I come bringing good news about hydrogen

The fourth big green energy technology

The is a techno-optimist blog, and it’s time for another techno-optimist post.

By now, everyone knows about the importance of solar and wind power. Some people still argue against these technologies, but the brute logic of falling costs has simply overwhelmed the naysayers. As for batteries, I recently wrote a big post about how they’re going to change everything — not just allowing storage for solar and wind power, but providing cheap energy portability for everything from drones to robots to appliances to bikes and cars.

Fundamentally, the reason that solar, wind and batteries are changing our world — and are set to change it even more — is that these technologies have followed learning curves. As we’ve built more of them, these things have become cheaper and cheaper. And the cheaper they’ve gotten, the more we’ve built.

If you spotted the learning curves for solar, wind, and batteries early on, you could make amazingly prophetic predictions about how cheap these things could get. Futurist author and investor Ramez Naam was not the first to notice the learning curves, but as far as I know he was the first to popularize the idea. In 2011 he wrote a blog post for Scientific American arguing that solar power was following an equivalent of Moore’s Law. Four years later he wrote a similar post about batteries, showing how they were also becoming cheaper exponentially as more and more were built.

Those who read Ramez’ posts and grasped their significance would be able to predict the solar and battery revolutions that we’re now seeing. Sadly — but also somewhat hilariously — most did not. Year after year for over a decade, the International Energy Agency predicted that solar capacity additions would flatten out immediately; instead, they kept accelerating.

Source: Visual Capitalist

But in fact, there’s a fourth crucial green energy technology that also appears to be following a learning curve: green hydrogen.

A lot of people roll their eyes when they see the word “hydrogen”, because there’s a popular idea that hydrogen is a failed technology. The reason people think this is that hydrogen cars, which some had predicted would replace gas cars, turned out not to be very viable. But the people who dismiss hydrogen are simply ignorant of all the other things hydrogen can do — and which we need it to do — besides powering cars.

I’ll talk about these uses in a second, but first a quick note about what “green hydrogen” means. This is not some sort of weaselly industry buzzword like “clean coal”. The classic way of producing hydrogen is from methane, or CH₄. This releases carbon — the C in CH₄ — into the atmosphere, which is bad. “Green hydrogen” means producing hydrogen by running electric current through water (H₂O) in order to break the chemical bonds and separate the H from the O. That’s called electrolysis, and the machine that does it is called an electrolyzer. Remember that in order to be “green”, the electricity that’s used to break apart the water should come from a green energy source such as solar or wind.

Anyway, so let’s talk about why hydrogen isn’t a failed technology, and there are lots of things we need it for.

The basic idea: Why hydrogen is interesting

There are basically four things we can use hydrogen for:

• Various chemical processes

• Heating

• Storing energy

• Moving energy around from place to place

In fact, we already use lots of hydrogen for the first of these items. The most important example is making ammonia, which is used for fertilizer, plastics, and lots of other stuff. Ammonia is NH3. We get the nitrogen from the air, and traditionally we get the hydrogen from burning methane — this is the famous Haber-Bosch process. But burning methane releases carbon. So in order to decarbonize our economy, we need to switch this dirty hydrogen to green hydrogen, electrolyzed from renewable sources.

But there are lots of other things we could use hydrogen for. One of these is heating. A lot of industrial applications require producing intense localized heat — to melt metals, break chemical bonds, fire ceramics, etc. Currently, that heat is often provided by burning fossil fuels (mostly methane), but we could conceivably burn hydrogen instead.

Third, hydrogen is a medium for storing energy over time. Energy from solar is intermittent — you need something to store the energy produced during the sunny times, in order to have electricity in the winter. Batteries aren’t great at storing energy for many months, so long-term storage needs other options.

Finally, most vehicles run on fossil fuels, and this needs to change. Hydrogen failed as a way of powering cars, but it might succeed for other vehicles like ships.

Notice that all of these uses have one thing in common — they’re all ways of replacing fossil fuels for things we already do. Tons of things in our economy rely on burning stuff; hydrogen is simply a flammable chemical that doesn’t release carbon when you burn it. This is why my optimism about green hydrogen is a lot more moderate and tempered than my optimism about batteries — whereas batteries enable us to do lots of things we can’t currently do at all, green hydrogen pretty much just replaces things we already know how to do.

But that doesn’t mean green hydrogen is simply a backstop technology — a way to sustain our industrial economy without cooking the planet. Solar power is getting so cheap that green hydrogen — electrolyzed by solar power — may allow us to make lots of other things cheaper than they ever were before. We could get cheaper fertilizer, cheaper chemical manufacturing, and so on. And that would make the world richer than it ever was before. I think of hydrogen as a way of translating cheap solar into a bunch of other cheap stuff.

What hydrogen isn’t good for — and what it might be good for

There’s still the question, however, of which of these industrial activities hydrogen will actually be practical for. We all remember the failure of hydrogen cars. Basically, electrolyzing water, compressing and chilling the hydrogen gas, and burning the hydrogen again iss simply not an efficient way of moving energy around. Though hydrogen cars beat electric cars (for now) on range and refueling time, they lose on pretty much every other conceivable metric, simply because of that fundamental inefficiency:

Source: Liebrich Associates

Hydrogen is also pretty bad for heating homes, for similar reasons. Electric-powered heat pumps are just much, much more efficient.

But there are lots of other things that hydrogen can probably do more competitively. The green energy consultant Michael Liebrich has created a very useful infographic of what applications we’re more and less likely to use hydrogen for. It’s called the “hydrogen ladder”:

Liebrich also has a long and very useful post where he explains why he put each of these applications where he did. You’ll notice that cars and domestic heating are very far toward the “uncompetitive” end of the spectrum, as we’d expect. Green hydrogen is not the solution there.

In fact, I’m willing to assume that hydrogen is basically useless for all building heating, and for all forms transportation except for ships. But decarbonizing shipping is pretty important! So that’s one thing to be optimistic about. Batteries aren’t very useful for ship propulsion, and the other main “green” alternative is biofuels. So hydrogen could be very useful there.

Another important use for green hydrogen is to replace dirty hydrogen in the production of fertilizer and other chemicals. As Liebrich notes, this is simply an unavoidable fact of decarbonization. In the short term, this is forecast to be the main source of demand for green hydrogen.

One additional area where hydrogen might help is industrial process heat. These processes account for a substantial amount of greenhouse gas emissions, and are very fossil-fuel-intensive. So this sector needs to be decarbonized.

Liebrich is surprisingly cautious about hydrogen’s prospects here, writing:

Low-and mid-temperature industrial heat - up to 160C and perhaps higher - will be much more cheaply and precisely supplied by heat pumps. Many people seem to think that high-temperature industrial heat has to be delivered by gas. That is simply not true, there are lots of ways of delivering high-temperature heat electrically. It's going to be a street fight for low carbon high-temperature heat between hydrogen and electricity, process by process, plant by plant[.]

Other experts generally agree. Hydrogen energy is less efficient than electric power, but it fits existing industrial infrastructure better (since we currently use methane, which is also a gas). Burning hydrogen can also reach a higher temperature than most other renewable heating technologies:

Source: Center on Global Energy Policy

So that could make hydrogen the solution in applications where you need higher heat.

The killer app is long-term energy storage

But of all the places where hydrogen could be important, the most promising by far is long-term energy storage. Solar power is the most important renewable energy production technology, but it’s inherently intermittent (and so is wind). This intermittency is a problem in both the short term — nighttime, rainstorms, etc. — and in the long term. Batteries can store energy well for a short period of time, and they’re getting cheaper and better every day.

But batteries are simply incapable of storing up solar energy for the long, dim winter months. Pumped hydro storage, the most popular currently existing technology, just doesn’t have much storage capacity in most places. Most proposed solutions, like compressed air, are still science fiction. Many experts agree that green hydrogen is emerging as the clear promising solution.

Hydrogen storage solves solar’s intermittency problem beautifully and elegantly. If there’s an electrolyzer and a hydrogen gas storage site next to a solar plant, then any time the sun is producing more electricity than we need, we can use the excess to make more hydrogen. Bit by bit, little by little, a solar plant will build up a hydrogen reserve to supplement it in the winter.

It’s a hell of a lot better of a use for excess solar power than mining Bitcoin!

Hydrogen-based seasonal energy storage — which would also provide resiliency in case of multi-day storms or blackouts — would allow us to reach the “last mile” of decarbonization. Combined with hourly storage from batteries, it would let us dispense with fossil fuel plants entirely.

The learning curve for electrolyzers

Anyway, this brings me to the reason I’m excited about green hydrogen technology. Like solar, wind, and batteries — and unlike many other technologies — hydrogen appears to be on a learning curve.

I learned about this fact from the climate writer David Roberts. Here’s an extremely useful podcast where Roberts interviews energy forecaster Doyne Farmer about learning curves:

Learning curves will lead to extremely cheap clean energy

Basically, some technologies have learning curves and others (such as fossil fuel extraction and nuclear power) seem not to have them. The technologies with learning curves eventually win out, because falling price and rising capacity reinforce each other.

Farmer has a new paper with Rupert Way, Matthew Ives, and Penny Mealy that looks at quantitative forecasts of the transition to renewable energy. They estimate learning curves for solar, wind, batteries, and electrolyzers. So far, the authors find, learning curves (which they call “Wright’s Law”) have predicted future renewables deployment far better than the more conservative methods used by the IEA or other forecasters. Here’s their picture for the cost of electrolyzers:

Source: Way et al. (2022)

If this curve holds, Way et al. expect hydrogen to play an important role in the energy transition, by filling in all the gaps where electric technology is infeasible and combustion is needed.

So should we expect the electrolyzer learning curve to hold? Just because solar and batteries got much cheaper than people predicted doesn’t mean every technology will do the same. In his interview with Roberts, Farmer argues that electrolysis — which is actually a pretty complex chemical process, at least if you want to do it efficiently — is conceptually similar to other chemical processes that have stayed on exponential learning curves for a very long time.

That’s no guarantee either — when it comes to technological progress, there are never any guarantees. But it’s reason for optimism. And experts who have looked into the details of the electrolyzer technologies themselves (there are multiple kinds) have concluded that there are substantial scaling effects to be taken advantage of. If you want to dive into the details here, I recommend this long but very readable report by the International Renewable Energy Agency.

More and more serious observers seem to be convinced of both green hydrogen’s importance and its promise. Here’s a great writeup from Bloomberg’s David R. Baker. Importantly, Baker notes that the rapid deployment of electrolyzers is no longer hypothetical:

In other words, it appears we’re looking at the fourth great green energy technology — the fourth energy solution that gets cheaper and cheaper the more of it you buy. As electrolyzers drop in cost, green hydrogen will fill in many of the nooks and crannies of decarbonization that solar, wind, and batteries can’t reach. And as solar gets ridiculously cheap, green hydrogen will be an important vehicle for transferring those cost savings to a whole lot of industries that need to burn something in order to operate.

Once again, human ingenuity is winning the day. That’s good news for all of us.

---