geothermal dreams




I've noticed a recent trend of advocacy for novel geothermal power designs as a complete solution for low-cost renewable energy.

There are an infinite number of flawed ideas, so I don't usually write about why specific ones won't work. Lessons on how to find flaws in ideas can certainly be valuable, but such analysis seems to even lose that utility without ephemeral context. But I'll try to generalize a bit.

The new advocates of geothermal power seem to be people who read something about human progress being about increased available energy, and it felt correct, so now that's Known and should be used.

They seem to all be current or former fans of nuclear power plants. Because they take technological progress as an axiom, the increased costs of nuclear power over time must, from their point of view, be caused by regulation. But of course, those cost increases haven't tracked regulations over time or across countries.

It seems the failure of nuclear power projects to actually deliver caused a displacement of that hope to geothermal power. Techno-optimism is the axiom, not the conclusion, so whatever is found from a search must be a solution.

What has supposedly changed about geothermal power? The answer seems to be "drilling by vaporizing rock with microwaves". There are multiple startups trying to do this now, including Quaise and AltaRock Energy.

Here's a talk on this by a researcher at MIT. You don't need to watch it.

This guy at MIT does stuff with microwaves, and he has some pictures of small slabs of rock with holes melted through them using microwaves. Well, that's nice, but I can put a hole in a thin slab of rock with a hammer. What happens when you have a 1km deep hole?

I shouldn't need to tell you the answer. Just imagine a deep hole in rock.

You heat up the bottom of the hole, and eventually you have a hole with molten rock in the bottom. This doesn't help with drilling. You could try to pump it up the hole, but it will melt, weaken, or dissolve your pipe. And a lot of energy gets lost from thermal conduction. And molten rock tends to be extremely viscous.

OK, so instead of trying to pump molten rock, you keep heating it until it vaporizes. This vapor starts going up the hole, and it condenses on the sides. It also condenses on your waveguide or whatever you have going down the hole, making it extremely hot.

I think the thought process of people who find that talk convincing doesn't get into that, and just goes: "MIT is doing this research and a startup working on this approach has been funded, so all the obvious problems must have been solved". (Well, I don't want to pit my credibility against MIT's here, but when I see a news article that starts with "MIT scientists have found a way to" I stop reading. Their PR department is out of control.)

The problem is, that approach doesn't work. You end up saying things like:

 • sure, this evidence doesn't seem to make sense, but the US government wouldn't just make up Iraq having a nuclear weapon program

 • what Theranos is trying to do doesn't seem possible now, and blood from a finger prick isn't even the same as blood from a vein, but I'm sure the investors and the big names on its board of directors would have checked that

 • Solyndra is making cylindrical solar cells, and sure, it seems pretty dumb to make solar cells that are mostly not facing the sun, especially since lower average light reduces their efficiency, but I'm sure the US government wouldn't give them a bunch of money if it was really that dumb an idea

 • sure, logically it seems like transmission patterns imply SARS-CoV-2 is airborne, and Taiwan and China are urgently trying to get masks, but surely the WHO and CDC wouldn't be saying it's only spread via surfaces if it's mainly airborne

 • US media is reporting that there's a scientific consensus that COVID-19 couldn't have originated from a laboratory, and surely it wouldn't just say there's a scientific consensus when there isn't

 • surely surgeons wouldn't recommend arthroscopic knee surgery if it didn't actually help

 • surely carpet companies wouldn't be putting fluorosurfactants in their carpets if it was actually hazardous

 • surely the FDA wouldn't approve aducanumab if it didn't actually work

 • surely media and scientists wouldn't have been saying that a very-low-fat diet is healthy without good evidence

 • surely there's a good reason for quinoa being considered healthier than rice with lentils

 • surely Japan wouldn't be putting even more money into hydrogen fuel work if it was a fundamentally flawed plan

 • surely the US government wouldn't spend billions on subsidizing ethanol fuel from corn if that had no chance of ever being economically viable

 • NASA has lots of smart people so surely they wouldn't spend billions on SLS and accomplish nothing


You get the idea.

Why is hydrogen fuel research such a perennial sink for billions of dollars from governments? The problems seemed obvious, so this puzzled me for a long time, but I think I understand now. Hydrogen fuel research is invested in because it has lots of papers and scientists saying it's worth researching, which in turn is because lots of money was spent on it. Scientists work on whatever they get grants for, the scientific grant system is rather broken, and everybody has to pretend that whatever their latest paper is about is actually useful somehow.

That cycle began, I suppose, because hydrogen fuel sounds futuristic and is relatively easy to make (commercially impractical) demonstrations for.

By the way, yes, polycrystalline diamond drill bits are great, but they're a smaller and more situational improvement over tricone bits than you might think, and diamond bits have been around for a while now. They certainly don't automatically make deep closed-loop geothermal economically competitive.

Anyway, there's currently nothing new that enables new approaches to geothermal that weren't previously viable. I should note that it is theoretically possible to drill and install pipe cheaply enough for closed-loop deep geothermal to be economically viable, but pushing forwards with an approach that can't work is counterproductive: once those projects fail, investors will take that as a sign to do something other than geothermal.

Now, I've talked to hardcore techno-optimists before. If you try to argue, it goes something like:
"Tech X is coming soon!"
"What enables Tech X now?"
"Method A makes X possible!"
"Here are 12 reasons Method A doesn't work, and also it's not a new idea at all."
"You're probably wrong somehow."
"OK, here are some people with credentials pointing out these problems."
"It doesn't matter if Method A doesn't work because there's Method B too."
"How many times am I going to have to go through this?"

It's not worth it. Just smile and nod. And that's certainly part of the problem: nobody's willing to tell Elon Musk why something like his Hyperloop paper is stupid, and the competent engineers have to play Kif Kroker to Zapp Brannigan executives. But I'm not satisfied with that as an answer.

China invested in silicon solar panel production, and now it's the largest producer by far. The US invested in failed companies like Solyndra and subsidies for rooftop solar, which is ~2.6x as expensive as utility-scale solar and never made economic sense.

The secret to cheap solar power turned out to be slicing thinner silicon wafers, but I've been told by some people involved that this sounded too boring.

China invested in HVDC lines, and now it has an HVDC grid. America invested in CO2 storage from coal plants, Gen IV nuclear reactor research, and hydrogen power, and got nothing.

China invested in electronics manufacturing and now Shenzhen produces the world's electronics. Taiwan invested in semiconductor production and has TSMC.

Risk is necessary, and some failure is inevitable, but high-risk ventures are only worthwhile if at least some of them succeed. Project failure rates have been increasing, and if current trends continue, America is in trouble. You can't run an economy on nothing but surveillance advertising and SEO and mortgages, and TSMC and Foxconn can survive without Apple better than vice-versa.


back to index