, and species; substantial amounts of carbon stored in dead wood, litter, and soil carbon; and other helpful qualities. In other words, they store a lot of carbon, and they have the best chance of holding onto it.

Reforestation can hope to recreate this state, but it takes many decades (perhaps centuries) of careful management to get there. So preserving existing undisturbed forest is most important, but reforestation – with appropriate tree species – can also be very constructive.

Afforestation can be quite problematic: if an area wasn’t historically forested, there's probably a reason. The land may be vulnerable to fire, disease, predation from plant-eating insects or animals, etc. Densification also may push a forest unsustainably out of balance3. Some analyses focus on carbon capture potential per tree planted, ignoring the fact that the ecosystem may not be able to support as many trees as were planted4.

A related term, “improved forest management”, covers a range of practices such as restoration of degraded forests, “fuels treatment” to address over-managed forests now at risk of catastrophic burns, better scheduling of harvests in working forestlands, and others. This doesn’t fit neatly into the previous categories, but helps illustrate how many dimensions there are to forestry.

The upshot is that planting an acre of trees is helpful only to the extent that those trees will survive and thrive in the long run. And the historical state of the land provides a strong hint as to how that will turn out.

Does a Tree Count if You’re Just Going To Cut It Down?

Not all tree-planting projects are even attempting to create natural forests. A common alternative is the “tree plantation”. Per Wikipedia:

A tree plantation, forest plantation, plantation forest, timber plantation or tree farm is a forest planted for high volume production of wood, usually by planting one type of tree as a monoculture forest.

In other words, rather than attempting to reproduce the complex diversity of natural forests, a plantation treats trees as crops, where a monoculture of identical trees will be periodically harvested and replanted. According to a paper in Nature, of acres pledged for forestation under the Bonn Challenge pledges (a major international forestation initiative), 45% are for plantations. The paper goes on to say:

…plantations hold little more carbon, on average, than the land cleared to plant them. Clearance releases carbon, followed by rapid uptake by fast-growing trees such as Eucalyptus and Acacia (at up to 5 tonnes of carbon per hectare per year). But after such trees are harvested and the land is cleared for replanting — typically once a decade — the carbon is released again by the decomposition of plantation waste and products (mostly paper and woodchip boards).

And later:

…we find, on average, that natural forests are 6 times better than agroforestry and 40 times better than plantations at storing carbon.

On the other hand, from Project Drawdown:

Degraded lands present potential locations for tree plantations. Managed well, they can restore soil, sequester carbon, and produce wood resources in a more sustainable way.

Or Wikipedia:

Because tree farms are managed to enhance rapid growth, a tree farm tends to sequester carbon more quickly than an unmanaged forest, considering only the sequestration side of the equation and not the carbon release due to rot, fire, or harvest. The fact that managed woodlands tend to be younger and younger trees grow faster and die less contributes to this distinction.

The point being, there’s no obvious consensus regarding the carbon impact of tree plantations. I’m not sure why there are such a wide range of takes. It may simply be that the impact depends heavily on where the plantation is sited and how the wood products are used, there may be some bad analysis floating around (I certainly wouldn’t want to rely on Wikipedia for this sort of thing), or the science may still be unsettled. In any case, it’s clear that broad claims of the global potential for tree plantations need to be treated with caution.

(Barbara Haya notes that plantations can have indirect benefits, such as replacing the need to harvest old-growth forests, and even displacing carbon-intensive concrete and steel as building materials.)

Shocking Conclusion: Forests Are Complicated

The carbon impact from planting a tree can be positive, neutral, or even harmful. It depends enormously on the sort of environment it’s planted in, how the land will be managed, the previous condition of the land, and other factors.

(As I mentioned last time, “albedo effects” can further complicate the climate story: planting dark-leaved trees on light-colored land can absorb more heat, overwhelming any carbon benefit.)

As a result, we can’t think about climate impact in broad terms of “billions of trees planted”. The details are hugely important.

Next time, I’ll hopefully be ready to actually present what we’ve learned about the potential of forestation to help combat climate change. Now that we all understand some of the important factors that influence the climate impact of planting trees, we’ll be ready to understand how that impact can be estimated at a global level – and the limits of such estimates.

Thanks to Barbara Haya of the Berkeley Carbon Trading Project for feedback on a draft of this post. All errors are of course my own.

We’re working on an analysis of forestation – i.e., planting trees – as a solution to climate change. This of course entails reading a bunch of papers. Making sense of what they say has been even harder than I’d expected.

We’re specifically looking for projections of how quickly new forests can absorb CO₂. One paper gave numbers that seemed to work out to 6.61 GtC/year1. Another paper reported 0.12 GtC/year. The difference between 6.61 and 0.12 is equivalent to the difference between the population of the United States, and the population of Colorado. That’s a pretty big discrepancy!

Climate science is tricky, but it’s usually not “off by a factor of 55” tricky. It turns out that most of the discrepancy came down to subtle reading errors on our part, along with at least one important omission in one of the papers. Either we’re dumb, or reading these papers is hard. I respectfully argue for the latter.

What you are about to read is my third attempt to explain how we came out with a 55x error in our reading of these two papers. I had to throw out my first draft and start over, because I realized I was looking at the situation incorrectly. Then I had to throw out my second draft for the same reason. In Never Trust a Number and Never Trust a Projection, I’ve talked about how how easy it is for numbers to mislead. Boy did that play out here!

Note: there’s a lot of technical detail in this post. Don’t worry about following the details; if you just skim along, you’ll still pick up the important points. In fact, feel free to skip over all of the numbers if you aren’t interested in that level of detail.

How Much Wood Would a Wood Planter Plant?

Let’s begin by briefly summarizing each paper, and the figures they give for carbon absorption by forests.

The first paper, published in Science in 2019, is The Global Tree Restoration Potential. It states that forestation can store 205 GtC by 2050, which works out to 6.61 GtC / year.

The second paper, Limited climate change mitigation potential through forestation of the vast dryland regions, was published in 2022. It states that between 2020 and 2100, trees could provide cooling “equivalent to the sequestration of only 9.7 Gt C”. 9.7 GtC over 80 years is 0.12 GtC / year.

As best I can understand, both papers use a similar approach. They use satellite images to create a detailed map of growing conditions – topography, soil, and climate – for the entire Earth. The map has a resolution of roughly half a mile2, which is a pretty good level of detail when we’re considering the entire planet.

They then feed in known data about existing forests: in one case, “78,774 direct photo-interpretation measurements of tree cover”. Using this, they train a machine learning model to extrapolate the potential for tree cover on each half-mile square of the Earth. By adding up the numbers for all land that is not already in use (for agriculture or other purposes), they get an estimate for the global potential of tree planting.

In what follows, I’ll refer to the two papers by their first listed authors: Bastin for the first paper, and Rohatyn for the second. This is shorthand for “Bastin et al” and “Rohatyn et al”.

We Thought 2050, We Thought Wrong

I mentioned that the Bastin paper models storage of 205 Gt carbon by 2050. It turns out that we hallucinated the 2050 part.

The paper doesn’t actually say how long new forests take to grow. It does talk about the year 2050 in other contexts3. But it never connects 2050 with tree growth.

Because no time frame other than 2050 is mentioned in the paper, we assumed that 2050 was also the time period for which they were modeling forest growth. But after reading a bunch of other sources (including the Rohatyn paper), it’s clear that a reasonable time for a forest to reach maturity is more like 80 years.

In our defense, the first paragraph of the paper mentions both “205 gigatonnes of carbon” and “by 2050”, and makes no effort to clarify that they are not connected. Academic writing frequently omits important context, presumably on the assumption that you’re already deeply familiar with the subject matter. If not, it’s easy to go astray.

If we divide Bastin’s 205 Gt by 80 years, we get 2.56 GtC / year. That’s still a lot more than Rohatyn’s 0.12 GtC / year, but at least we’re getting closer.

Bastin Didn’t Model Forests Correctly

The issue of Science in which the Bastin paper appears, also contains some comments pointing out several important flaws in the paper.

First, the paper miscalculates carbon sequestered per acre. It assumes that if a mature forest contains X tons of carbon, then planting a similar forest will absorb another X tons. However, the areas to be planted – mostly “sparse vegetation, grasslands, and degraded bare soils” – already contain quite a bit of carbon (much of it below ground). If a certain patch contains 200 tons, and eventually grows into a mature forest containing 500 tons, then forestation has only pulled down 300 tons of new carbon. According to two of the comments4, the Bastin paper fails to take this into account, and as a result gives estimates that are about double what they should be.

Second, it ignores “albedo effects”. Leaves are relatively dark, so planting trees on light-colored land has an effect similar to wearing dark clothing on a sunny day: the trees absorb sunlight and get warm. This heating effect reduces the climate benefit; in many areas, additional tree cover would be outright detrimental.

Third, they earmark natural grasslands and savannas for forestation. One comment notes that “herbivores and browsers—including insects—are ecological engineers stopping huge areas of natural grasslands and savannas from converting to forest”; another points out that in these areas, “fires and large herbivores have maintained low tree cover for millions of years”. Basically, there's a reason these areas, when pristine, weren't forests, and so attempts to turn them into forests are likely to fail.

At this point, I was developing the impression that the Bastin paper was written by a team that mostly understood machine learning, and didn’t know how to properly apply their results to climate change. The comments suggest that the estimate of carbon absorption is too high by factors ranging from 2x to 5x. If we compromise on 3x, then 205GtC becomes 68GtC; over 80 years, that’s 0.85 GtC/year, vs. 0.12 GtC/year for Rohatyn et al. Getting closer!

We Misunderstood Rohatyn’s Model

The next problem lies in our reading of the Rohatyn paper’s abstract. It presents a number which we thought applied to the entire Earth, but it turns out we were way off. Here’s the relevant section:

Forestation of the vast global drylands has been considered a promising climate change mitigation strategy. However, its actual climatic benefits are uncertain because the forests’ reduced albedo can produce large warming effects. Using high-resolution spatial analysis of global drylands, we found 448 million hectares suitable for afforestation. This area’s carbon sequestration potential until 2100 is 32.3 billion tons of carbon (Gt C), but 22.6 Gt C of that is required to balance albedo effects.

The last sentence is fairly clear in projecting a net climate benefit of 9.7 Gt of carbon over 80 years5. And the first sentence refers to “forestation”, a broad term which basically refers to any planting of new trees. However, the last two sentences switch to discussing afforestation, defined as “the act or process of establishing a forest especially on land not previously forested”. In other words, “afforestation” excludes the restoration of historical forest land (reforestation) or the planting of additional trees in land that already includes some tree cover (densification). Initially, none of us picked up on this subtle shift from “forestation” to “afforestation”; I finally noticed it on a fourth reading!

The other issue is that they are only modeling afforestation on “drylands”, defined as “an aridity index of less than 0.65”. Basically, this seems to exclude rain forests and other humid areas. So, 9.7 Gt is not their model of global carbon drawdown potential, it is only their model for afforestation in dry areas. They do go on to present a projection for all possible forestation activity across the entire globe, and it is a much higher number: 113.6 GtC.

Using 113.6 GtC for the Rohatyn paper, and dividing by 80 years, we get a carbon drawdown rate of 1.42 GtC/year, vs. our updated 0.85 GtC/year for Bastin. Our corrections have pushed the numbers so far that they’ve actually switched places; the Rohatyn figure now looks higher.

Bastin Strikes Back

Remember when I said that Bastin et al didn’t model forests correctly? Turns out that might be wrong.

Recall that the criticisms of the paper were as follows:

1. Giving forests credit for carbon that was present before the forest was planted.

2. Ignoring the impact of herbivores and fire.

3. Ignoring albedo effects (forests are dark, and so absorb the sun’s heat).

It turns out that Bastin et al posted a detailed response, which directly rebuts the first two points. They go into detail as to how they computed carbon per forest acre. It was a sophisticated analysis, and does take those first two factors into account.

My idea that Bastin’s team were experts in machine learning rather than climate science was way off base. If I had bothered to look as far as far as the footnotes on their first page, I would have seen that Bastin himself is at the “Department of Environmental Systems Science, Institute of Integrative Biology, ETH-Zürich”; he has a degree in tropical forestry.

They do acknowledge that they ignore albedo effects, stating that this is “beyond the scope of the present study”. Another example of academic writing omitting important context.

The details of the response are interesting. Bastin et al call out some errors that their critics made in interpreting their own paper – showing that even academics in the field can misread one another’s work. I’ve included some snippets in an appendix at the end of this post.

I don’t know how to update our math based on this. Earlier, I downgraded Bastin’s figure by 3x based on the three criticisms above. Accepting the rebuttal of the first two, but noting that the third criticism stands, I’m going to wave my hands and change the 3x to 1.5x. That gets us to a Bastin figure of 1.7 GtC/year, vs. 1.42 GtC/year for Rohatyn. Pretty close!

Boy, I’m Glad That’s Sorted

After many rounds of review, we’ve refined our interpretation of the two papers until they yield estimates that are only 20% apart (1.7 vs. 1.42 GtC/year). Such close agreement must mean that we’ve gotten to the actual truth, right?

Hah! Don’t count on it. Depending on how you count, we’ve identified up to eight distinct errors in our initial analysis6. If there were eight, there were probably more than eight. It took four readings of Rohatyn to figure out that the headline number only applied to afforestation; who knows what I’d find on a fifth reading?

Eight corrections in, the figures more or less agree, but that could easily be coincidence. If they had happened to agree after five or six corrections, I’d probably have stopped there, meaning I might have missed two or three errors. As the saying goes, “if you torture the data long enough, it will confess to anything”.

It’s also worth remembering that much of the fundamental science here is not settled. For instance, there is a lot that’s still not understood about underground carbon stores. Both papers are rough attempts to model a complicated planet, and they could both be off in the same direction.

At the end of the day, the fact that we’ve found two numbers that more or less match doesn’t mean that we have truth, it just means we have two numbers that match.

Let’s Be Careful Out There

On a first reading of the Bastin paper, we determined that forestation can remove about 6.61 GtC / year. On a first reading of Rohatyn, we got 0.12 GtC / year. After a lot more work, we’re converging on a number that is nowhere near either of those initial figures.

In other words: imagine that you have a scientific question in mind. You find a paper published in a prestigious journal. Your question is answered in the first paragraph, but you want to make sure you’re understanding it correctly, so you take the time to read the entire paper. Even so, there’s quite a good chance that the conclusion you’re coming away with is way off. On these two papers, in our attempt to read them correctly, we went zero for two.

One big problem is that these papers omit context. Bastin didn’t mention their time frame, nor that they were ignoring albedo effects; Rohatyn buried the fact that their headline figures only covered dryland afforestation.

The various comments and rebuttals were helpful, but you have to dig around to find them – they’re not in the PDF.

How do you guard against these problems? The best advice I can give:

1. Do your best to find multiple sources. If they disagree, dig in to understand why.

2. Maintain an attitude of humility. Your first reading may be incorrect. Your second reading may be differently incorrect.

3. Never Trust a Number!

   Thanks for reading! Subscribe to get another post every week or two (no fuss, no spam). You’ll make a blogger happy, but perhaps that’s a price you’re willing to pay.

Appendix: Snippets from Bastin Et Al’s Response To Critics

As promised earlier, here are some excerpts from Response to Comments on “The global tree restoration potential” (Bastin et al’s response to the comments on their paper). I’m not going to provide full context or explanation, but by skimming these excerpts, you should get some idea of the nature of the (extensive!) disagreements and misunderstandings.

The discrepancies between our estimate and their estimates arise from (i) misinterpretations or confusion between the definitions of forest cover and associated carbon pools, and (ii) a lack of sufficient detail in the original manuscript on how existing carbon in potential restoration areas was removed for estimating the global restoration potential. We clarify these points here.

Three of the four examples provided are based on a different definition of forest: namely forest area, rather than tree canopy cover.

We estimated that there is 1657 Mha of forest area available [table S2 of (1)], which contains 900 Mha available as cumulative tree canopy cover. Because these papers (7, 10, 11) were addressing forest area, the carbon density estimates would need to be scaled to 1657 Mha instead of 900 Mha. Correcting for this consideration of forest area almost doubles the carbon estimates proposed by Lewis et al.

The numbers provided by Lewis et al. in their restoration study (7) concern only two of the five carbon pools for vegetation ecosystems (i.e., aboveground and belowground plant biomass). Restoring forest ecosystems would actually have an impact on all five pools of carbon, including soil, litter, and dead wood (12). In our analysis, we included all five, which drastically increases the amount of carbon expected to be stored in restored forests.

After subtraction of the existing carbon content from the potential global carbon content that could be stored in areas available for restoration, the global carbon gain from tree restoration potential ranges between 133.2 and 276.2 GtC with a mid-range value of 204.7 GtC. This range reflects the uncertainty in calibrating the biome-specific carbon density values to a baseline percent tree cover (see Table 1).

We completely agree that changes in forest cover resulting from restoration would also affect the climate through a range of mechanisms including changes in surface albedo and evapotranspiration. … Calculating the changes in albedo and evapotranspiration associated with restoration is beyond the scope of the present study.

Veldman et al. stress that our model had low predictive power across many of the open-canopy biomes, suggesting that it fails to account for natural fire and the presence of large mammals. Here, they have misinterpreted the uncertainty of our model. First, natural fires and large mammals exist in protected areas. They are therefore indirectly accounted for in our model. Second, as natural fire cannot be distinguished from human-made fire, it cannot be accounted for as a variable of the model to extrapolate the natural tree cover outside protected areas. Third, the high uncertainty in intermediate tree cover is due to the general low occurrence of intermediate tree cover.

If land use is a real problem, it won’t show up for a while. To do anything about it, we’d need to see the problem coming well in advance. For that reason, I’m interested in identifying these sorts of scaling limits. Just because solar and wind power installations are skyrocketing today, doesn’t necessarily mean they can take us as far as we need to go.

Land Usage Figures Are… All Over the Map?

The amount of land needed to produce a given amount of electricity depends on how you’re producing it. Check out this chart:

You can gloss over the details. Each red dot indicates a mid-range estimate of the land usage for the corresponding energy source. Note that the numbers for hydropower and nuclear power are 33 and 0.3, respectively: hydropower uses 100 times more land to produce the same amount of electricity. (This despite the fact that the nuclear figure includes “land used for the mining of materials used for its construction, fuel inputs, decommissioning, and the handling of waste”.) This doesn’t mean hydropower is bad; it just means that land requirements will very much depend on our mix of electricity sources.

Renewables Don’t Use Up All The Land They Use

The amount of land used for renewable power depends on your definition of “used”. A certain amount of solar can go on rooftops. Wind power is very dense if you include only the space used by the turbines (“footprint”), but much less dense if you consider the necessary distance between turbines (“spacing”). Fortunately, the space between turbines can often be used for other purposes, such as farming. Solar power can also coexist with agriculture (“agrivoltaics”). Wind turbines can be located offshore, and solar panels can be placed on top of – and provide useful shade for – parking lots, canals, and other surfaces. So depending on the situation and how you look at it, there are three ways to measure the space used for renewable power generation:

• Spacing – all of the land involved, including the space between turbines, solar panels, or other units.

• Footprint – only the area physically occupied by the generation equipment.

• Dedicated1 – footprint, but excluding rooftop solar, solar panels floating on a canal, or other deployments that don’t take away any area from other uses of the land.

When reviewing published statistics for land usage, it’s critical to understand which measurement they’re using. This is touched on at the bottom of the previous chart, which shows that for wind power, the difference between spacing and footprint can be 250:1!

If you want to think about the potential capacity for a given source of energy, you need to use the spacing figures. The wind turbines we build in Nebraska can’t have a total spacing area that’s larger than Nebraska2. But if you want to know how much land will be “used up” for renewables, the “footprint” or “dedicated” figures are more relevant. (This is primarily an issue for wind power. As a result, the exact relevance of spacing vs. footprint will depend on our ability to have wind turbines co-existing with agriculture.)

Doing the Math

Let’s estimate an upper bound on total land requirements for clean electricity.

We found estimates of land usage intensity from three sources: Land-use intensity of electricity production and tomorrow’s energy landscape, Energy Sprawl Is the Largest Driver of Land Use Change in United States, and Decarbonization Won’t Require As Much Land As You Think. Using these figures, we can estimate the amount of land that will be required for electricity production in a net-zero future.

As should be familiar by now to readers of this blog, the sources don’t agree. For the majority of energy types, estimates differ by a factor of 2 or more. One source has the ratio of land usage for solar power vs. geothermal as 3:1; another has 40:1! Our detailed research notes, including some additional sources, are here. At the beginning of that link, you’ll find a small table summarizing the land use figures from each source.

In Energy Consumption in 2050, we estimated an upper bound for global electricity production in 2050 of 300 quadrillion BTUs. That’s about 88,000 TWh.

Setting aside biomass, and using “footprint” figures for wind power, the highest land usage figure we found for any form of energy in any of our sources was 23.5 m²/MWh/y, or 2350 Ha/TWh/y. To produce 88,000 TWh would thus require 207 million hectares. That’s assuming we use only hydropower (one of the least-dense sources), and using the most pessimistic estimate for hydropower land requirements across all three sources. So actual land usage should be much less than 207 million hectares.

As a cross-check, one of our sources provides this diagram, showing land usage for electricity production across a variety of scenarios:

This is a really nice information source, because it shows multiple scenarios and provides error bars. Looking at the scenarios for 2050, and excluding “Renewables (Spacing)” (which, again, is mostly the space between wind turbines), the numbers are generally in the range 100 to 150 Mha – well below our upper-bound estimate of 207. Even when including spacing, estimates range from around 170 to 280.

The first bar (“Current”) shows that today, around 85 Mha is already devoted to electricity production. So our conservative upper-bound estimate of 207 Mha in a net-zero 2050 scenario represents an increase in land usage of around 122 million hectares3 from today.

That’s a lot, but not A LOT a lot

122 million hectares is about twice the size of Texas. So we’re talking about a lot of land. But relative to the Earth, Texas isn’t really all that big. The additional land requirement between now and 2050 is only about 0.9% of the world’s ice-free land area.

(Remember also that this is based on our generous upper-bound estimate of 300 quadrillion BTUs worth of electricity, and taking highest land usage estimate for the least-space-efficient renewable source, hydropower.)

What about spacing for wind turbines? Even if we unrealistically assume that all electricity were to come from wind power, specifically on-shore wind power, and use the highest “spacing” estimate, that works out to 1119 million hectares, or 8.5% of the world’s ice-free land area. That’s less than 1/4 of the land currently used for agriculture (and remember that wind power can often co-exist with agriculture).

Land Use Will Be a Policy Challenge, Not a Physical Challenge

When the outcome is a clean, renewable solution for the entire planet’s electricity needs, it doesn’t seem unreasonable to use an additional 0.9% of the world’s ice-free land area. That’s about 1/40th of what we currently devote to agriculture. (And if we want to reduce overall land use for human activities, we’d do much better to focus on agriculture rather than energy production.)

It’s worth noting that the land requirements for renewables are static, whereas fossil fuels require more land each year: as wells dry up and coal mines empty out, we keep needing to find more. For instance, Germany periodically razes villages to make room for coal mines. In fact, coal power uses almost 2/3 as much land as solar4, even when disregarding rooftop solar. You don’t see much discussion of the land requirements for coal mining. The requirements for renewables won’t be so very much greater.

There will be huge challenges in deploying renewables at this scale, but they won’t take the form of “not enough land”. The challenges will be finding the right kinds of land in the right places, getting permission to use them, connecting them to the grid, and agreeing on who should pay for everything. In other words, policy challenges.

So don’t let anyone tell you that we won’t have enough land for renewable power. Focus on finding policy solutions that will get enough folks on board to allow us to build it.

However, as the saying goes: “Don't be so open-minded that your brains fall out”. Sometimes there’s a reasonable-sounding idea, taken seriously by reasonable people, which turns out to not make sense.

For instance: fossil fuels are mostly decomposed plants. At some point, it occurred to someone that we could emulate nature and process fresh crops to create fuel. The result is a renewable fuel source – every year, we can plant another crop. Best of all, this doesn’t contribute to warming, because plants get their carbon straight from the atmosphere. Great idea, right?

Unfortunately not. I am here to say that growing crops to produce fuel for electricity is a ridiculous idea, and we should simply drop it. People talk about it as part of the climate solution, and they just shouldn’t. This emperor has no clothes.

(The analysis here is specific to electricity. I am also skeptical of growing crops for other energy purposes, but that will have to wait for another post.)

NOTE: as often on this blog, I’m summarizing a complex topic and attempting to give a straightforward conclusion. If I say something that seems oversimplified or straight-up wrong, let me know!

An Extremely Brief Introduction to Biofuels

People have been using plants for fuel since the discovery of fire. The technical term is “biomass”. This is just a fancy word for using wood, manure, or other plant material or animal waste to produce energy. Traditionally by burning it, of course, but the raw material can also be processed into a more concentrated fuel form, such as ethanol, before burning it.

Some biomass comes from waste materials (agricultural plant waste, lumber scraps, municipal waste, manure, etc.) Sometimes, crops are grown specifically with the intent of producing biomass for fuel. This is known as “dedicated biomass”.

Location, Location, Location

To power the move to renewable energy, we’re going to need a lot of land. How much land? We set out to investigate that, and I’ll be diving into the results in the next post. The figures are, if you’ll forgive me, all over the map:

For now, you just need to know that each box represents a source of electricity. The higher the box, the more land that source uses. You can see that the last box stands out above all the others: “BioDed” – dedicated biomass.

If you look carefully, you might notice that the vertical axis uses a log scale. You’d then be able to work out that, by these figures, dedicated biomass uses almost ten times as much land as the next source down.

However, that’s understating the case. That next source down, “Wind+”, represents land requirements for wind farms including all the empty space between the turbines. That space can often be used for other purposes (e.g. farming). If you consider just the space actually occupied by the wind turbines and their access roads, you get “Wind-”, way over on the lower left of the chart.

We pulled land usage figures from three different publications1. For each energy source, we pulled out the figure for LUIE (land-use intensity of energy), i.e. the amount of land required to produce a given amount of electricity. The most optimistic figure we found for dedicated biomass was 580 m²/MWh/y. The most pessimistic estimate for any other energy source2 was 23.5. So, using the figures that put dedicated biomass in the most favorable light, we find that it requires 25x more land than any other source of energy.

25 Times Worse, That’s Bad, Right?

Let’s run some numbers. According to Google, the world’s total arable land is about 1.38 billion hectares. Using the biomass LUIE figure of 580 m²/MWh/y, we get a total potential electricity production of 23,793 TWh / year. Based on the figures we came up with last time for electricity usage in 2050, this would probably meet 1/4 to 1/2 of the world’s requirements. In other words, if we devote the entire arable land area of the planet for biomass – no food crops, no forests, heck not even any cities in arable land – we still wouldn’t come anywhere close to meeting our needs for electricity alone.

As the mathematician Richard Courant commented, after viewing a prototype of a 1950s rocket design based on throwing little nuclear bombs out the back to push the rocket forward:

Zis is not nuts, zis is super-nuts.

But Wait, It Gets Much Worse

In practice, biomass crops will typically entail large monoculture farms (e.g. corn), requiring fertilizer, irrigation, etc.

The large land usage has huge impacts on ecosystem health, biodiversity, deforestation, etc.

The climate benefit is often negligible. For instance, in the wake of the 1970s energy crisis, the US began planting corn for conversion to ethanol (which is then mixed into the gasoline supply). This has long been notorious for its various harmful impacts. From a Reuters article last year:

The research, which was funded in part by the National Wildlife Federation and U.S. Department of Energy, found that ethanol is likely at least 24% more carbon-intensive than gasoline due to emissions resulting from land use changes to grow corn, along with processing and combustion.

“24% more carbon-intensive than gasoline” is not a ringing endorsement for a supposedly renewable energy source. The US really ought to stop doing this.

I Just Don’t See Two Sides to This Question?

Based on the data we found, I’m here to say that growing crops explicitly for use as biomass for electricity production is a ridiculous idea, absolutely inane, and it’s a waste of time to even talk about it. I usually try to talk about what I’m in favor of, rather than what I’m against; there’s plenty of room to try multiple approaches. But dedicated biomass strikes me as a “solution” with very little room for benefit of the doubt, and it’s absorbing substantial resources right now.

It’s true that biofuels, unlike (say) wind or solar power, can be easily stored. However, there are a lot of other ways to address the storage problem, that don’t have nearly as much environmental impact as dedicated biomass3.

So if anyone ever tries to talk to you about dedicated biomass as a source of electricity, please point them at this post – here’s the link, it’s climateer.substack.com/p/biomass – and ask them to either refute the math, or drop the subject and never bring it up ever again.

If you think I’m looking at this the wrong way, let me know in the comments! I promise to take feedback seriously; if I’m wrong about this, I really would love to know it, and I promise to update the post. It follows that if you’re reading this paragraph as it currently stands, then no one has provided me with an even vaguely plausible argument that devoting land specifically for the production of biofuels is anything other than a straight-up bad idea.

FWIW, Bill McKibben seems to agree.

In this post, I’ll explore what we found when exploring the literature… and challenges we encountered along the way.

We specifically looked for projections for the year 2050, and when possible, scenarios that involve achieving net-zero emissions in that year1. Why 2050, and why net zero? Well, my overarching goal is to help optimize the path to a net-zero planet. To do that, we should think about scenarios that are aspirational, but not impossible. (We’re trying to shape the future, not passively predict it.) That seems to point to net zero somewhere around 2050, and it’s a nice round number.

It’s Hard To Find A Straight Answer

The Internet is, of course, awash in reports, studies, white papers, explainers, and the like. All of which are full of numbers. So many numbers!

And yet, when you’re looking for a specific number – a straight numerical answer to a specific question – it can be surprisingly difficult to find. This email that Eliana sent while doing the research beautifully illustrates the challenges:

1. I couldn't find raw data for the IPCC 2018 report. I found supply data in the second chapter in table 2.6 but had no such luck for demand. Maybe we have to eyeball it from Figure 2.14. Figure 2.19 also has demand/consumption data so we could try to eyeball it there. Looks like the median is just over 200 EJ for electricity and maybe 420 EJ for total energy for 1.5 degree scenarios.  

2. In the 2019 IRENA report, Figure 8 suggests total final energy consumption will be 351 exajoules in 2050. Electricity will be 86% of that (177.99 EJ or 47,775 TWh). In Figure 9, though, the consumption is only 47,056 TWh. Do you think it's worth trying to understand the missing TWh there? Does my math/approach make sense?

3. Also confusing in the IRENA reports--the 2018 report says that electricity consumption in end-use sectors will be "over 42,000 TWh" (see page 38 here). Scroll down a bit though and the infographic on page 40 shows a different number: 41,508 TWh. Do you think that difference is worth trying to figure out? It's possible the consumption number is pre-loss. 

4. I wanted to find the estimates for the reference case in the IRENA report but all I could find was energy supply (which I equate to generation but maybe incorrectly? My research suggests they are equivalent for these purposes) and electricity consumption.

5. For the IEA data, they have an export of raw data (attached) but it seems to differ from what is in the report. For example, page 113 of the report says electricity demand will be 60,000 TWh in 2050. Electricity demand in the raw data = 169 EJ (~47,000 TWh). 

In general, I aimed to find the raw data instead of trying to estimate from a chart/infographic but IPCC and IRENA did not make it possible to do that.

Numbers presented without adequate explanation; apparent inconsistencies within a single report; critical values that have to be eyeballed from a chart; values not presented in the form you need. I used to think, when I had difficulty finding a number I needed, that it was a personal failing – I had forgotten some critical library science lesson from middle school, or I was too lazy to do the proper legwork2. But in actuality, this is just legit difficult!

To be more precise: it’s often not difficult to find “sort of” the number you’re looking for. But finding a clear and direct answer to precisely the right question – even a seemingly straightforward question – can be another matter.

To Get a Clear Answer, You Need a Clear Question

“Energy usage” isn’t really a very good number to look for; it’s surprisingly ambiguous. We wound up looking for four numbers from each source:

• Energy generation

• Energy consumption

• Electricity generation

• Electricity consumption

You may be wondering about the difference between “generation” and “consumption”. The amount of power used is less than the amount generated, because there are losses along the way. When burning natural gas to produce electricity, quite a bit of the energy in the natural gas disappears up the smokestack as waste heat. Some electricity is lost to resistance in the transmission and distribution lines. And so forth.

I had planned to give precise definitions of what we’re measuring when we talk about “generation” vs. “consumption”, but to make a long story short, I failed. I’m not even convinced that the various data sources we’ve drawn on here are using consistent definitions. To give you an idea of how messy this gets when you dive in, the IEA definition of “primary energy consumption” contains 15 bullet points, one of which is:

Geothermal electricity net generation (converted to Btu using the average annual heat rate of fossil-fuel fired plants), geothermal heat pump energy and geothermal direct-use energy

I could try to explain what I think “converted to Btu using the average annual heat rate of fossil-fuel fired plants” means, but that would be missing the point. The point is that these seemingly simple terms, when applied to the real world, turn out to defy simple definition.

Enough Words, Show Me The Data!

OK, OK, here you go. (See the “Appendix” at the end of this post for an explanation of the six data sources we used.)

Looking over these figures, plenty of oddities stand out. I list some below. Feel free to skip the list; the theme is that the models disagree with one another in ways that imply there is a lot of uncertainty here, and there is likely a considerable amount of miscommunication, if not outright errors.

• The “all energy consumption” figure for the EIA report is a major outlier, over twice as high as the next-lower projection. Presumably this is because it’s assuming “business as usual”, unlike the other sources we picked.

• Even setting aside the EIA figures, the projections vary substantially, generally by 40 to 50 percent.

• All of the models show overall energy usage as being much higher than electricity usage, i.e. they show quite a lot of usage of energy in forms other than electricity. I’m not sure how this is compatible with the net-zero and/or 1.5°C scenarios used by some of these models. There are some ways to use non-electrical energy while maintaining net-zero emissions3. But I’m skeptical that a realistic global scenario would involve a large proportion of this sort of thing, and this makes me think that there is a communication gap somewhere.

• The IEA figures show a very large gap in generation vs. consumption, especially for electricity4. I have to think that this represents a mistake of some sort (possibly ours). Never Trust a Number!

• Electricity figures for IRENA 2.0°C scenarios are lower than for IPCC 1.5°C scenarios. Since electrification is closely associated with emissions reduction, this seems difficult to reconcile, unless IRENA is assuming much slower economic growth.

• The IRENA numbers changed noticeably from 2018 to 2019, emphasizing how these projections are a moving target.

You can find the numerical data in copy/pasteable format, along with Eliana’s complete research notes, here.

What Can We Conclude?

Suppose you’re trying to model one of the questions I discussed at the top of the post; for instance, land usage for solar panels in 2050. What figures should you use for overall electricity demand?

If there is one single lesson I’m taking away, it’s that there is no single figure. We, the collective global “we”, are simply not in a position to make a precise prediction 28 years out. Not by a long shot. Solar eclipses: predictable. Solar power: no way.

However, we shouldn’t throw up our hands and walk away, either. We can’t usefully project a single figure, but we can get some sense of a plausible range. Taking electricity consumption as an example:

• The six data sets project values ranging from 126 to 190 quadrillion BTU.

• The higher figures come from scenarios that project lower emissions (presumably these scenarios assume a more rapid shift away from other types of energy). If we are trying to model an aspirational, 1.5° scenario, we should focus on those higher figures.

• None of these sources provide confidence intervals (sigh), but the IPCC report at least includes a graphic showing the amount of variation between the 19 separate models5 on which it is based. I struggle to draw anything from this graphic more concrete than “there is a lot of variation”6. This is not the same thing as getting confidence intervals from the individual models, but it at least gives us some view into the amount by which these projections will change based on differing assumptions or methodologies.

The upshot is that in an aspirational, 1.5°C and/or net-zero-by-2050 scenario, global electricity consumption in 2050 could easily reach 190 quadrillion BTU, and the range of plausible scenarios extends quite a bit higher. For rough napkin math purposes, if I were thinking about potential bottlenecks in electricity production and distribution – which is exactly why we did this research! – I would be inclined to plan for figures as high as, say, 250 to 300 quadrillion BTU. That is of course a very hand-wavy number, but I don’t know how to do better.

To be clear, 300 QBTU is my figure for “we might need that much electricity, so let’s make sure there aren’t any bottlenecks that would prevent us from getting there if we need to”. The actual figure is likely to be quite a bit lower.

Seriously, After All That Work, We Just Get “I Dunno, Could Be 300 Quadrillion”?

I know, right?!?

This isn’t as bad as it sounds. It’s important to keep in mind the purposes to which we might put these projections. If you’re trying to make concrete business decisions – whether to start construction on another wind turbine manufacturing plant, or how much raw material to order for your solar panel factory – you need much more precision. But no one is making decisions like that for 2050 yet!

The applications I have in mind are order-of-magnitude questions, like “will we have enough land for all the renewables we’ll need to deploy in the next few decades”. For a question like that, it’s not critical to have a precise figure. But it is critical to have a clear idea of the range of plausible scenarios. I’m happy we’ve done this exploration, and I look forward to building on it.

There’s certainly room to dig deeper into the published work, but that’s beyond the scope of what we’re likely to undertake right now.

It Shouldn’t Be So Hard To Compare Sources

A lot of work went into this post. For the most part, that’s not because the question was inherently tricky. It’s because the sources are hard to work with. News articles and executive summaries present conclusions, but don’t provide enough detail and context to allow you to check their assumptions, or make apples-to-apples comparisons with other sources. Academic and quasi-academic sources drown you in detail, while still not doing a good job of presenting data in a systematic or standardized way or explaining assumptions and terms.

We can do better.

I’ve talked before about the need for better resources to explain the range of expert opinion on climate topics. This will start coming up more often in this blog, as we explore more of these uncertain or controversial topics. I’m hoping that this blog can serve as one small part of the solution. To that end, if you have any light to shed on the topic of this post – energy trends to 2050 – please comment!

Appendix: Data Sources

The projections discussed here come from the following sources:

EIA

A branch of the US federal government, “The U.S. Energy Information Administration (EIA) collects, analyzes, and disseminates independent and impartial energy information to promote sound policymaking, efficient markets, and public understanding of energy and its interaction with the economy and the environment.” They have published the International Energy Outlook every year since 1985. In this report, they model global and regional energy generation and consumption based on existing laws and regulations. It assumes current trends/relationships between supply, demand, and prices. The EIA develops this model as "a baseline for estimating the effects of policy or technology changes" (https://www.eia.gov/pressroom/releases/press487.php). In other words, this is a “business-as-usual” model, minimizing the impact of future technical, commercial, or policy developments.

IEA

The International Energy Agency “is a Paris-based autonomous intergovernmental organisation, established in 1974, that provides policy recommendations, analysis and data on the entire global energy sector, with a recent focus on curbing carbon emissions and reaching global climate targets, including the Paris Agreement. The 31 member countries and 11 association countries of the IEA represent 75% of global energy demand” (Wikipedia).

The IEA special report, titled Net Zero by 2050: A Roadmap for the Global Energy Sector, was requested by the UK government to inform conversations that would occur at COP26. It purports to be "the world's first comprehensive study of how to transition to a net zero energy system by 2050" (https://www.iea.org/news/pathway-to-critical-and-formidable-goal-of-net-zero-emissions-by-2050-is-narrow-but-brings-huge-benefits) and examines three scenarios: Stated Policies, Announced Pledges, and Net-Zero Emissions. The Stated Policies Scenario only takes into account existing policies. The Announced Pledges Case takes into account all commitments by governments, regardless of whether policies exist to support those commitments. Net-Zero Emissions models the actions required to achieve net-zero emissions by 2050. We used the figures from the Net-Zero Emissions model.

IPCC

The Intergovernmental Panel on Climate Change is a UN organization that undertakes the herculean task of consolidating the scientific literature on climate change (and mitigation) and producing consensus reports. The IPCC's special report Global Warming of 1.5º C was drafted in response to an invitation by the United Nations Framework Convention on Climate Change as part of the adoption of the Paris Agreement. The report aims to identify the impacts of global warming of 1.5 degrees Celsius above pre-industrial levels to inform policy decisions and strengthen the global response. There were 19 modeling frameworks and 529 scenarios utilized, all of which represented possible scenarios in which a warming of 1.5 degrees would occur.

IRENA

From the web site:

The International Renewable Energy Agency (IRENA) is a lead global intergovernmental agency for energy transformation that serves as the principal platform for international cooperation, supports countries in their energy transitions, and provides state of the art data and analyses on technology, innovation, policy, finance and investment. IRENA drives the widespread adoption and sustainable use of all forms of renewable energy, including bioenergy, geothermal, hydropower, ocean, solar and wind energy in the pursuit of sustainable development, energy access, and energy security, for economic and social resilience and prosperity and a climate-proof future.

IRENA's 2018 report builds upon a joint report published with the IEA in 2017 titled Perspectives for the Energy Transition: Investment needs for a low-carbon energy system. With the publication of Global Energy Transformation: A Roadmap to 2050, IRENA updates its analysis based on rapidly changing policies and markets. The report sets out to provide a framework for the energy transition for a scenario in which global warming is limited to 2 degrees Celsius with 66% probability.

We present three distinct sets of figures from IRENA:

1. The 2018 report described above.

2. Another set of figures from the 2018 report, representing a “business as usual” scenario (rather than assuming sufficient progress to limit warming to 2°C).

3. Figures from an updated report issued in 2019:

This 2019 edition updates IRENA’s analysis of key countries and regions, and it presents a deepened perspective on electrification with renewable energy – the key enabling solution of the energy transition. The report also details new findings related to the costs, subsidies and investments needed for the transition. IRENA’s socio-economic footprint analysis delves into the implications of the transition, providing footprint measurement in terms of GDP, jobs and welfare. A discussion of the socio-economic implications of carbon taxation is presented. Climate damages have been included into the macroeconomic analysis, bringing about important socio-economic consequences. The need for holistic employment and just transition policies is highlighted by analyzing the implications of the transition on whole-economy and energy sector jobs. The focus also has been strengthened on how high shares of variable renewable energy (VRE) can be integrated into energy systems. In addition to discussion on the role of electrification, solutions for decarbonising heating, cooling and transport demand are also presented.

No one knows how much energy we'll use in 2050.

Stated like that, it may seem obvious; the future is, famously, not yet written. But… isn’t that the sort of thing that “they” are supposed to know? I mean, I’m not sure who “they” are – I’ve never actually met a “they”1 – but, there’s, like, economists, and all the people who work on the IPCC reports, and so on. They have, you know, models and stuff, right? I mean, someone must know, because we’re always reading articles about how this and that is going to happen by such and such year.

Indeed, people do routinely act as if they did know things like how much energy we’ll use in 2050. Even in academic settings. For instance, I googled “global solar energy in 2050”, clicked the first interesting link, and arrived at the paper Low-cost renewable electricity as the key driver of the global energy transition towards sustainability. Skipping to the Conclusions section, I find:

The solar PV industry is capable of providing all required capacities, as shown by Verlinden [98], since 70 TW of PV capacities can be ramped up by 2050, which is about 10% more than 63.38 TW found in this research.

This confident statement – “the solar PV industry is capable of providing all required capacities” – appears to be based on the fact that the projected capacity is 10% larger than the projected need. If either figure turns out to be off by 10% in the wrong direction, the conclusion would be invalidated. How likely is it that projections of energy usage in 2050 will be off by 10%? Well, it’s easy to find projections that disagree with one another by much more than that, so the probability that at least some of these projections are off is 100%!

Some Questions Have No Answers; Some Have Too Many

I’ve long wanted to have a clear idea of how much energy the world will use a few decades down the road. Energy usage is an important input to many other interesting questions about how the green transition will play out. However, I’d hesitated to dig in, because I suspected it would be difficult to find a useful answer.

That changed recently, when – as I’m excited to report – Eliana Schwartz and Janki Patel joined to help with research and analysis for this blog (see below). With folks less lazy than me more collective time and energy to invest, it was reasonable to dig into the published literature on 2050 energy usage. I’ll go through the findings in a subsequent post, but if you’d prefer to dive in yourself, you can find the data Eliana gathered (along with her extensive notes) here.

Across the various sources, there are figures ranging from 3262 to 8863 quadrillion BTU. Both are from credible sources: the International Energy Agency (IEA) and the U.S. Energy Information Administration (EIA), respectively.

The difference between these numbers is huge: a ratio of 2.7 : 1. That’s far more than the 10% margin in the “Low-cost renewable electricity” paper. But even this huge discrepancy is only one part of the problem.

Confidence Intervals

The EIA projection for 2050 energy usage is not really 886 quadrillion BTU. If you click through the footnote to find the original source, you’ll see that it says “886.3”.

This figure probably comes from some complex mathematical model that actually spit out a number like 886,319,455,081,534,769 BTU. Obviously the model is not accurate to that level of precision, so someone rounded it.

When presenting figures like this, you’re really supposed to round down to a level of precision that matches the accuracy of your model. 886.3 has one digit after the decimal point, implying that true figure should be between 886.2 and 886.4. Obviously, the model can’t possibly be that accurate.

In any case, rounding is a blunt tool for conveying precision. You can better convey the expected accuracy of a figure by using a “confidence interval”. As an example of the right way to do it, page 41 of the IPCC WG1 AR6 technical summary states:

For the decade 2011–2020, the increase in global surface temperature since 1850–1900 is assessed to be 1.09 [0.95 to 1.20] °C.

“1.09” is the estimate, and “0.95 to 1.20” is the confidence interval. A footnote explains that the IPCC’s modeling work projects a 90% probability that warming in 2011-2020 was between 0.95 and 1.20 degrees C. This gives us a clear idea of the level of uncertainty in this estimate – about 12% in either direction. (And this is an estimate, a very careful and thorough estimate, of what happened in the past! The uncertainty for a projection three decades into the future will be much larger.) Computing confidence intervals adds extra work, but when working with uncertain figures, it’s vitally important to communicate how uncertain the values are.

Unfortunately, it’s only possible to evaluate known sources of uncertainty. So even when a confidence interval is reported (and was computed correctly), you shouldn’t treat it as gospel. But if projection doesn’t include a confidence interval, you can’t tell how seriously to take it, and you have to suspect that the author doesn’t know, either.

Assumptions

A broad statistic like “global energy usage” depends on, well, everything: population growth, economic growth, relative importance of different economic sectors, efficiency improvements, shifting consumer preferences… everything. To project this into the future requires making assumptions about how every aspect of the global economy will evolve. Sometimes these assumptions are explicit, sometimes they’re implicit, but they all affect the result.

Unfortunately, sometimes these assumptions aren’t very good. Consider the US EIA figure cited above. It comes from this report, dated October 6, 2021. Poking around, on slide 11, I find:

Electric vehicle stock… represents 31% of total passenger vehicle stock by 2050

That is, the EIA is assuming that in the year 2050, 31% of passenger vehicles will be electric. Given current rates of EV adoption, this strikes me as implausibly low (consider that GM has pledged to stop selling gasoline vehicles entirely by 2035, and EVs have already reached 35% market share in China, the largest market)4. EVs use far less energy than internal combustion vehicles, so by under-estimating the shift to EVs, the EIA is over-estimating overall energy usage.

These particular publications don’t attempt to keep their assumptions secret. The EIA report explicitly assumes we continue “business as usual”, whereas the IEA projects a scenario in which we reach net zero emissions. But that still leaves the onus on us, the reader, to decide which projection (if either) to rely on. And it’s not always easy to tease out all of the relevant assumptions; even the authors may not be explicitly aware of every assumption they’re making.

All Models Are Wrong; Are Some Useful?

You may have seen the quote “all models are wrong, but some are useful”. I think the intended meaning is that some models, under some conditions, provide answers which are sufficiently accurate for some purposes. One way of thinking about this is that a model is useful if you apply it under the conditions and purposes where it provides reasonable answers. In other words, the onus is on the user, not the model.

If you open the EIA report, scroll down, read that global energy consumption in 2050 will be 886.3 quadrillion BTU, and use that to start making firm conclusions about precisely what the world will look like, you’re relying on the model for a purpose for which it is not useful.

On the other hand, if you compare the conclusions of the EIA report with other analyses, you might be able to get an idea of the range of plausible trajectories, and you might be able to learn something about how different assumptions affect projected energy usage. That is a purpose for which the model might be useful. I’ll explore this further in an upcoming post.

For today, the key takeaway is to remember that the future is uncertain. The Internet is littered with confident, quantitative projections of the future, awash in numbers that have no confidence intervals and don’t clearly state their assumptions. Don’t place much confidence in these numbers, and don’t assume they apply to your purpose. In Never Trust a Number, I explain how easy it is to be led astray by numbers in general. When a number attempts to describe the future, it’s even less trustworthy.

A corollary is that we tend to be overly confident in our ability to predict the future, and thus, overly hasty to dismiss certain technologies as unnecessary. The inherent uncertainty in our models is another reason to err on the side of developing multiple paths to decarbonizing every sector of the economy.

Postscript

The Climateer team has expanded! I’d been doing this in my spare time, but there are now three of us working in our spare time:

• Eliana Schwartz is a Senior Product Manager at WattBuy, working on home electrification and decarbonization. She is passionate about contributing to research that helps demystify climate and energy data. Her LinkedIn is here.

• Janki Patel works with firms on bringing sustainability into their business strategy. She recently graduated from UC Berkeley's MBA program, where she focused her time on trying to understand the intersection between equity and climate change. You can connect with her on LinkedIn.

I’m very excited to have Janki and Eliana on board. They’re able to engage in research and analysis that I wasn’t previously able to undertake, which will expand the scope of what we can explore here. We’ll be posting more frequently. And it’s wonderful to have their additional perspectives.

I haven’t actually run out of topics. In fact, my notes file lists about 50, many with a page or more of seed material. However, most of these potential topics would require a lot of research to flesh out, and I’m too lazy / busy / impatient2 to put in that work.

My goal for this blog, to the extent that I have a coherent goal, is to shed light on the most effective paths for mitigating climate change; basically, topics that can help us make choices and set priorities. Here are a few of the items on my list:

• Which important emissions sources are least addressed?

• Are there limiting factors to the greening of the economy that we need to be addressing? (E.g. skilled workers; key raw materials such as lithium, nickel, or copper; transmission lines; land. Which of these are on track to be a problem?)

• What are the most important actions individuals can take? (As a consumer; as a voter / citizen; as an employee; as an investor; as a donor.)

• How much of a role can improved land management (including agricultural practices and forestry) play in drawing down carbon?

• How close are we to the point where demand for fossil fuels will start to drop significantly, and is the market reacting properly to that, or should we do more to discourage new supply?

These questions all have one thing in common: they require quantitative evaluations of how things are playing out in the real world. By contrast, most of my topics so far have been binary propositions whose truth can be reasoned from first principles. That’s not a coincidence. Guess which type of question requires more research?

To make further progress, I can’t keep dodging the messy empirical questions. Unfortunately, that way lies danger or, worse, hard work.

Quantitative Evaluations Of Climate Topics Are Hard

Propositions like “we need to address all of the emissions sources”, or “electricity is harder to store than fossil fuels, so we need better coordination of production and consumption” are sort of self-evident. A few motivating examples are sufficient to explain the idea and its importance, and then I can move on to discuss potential solutions.

Quantitative questions and relative comparisons, by contrast, are less amenable to quick conclusions. Take the topic of “regenerative agriculture” (a set of agricultural practices, such as no-till farming, that are thought to provide a number of benefits, including sequestering carbon in the soil). It’s easy to find support for the idea:

• National Geographic: “The solution to climate change is just below our feet.”

• oneearth.org: “A growing consensus is emerging among soil scientists that regenerative agriculture – agricultural practices that remove carbon from the atmosphere and put it back in the soil – could deliver a huge win for the climate.”

• A recent Canary Media writeup notes that “The IPCC’s latest report lists soil carbon sequestration as one of the lower-cost, more readily available options for carbon dioxide removal.”

Sounds like this is a settled question, yes? And yet, there appears to be a lot of question as to how much carbon can be captured, which techniques are most helpful, and how to measure the impact. There is even some question as to whether some of these practices have any net climate benefit at all. For instance, in a recent podcast, Eric Slessarev (a staff research scientist at Lawrence Livermore National Laboratory) notes two significant problems with no-till agriculture:

• The carbon captured in the soil is in the form of organic carbon, which is readily consumed by soil microbes. As a result, the carbon may only be stored temporarily.

• We’ve recently been learning that there is a complex interplay between carbon at different depths in the soil. No-till agriculture can actually reduce the amount of carbon deeper in the soil. Unfortunately, many studies have only looked at the top 12 inches, and therefore have failed to account for this reduction in deep-soil carbon; the results may therefore be misleading.

So, how much carbon can we realistically expect to capture via regenerative agriculture, and how long will that carbon remain out of the atmosphere? I don’t know, and at this point, it seems that no one really knows for sure. There are a lot of questions for which we simply don’t have definitive answers yet. Many important questions in climate change are questions about how the future will play out, so they’re actually predictions3, which are always fraught.

Blindly Repeating Someone Else’s Take Is Rarely a Good Idea

Uncertainty is unpleasant. We like knowing the answers to things. I find it frustrating that I don’t understand how much of a role regenerative agriculture can play, which bottlenecks deserve attention, or exactly where we sit on the fossil fuel demand curve.

For every messy real-world question, the Internet provides an unlimited supply of easy answers. Earlier, I listed a few sources which were extremely positive regarding the potential for regenerative agriculture. It’s very tempting to take these at face value. “Regenerative agriculture” just sounds wholesome (someone did a nice job of branding there), and the idea has a win/win appeal: simply by updating the way we manage our croplands, we can mitigate climate change and improve soil health, without significant cost.

However, on any topic that’s even mildly controversial, I’ve learned that it’s dangerous to repeat a conclusion if I’m not prepared to explain the facts and arguments that support it. I start to feel ownership of the idea. When I read contradictory evidence, I discount it. When challenged, my natural reaction is to double down; but since I don’t have a deep understanding to fall back on, this pushes me toward unproductive forms of argument.

Even for topics where there seems to be a clear correct answer, there will be people who disagree, and I’d like to be able to hold up my end of a detailed, thoughtful, fact-based discussion. Of course, we can hardly all take the time to become informed experts on every topic. So it’s important to pay attention to the distinction between “opinions I hold because I understand the objective facts and the arguments based on those facts” and “opinions I hold because someone I like espoused them”. If you’re parroting someone else’s conclusion without understanding how they got there, then you can’t explain it, defend it, recognize the level of uncertainty it reflects4, or update it to reflect new data. You can angrily repeat it on Twitter though! But maybe you shouldn’t.

Navigating Uncertainty

For topics where there is no expert consensus, how to proceed?

I touched on this in Why Everyone Disagrees About Climate. Sometimes, with some digging, it may become clear that one side of the dispute is simply out of date, or their arguments have been addressed, and a clear answer emerges.

Often there won’t be a clear answer. My intuition is that there is still value in wading into the messy, unanswerable questions. At a minimum, we can highlight the uncertainty, and try to delimit its boundaries. It may be possible to sharpen the conversation by highlighting the key points of debate. Or it may be possible to identify some conclusions we can draw despite uncertainty.

What is certainly true is that none of this will be possible without a lot of legwork. In Never Trust a Number, I argued that there are no shortcuts to understanding. This is especially true for unresolved, messy real-world questions like “how does the likely future supply of lithium match up with likely future demand”. By reading predigested opinions from experts, we can fast-forward the learning process a bit, but only a bit. It’s necessary to read a range of opinions, understanding the arguments and their merits.

This brings me back to the start: I’ve started having trouble coming up with new topics to blog about, because I’m running out of “easy” topics. I’m going to try diving deeper into the messy, quantitative, future-prediction questions. I expect that I won’t be posting as often, though I’ll probably come up with some research-lite pieces to mix in along the way. In the meantime, keep in mind:

• Navigating climate change requires tackling a lot of messy, empirical, future-looking questions.

• Don’t get too attached to any specific idea, especially ideas that you’re repeating without understanding.

• But don’t just throw up your hands. Even when a question can’t be answered precisely, it can often be answered loosely, and we can get a lot done with that.

P.S. I just listened to David Roberts of Volts being interviewed by Jason Jacobs of My Climate Journey. It’s an outstanding exploration of how to think about the climate situation, and very accessible. Highly recommended for anyone who reads this blog. You can find it here (Volts) or here (MCJ).

Aaand then I found myself repeatedly bumping into avoided-emissions offsets that seemed really important and worthwhile. For instance, in much of the developing world, it’s common for refrigerants to be vented into the atmosphere when a refrigerator or A/C unit is being serviced. These refrigerants are powerful greenhouse gases; disposing of them properly would be a big win. A startup, Recoolit, is working to capture and destroy these refrigerants. Their business model? Selling offsets.

Organizations like the Environmental Defense Fund are promoting REDD+, a system for protecting tropical rainforests using – you guessed it – carbon credits (aka offsets).

I really like the work Recoolit is doing. They’re addressing an important need, in an under-addressed niche, using a clever approach. And offsets are the obvious funding source. I found myself going through mental contortions to try to reconcile this with my conclusion that avoided-emissions offsets don’t make sense. For instance, I found myself saying:

…by the time a batch of refrigerant has found its way into a window A/C unit in Jakarta (Recoolit’s headquarters city), it has escaped effective control, and so… in some sense, it has already been, like, “emitted”? It’s not yet loose in the atmosphere, but that’s where it’s likely to wind up. And so collecting that refrigerant isn’t “avoiding” emissions, it’s actually negative emissions, like DAC (pulling greenhouse gases out of the atmosphere)? Kinda? If you look at it the right way?

This is maybe not completely wrong but it’s definitely a stretch. And at this point I was trying to hold onto the mental position that DAC is good, but avoided-emissions offsets are bad, unless they’re the Recoolit kind of avoided emissions which are actually good, and…

Who Should Buy These Offsets?

I wound up raising this on the My Climate Journey Slack board, where the thread wandered onto the question of who is buying the offsets. (Many thanks in particular to McGee Young, founder and CEO of WattCarbon, for a stimulating discussion.)

Suppose that a tropical landowner is planning to raze some rainforest; PG&E (my local utility) pays them not to; and then PG&E claims an “offset” which allows them to meet some net-zero target while continue to operating a gas plant. This is the precise scenario I’ve argued against: the gas plant is still operating, so we have not achieved zero emissions. (See that previous post for a longer explanation.)

If, say, Warren Buffet were to purchase that same offset, there’s no problem. Warren Buffet, personally, is not generating a lot of emissions1. Anything he might do to reduce emissions, regardless of where he does it, would be strictly to the good.

So, offsets seem to be a fine thing only when the purchaser doesn’t use them to counterbalance some polluting activity. But then, why call them “offsets”?

Review: How Are We Planning To Get the World to Net Zero, Again?

To make sense of all this, I think it’s necessary to review the general concept of how the world gets to (net) zero emissions. There’s not really any canonical plan, but I think most serious participants are working from a framework that looks something like this:

• The goal is to zero out all manmade greenhouse emissions, so that the level of greenhouse gases in the atmosphere stops increasing.

• The overwhelming majority of emissions-producing activities need to stop. No more extraction of fossil fuels, minimize or mitigate methane-producing agricultural activities, no more production or (especially) release of greenhouse-inducing refrigerants, etc.

• In some cases, rather than ending the activity, we might modify it so as to eliminate the emissions. (For instance, capturing and sequestering the CO₂ released from limestone during cement manufacturing.)

• In a few cases, when neither of the preceding options are practical, we might continue the polluting activity, but counterbalance it by pulling CO₂ out of the atmosphere.

This can be summarized as: each emitting activity must either cease, be modified to not emit, or be counterbalanced by negative emissions. Which then begs the question: who is responsible for all this?

The obvious starting point for discussion is that everyone is responsible for mitigating their own emissions. Homeowners must heat their homes and water without burning natural gas. Utilities must provide emissions-free electricity. Steel mills must produce carbon-neutral steel. And so on.

Then we get into adjustments based on questions of ability (a low-income family can’t float the up-front cost of a heat pump), motivation (try convincing Kim Jong-un that it’s time for North Korea to stop using coal), and equity (explain to Africans that it was OK for the West to industrialize using coal, but they’d better not). I couldn’t even begin to explore the complexities here, so I’ll merely note that any complete solution will need to include subsidies, regulatory coercion, etc.

Offsets Should Come From Exceeding Your Responsibility, Not Meeting It

As soon as we say that everyone is responsible for their own emissions, the inescapable problem with “offsets” becomes clear. Let’s revisit the example of PG&E paying a tropical landowner to protect a patch of rainforest. The landowner has no standing to provide an “offset” to PG&E; any credit for protecting that forest is needed to discharge the landowner’s own responsibility.

This also makes it clear why offsets for DAC (Direct Air Capture) are fine. If my responsibility is to have zero emissions, but I build a DAC plant on my land, then I have negative emissions – I have exceeded my responsibility. I can sell a credit for those negative emissions without breaking the responsibility model.

This is why offsets / credit markets are often used in conjunction with a cap-and-trade system. Under cap-and-trade, participants are not required to have zero emissions; they are allowed to emit up to their “cap”. If they emit less than their cap, that generates a salable credit. It’s easier to exceed your responsibility when your responsibility is not “halt all emissions”.

Of course, cap-and-trade only works if the cap is eventually reduced to zero, at which point the “avoided emissions” credits will dry up. In this model, offsets are allowing us to optimize the sequencing of our mitigation efforts – pick the low-hanging fruit first – while still requiring us to eventually zero out everything.

Finally, it’s worth reiterating a couple of other issues with offsets that I mentioned in my earlier post. Offsetting carbon emissions doesn’t do anything to mitigate other impacts, such as local air pollution. And offsetting the emissions from, say, a steel mill doesn’t help us climb the learning and scaling curves on producing carbon-free steel2, as we will eventually need to do.

So… How Do We Save the Rainforest, Then?

Avoided-emissions offsets are funding important work. If we can’t use offsets, then how do we fund that work?

I don’t know. Earlier, I noted that any complete vision of a net-zero planet will have to incorporate subsidies and regulatory coercion to get everyone to do their part. Coercion can only go so far; it’s hard to apply beyond national borders, and limited by political will. That leaves subsidies. Who will pay for them?

In my example of the classic-but-flawed offset model, PG&E is paying the subsidy to protect a patch of rainforest. If we rule out that approach, then we need to find another payer. The biggest challenges, again, come when we cross national borders. For instance, here in the US, the Inflation Reduction Act provides subsidies for a wide variety of actions within our borders. But for developing nations that can’t / won’t fully fund their own transition, where can the money come from?

Again, I honestly don’t know. This strikes me as one of the biggest challenges we face. Here are a few possibilities, presented more to flesh out the difficulty than to claim a solution:

• Offsets do become more practical under cap-and-trade. Needless to say, establishing a cap-and-trade scheme covering the entire planet would be challenging!, but it would provide a pathway to funnel mitigation funds across borders.

• Suppose we can imagine a reasonable cap-and-trade system, without actually trying to get global buy-in to enforce it. We could modify a conventional offset scheme so as to emulate the effect of cap-and-trade. For instance, suppose that in some third-world country, there’s an electricity plant which emits 10 Mt/year of CO₂. If PG&E pays to replace that plant with solar + batteries, we could give them a credit equivalent to the plant’s hypothetical emissions cap - say, 9 Mt/year today, gradually declining to zero. There are problems with this approach: it would leak like a sieve (many emitters would ignore their obligations under the “cap”), and we might wind up with multiple competing systems, allowing operators to participate in whichever system treats their particular circumstance the most favorably. But it would still be better than just crediting PG&E for the full 10 Mt/year.

• Cross-border government funding. The problem here, of course, is the political challenge of finding sufficient funding.

• Philanthropy. Again, this is great if the money is there, but it won’t be, in sufficient quantity.

Giving Green recently published A business case for beyond net zero, which contains some interesting ideas and promising case studies.

One way or another, we need to find a way to fund projects that won’t happen without external funding, and do it in a way that doesn’t give polluters a loophole to keep polluting indefinitely. Just as we’ll eventually need to find a way to pay for extra DAC to roll back some of hour historical emissions. I don’t know how to do that – but it’s a question we’re going to need to answer. Post your thoughts in the comments!

You can’t spend more than 10 minutes reading about climate change without running into opposing claims. I’m not even talking about denialism or politically motivated positions. I’m talking about well-intentioned, serious folks confidently saying contradictory things.

For example, it’s a commonplace that the industrial sector will be a major stumbling block on the road to zero emissions. Here’s a typical framing, from the Brookings Institute (emphasis added):

Steel, cement, and chemicals are the top three emitting industries and are among the most difficult to decarbonize, owing to technical factors like the need for very high heat and process emissions of carbon dioxide, and economic factors including low profit margins, capital intensity, long asset life, and trade exposure.

Sounds reasonable, no? “Very high heat”, hmm, yes, obviously difficult. “Process emissions”, I don’t know what those are, but they certainly sound serious1.

Meanwhile, Canary Media reports (emphasis added):

…the options to reduce the carbon-intensity of concrete … are broadly available, practical and cost-effective today, building decarbonization experts say.

And in fact, a variety of companies are developing carbon-neutral or even carbon-negative cement. As for steel, an interview earlier this year, Rebecca Dell of ClimateWorks notes that there are several strong contenders for decarbonization: substituting hydrogen for methane in "direct reduction”, or switching to one of several electricity-based approaches. Modest shipments of carbon-free steel have already begun. Maybe the industrial sector isn’t so hard to decarbonize?

And of course this isn’t the only topic where opinions diverge. Reforestation, agricultural carbon sequestration, shipping, scaling mineral production, the role of geothermal and nuclear power, grid stability, viability of hydrogen… it’s almost harder to think of a major topic where there isn’t widespread disagreement.

If this leaves you confused, you’re not alone. I’m right there with you.

Some Contradictions Vanish On Closer Inspection

In the example above, we have some people saying that the industrial sector is “hard to decarbonize”, while other people are busily decarbonizing it. That’s not actually a contradiction2! Heavier-than-air flight was legit “hard”, but we’ve gotten pretty good at it.

It’s easy to lose track of the fact that divergent viewpoints may not actually contradict, because messages tend to get simplified into sentiments like “cement is a problem” vs. “cement is not a problem”. The truth is usually in the messy middle: cement is a solvable problem, if we put in the work. If you look for simple answers, you’ll wind up with contradictions.

Failing To Allow For Change

Heat pumps have a reputation for struggling in cold environments, because historically, they did struggle. However, Consumer Reports recently reported3:

But that’s old news. When properly installed, plenty of today’s air-source heat pumps… can keep your home toasty even amid bone-chilling cold, using far less energy than other types of heating systems.

The article goes on to note that a poorly designed installation can still have problems. So: do heat pumps work in cold climates? In the past, no; today, yes; but only if done properly. There are lots of people who “know” it’s a problem because they aren’t up on recent developments, or had a bad experience with a bad installer. Others have had good experiences. People in either situation will happily share their views, and don’t always do a good job of clarifying the scope of their knowledge.

Similarly: the costs of solar, wind, and battery power have been plunging. Myriad other technologies are moving from idea to lab to pilot project. The facts keep changing, and people don’t always notice, or don’t update their views to reflect the implications.

It’s hard enough to keep up with the changes that have already happened; projecting forward is even trickier. For example, we’re likely heading toward an era where electricity will be super-cheap if you’re willing to be flexible in your consumption. (“Flexible” meaning locating near good sources of renewable energy, shutting down when solar + wind aren’t running, etc.) There are proposals to use this cheap, intermittent power for energy-intensive processes such as desalination or hydrogen production. Someone who hasn’t internalized the idea that tomorrow’s grid may look very different from today’s, will write off these ideas as uneconomic.

An oft-stated example that really bugs me: “EVs don’t help because we still use coal power on the grid they charge from”. The vehicle transition and the grid transition are both necessary, and the best thing is to run both transitions in parallel, so that we converge as soon as possible on all-electric vehicles on a zero-carbon grid.

Over-Generalizing

Another way contradictions can arise is when people over-generalize from limited experience: “My solar roof worked out great, so solar can solve everything.” (Try getting through an Alaskan winter on nothing but solar power.)

Another favorite is extrapolation from one flawed proposal to write off an entire field: “batteries don’t have the energy density to carry an airplane more than a few hundred miles, so we can’t decarbonize aviation”. (Aviation really is a challenge, but hydrogen power, atmosphere-derived fuels, and other approaches hold promise. We don’t need to use batteries for everything.)

A common mistake is failing to allow for scaling limits. For instance, singing the praises of hydropower, without noting that it can only be used in specific locations; or of planting trees to draw down CO₂, when there isn’t enough suitable land to solve the entire problem4.

How To Navigate the Maze

Spoiler: it’s hard!

The first step in resolving contradictory claims is to move past the headline and dig into the underlying arguments. What data is each author drawing on? What is their argument, and what exactly is their conclusion?

Then you can start to look for ways of resolving the apparent contradiction:

• When framed carefully, do the conclusions actually contradict?

• Is either author working from stale, suspect, or anecdotal data?

• Are their arguments sound? Is this good-faith, objective analysis? Or is one source relying on motivated reasoning?

• Does either author directly address the opposing argument? Is the rebuttal valid?

• Which argument aligns with your understanding of broad trends?

We Need Less Monologue and More Dialogue

When two domain experts5 contradict one another, it's a challenge for the lay reader – that's me, and (statistically speaking) you – to sort things out. It would be great if the experts could hash things out directly!

Unfortunately, this doesn’t seem to happen very often. On the My Climate Journey podcast, Jason Jacobs has been making an effort to bring in divergent viewpoints, which is a nice start.

It’s also just hard to keep track. Suppose I read a thorough discussion of whether steel manufacturing is “hard to decarbonize”. Then next month I’m reading about the steel problem again. I may not remember the previous discussion well enough to know whether it addresses the points made in the new article.

In my fantasy world, there would be a sort of living Climate Tech FAQ, recapping the relevant facts and common opinions on a broad range of topics, and able to serve as a cheat sheet – where is there expert consensus, what ideas have been superseded (or debunked), which seeming contradictions are just issues with framing or terminology, and where are there legitimate open questions. And since I’m fantasizing here, the site would sponsor thoughtful long-form discussions to help reconcile divergent views.

In the real world, the only advice I have is to apply good ol’ critical thinking skills, as discussed above. In particular, because things change so quickly, always be on the lookout for over-generalization, overly broad claims, and arguments based on stale data.

From Leah Stokes, a political scientist (and climate policy expert) at UCSB:

I’m trying to electrify my home in Santa Barbara. The time it takes to get permits to change my house is about a year. I’m still burning gas in my house for that year. Now we’re going back and forth about what kind of heat pump I can use. None of the system is oriented around climate being the most important thing.

In the New York Times, Ezra Klein writes (emphasis added):

[Many current environmental laws] were built for an era when the issue was that the government was building too much, with too little environmental analysis. The core problem of this era is that the government is building too little, in defiance of all serious environmental analysis. … Modest expansions to affordable housing or bus service are forced to answer for their environmental impact. But the status quo doesn’t have to win any lawsuits or fill out any forms to persist.

This tweet from Alan Cole really drives the point home:

The US has 42 MW of offshore wind production that is operational, 932 MW under construction, and 18,581(!) MW tied up in permitting.

To decarbonize the economy, many things have to come together – scientific, technical, economic, political, and societal changes are all needed. In recent years, there has been rapid progress across the board. We’re ready to build the carbon-free economy. The one big impediment that we haven’t addressed? It’s the bureaucratic barriers that get in the way of all change, even environmentally beneficial change.

Friction Is Insidious

Red tape adds delay, uncertainty, cost, and hassle to a project. But that’s not the end of the problem; in fact, it’s the best-case scenario. Worse is all the projects that are never initiated in the first place, because the barriers make the project uneconomical or simply too much trouble. Those projects won’t show up in any statistics, but they are a real cost of regulation.

Delays are also hard on innovation. To get a new technology from the laboratory to planet-wide scale requires many rounds of deployment and iteration. Permitting, public commentary, and approval processes can add years to each round, adding up to decades of delay before we get emissions under control. We don’t have decades to waste.

The Inflation Reduction Act contains a wide array of provisions, all aimed at building more stuff. If we can’t build – if we can’t build now – we’ll squander a massive opportunity. For example, there are tax credits to stimulate sales of electric vehicles, but right now everything is sold out for months in advance. I’m in the market for a new car, and I literally can’t find an EV to test drive; no dealer has one on the lot. What’s the point in stimulating sales of a product you can’t buy?

Change Used To Be Bad, Now It’s Good

Humankind has been transforming the environment since the dawn of agriculture, if not longer. Environmentally speaking, almost all of these changes have been negative. So our entire history, culture, and intuitions developed in an era where protecting the environment means preventing change. Virtually all changes were bad, so our entire mindset is oriented around protection through change prevention.

In other words, from an environmental perspective, the status quo was always preferable to the likely future, and so the environmentally friendly approach to new projects has been to oppose them1.

For perhaps the first time in human history, we’re entering an era where the status quo – a fast train to intolerable warming – is worse than the achievable future. Through the heroic efforts of countless people, we now have path to replace our fossil-fuel-based economy with something cleaner. Regulations that retard progress are no longer automatically aligned with protecting the environment. The status quo is bad, so we need to eliminate barriers to changing it.

Regulation Favors The Incumbent

Regulations tend to lock in the status quo. Per the earlier quote from Ezra Klein, the world is full of coal plants and gasoline refineries, that “[don’t] have to win any lawsuits or fill out any forms to persist”. Even when it comes to new construction, large incumbent businesses have the experience, regulatory connections, and pocketbook to navigate the system more easily than a newcomer, even a newcomer that is theoretically “better” in the ways that the regulation is meant to encourage. (See also: “regulatory capture”.)

Let ‘Em Drill, They Won’t Anyway?

Suppose we could wave a magic wand and make it easier (and faster!) to get permits for wind farms, new electricity lines, and other green projects; but as a tradeoff, we also had to remove barriers to fossil fuel projects.

I was going to write about how this would be a great deal because fossil fuel projects aren’t being built anyway. The writing is on the wall: gasoline cars are on the way out, renewable electricity is on the way in, and there are headlines like Half world’s fossil fuel assets could become worthless by 2036 in net zero transition. I keep reading2 about this; for instance, supposedly, one of the reasons oil prices are so high right now is that no one wants to invest in new capacity, even with the restrictions on Russian supply, basically because there's no perceived future in the fossil fuel market. If someone starts drilling a new oil well today, it might be profitable for a few years, but it will become a stranded asset before the initial investment is paid off.

Unfortunately, I didn’t keep track of any sources for this, and when I went now to look for evidence, I couldn’t find it. (Admittedly, I didn’t look very hard. I have a short attention span for research.) So rather than asserting that throwing open the regulatory floodgates wouldn’t lead to many new oil or gas wells, I’m going to leave it as a question to be pondered. Would it be a good deal to streamline barriers to change if we were forced to include fossil fuel projects in the deal? (If you have evidence in either direction, send it my way!)

In any case, it’s not a good idea to squash supply of fossil fuels until renewables are ready to step up. With Russian oil and gas supplies somewhat cut off, the coming winter – possibly several winters to come – are going to be horrible in Europe and elsewhere. We don’t want any of that to be blamed on the environmental movement. The message needs to be “look how unreliable fossil fuels are, let’s build renewables”, not “the economy is collapsing and people are freezing in their apartments because those green freaks didn’t let us drill”.

It’s Time To Get Serious About Saving The Planet

The lengthy process to get things built in the US is seriously impeding our ability to reduce emissions. That includes, though is certainly not limited to, laws like NEPA (which requires an environmental impact statement for federal actions with major environmental impact).

It’s possible to streamline these barriers. For instance, a bill requiring “instant, online solar permitting” for residential projects in most of California just passed the legislature and awaits Gov. Newsom’s signature.

We do need to be careful not to throw the baby out with the bathwater. The mere fact that a project yields carbon-free power or otherwise reduces greenhouse emissions doesn’t trump all other considerations; some project sites are legitimately inappropriate, and most projects probably need to be adapted in one way or another to mitigate impacts on the surrounding community. The tendency to site projects in disadvantaged communities won’t end when we go green, even if living next door to a wind farm is preferable to an asthma-inducing power plant. We need checks and balances to ensure that these considerations are addressed. But we need to do that in a way that doesn’t add years of delay and uncertainty to each project.

For a long time, it was reasonable to say that the chief barriers to mitigating emissions were scientific, technical, or economic. We needed more research, more pilot projects, and more funding. But increasingly, the research has been done, the pilots have been successful, and – especially now that the IRA has passed – the funding is there. Now, our big challenge is to build. Five-year review review cycles are not a law of nature, they are a choice. It’s time we make a different choice.