Hacker News Reader

Comment by lven | Hacker News Reader

his got me thinking about the earth's heat flux to the surface. It comes from primordial heat and radioisotope decay of thorium and uranium. Wild that geothermal energy is really just nuclear decay energy, just like the mars RTGs. But disappointing that it ain't all that much at just 47 TW, compared to the sun's 200,000 TW on the Earth's surface. And of those 47 TW, only 14 TW are on land. Not even close to enough to power mankind and our 20-40TW of primary energy today, and hopefully much much more in the future. Cool stuff, but we gotta think bigger.

View on HN

That 14 TW is just what leaks out naturally. But earth is an excellent insulator. Drilling down deep, all the energy extracted would be in addition to the 14 TW, and really as many TW as you could ever use.

The main problem with geothermal is that it is quite a bit more expensive than solar and wind, with an "opex" renewables don't have: regular overhauls of the steam turbines. So, a bet on geothermal would be a bet that storage built after there is enough renewable generation to charge it from will not also be cheap. Not a good bet: storage is, for the most part, just E = mgh and E = Fx, requiring only mid-20th-century engineering. (That is not to say more exotic methods won't also be used. Synthetic ammonia, good for fuel and fertilizer, is also a good storage medium. And, better battery chemistries surface all the time, with megafactories for certain of them already under construction.)

Just now, most storage schemes you read about have not had the professional engineering treatment, so are over-complex and expensive. When actually built out, much cheaper, simpler, and more reliable versions of those things will be used. (But Energy Vault's will not be. I predict they will pivot to something orbital soon, to string the saps along.)

Drilling for geothermal might get a lot cheaper, using the tech described in the article, but you still have the steam turbines. That new gadget that turns heat radiation directly into electricity with no turbine might eliminate that expense and make geothermal competitive, but it also might turn out too expensive to use.

Geothermal is decidedly limited. The deep-bore technology Homer-Dixon discusses is highly ambitious and would be tremendously expensive.

There are extensive geothermal power facilities, however, and a surprising amount of the readily-accessible generating capacity is already utilised throughout the world, the primary exception being in the African Rift Valley.

What Geothermal offers that wind and solar don't is high-reliability dispatchable peaking and night-time base-load power. Those loads are low but relatively predictable. There's a challenge with geo in extracting too much thermal energy too quickly, and fractional use on an as-needed basis might result in both longer-term utilisation per site (before the rock is cooled below practical usable levels) and more reliable generation capability for a given site.

A future low-carbon / renewables energy regime will be based on a mix of sources. Geothermal is an excellent option for that mix.

The drilling method described is distinguished from conventional methods by being much cheaper, besides being able to penetrate much deeper; and usable anywhere with stable geology, not just in a few select hot spots, as now.

So, if geothermal can get its operations cheaper, it could be good competition for wind and solar. Once plenty of storage capacity is on the grid, overproducing will not ever be a problem: the grid will absorb as much energy as any well can produce, provided the price is right.

This last is why nukes have no future: they cannot produce power at a price the market will pay.

I'll confess to being unfamiliar with the specifics of the technology. I'm also generally favourably inclined toward Homer-Dixon's work.

That said, the technical challenges, and costs, of drilling through many kilometers of rock are considerable. As are those of operating geothermal generation sites based on such bores.

The far less challenging Habenro project in Australia is a case study of early estimates being proving exceedingly optimistic:

I'd become aware of that specific project about a decade ago through ... a highly-positive puff-piece published in an overly-friendly Australian energy publication. The tone and fast-and-loose benefits descriptions set off a few alarms. The project ended up as a failure, the company running it pivoted out of geothermal energy.

https://old.reddit.com/r/dredmorbius/comments/1wpa90/how_not...

http://www.theregister.co.uk/2013/05/02/geodynamics_habanero...

I suspect most of the other links in my Reddit piece are dead, though archive copies may exist.

Energy Watch article archive: https://web.archive.org/web/20130805195650/http://www.energy...

ASX link seems dead.

Advertiser on SA Geothermal: https://web.archive.org/web/20130420153842/http://www.adelai...

AGEA: https://web.archive.org/web/20140126042126/http://www.agea.o...

All of that said:

- Geothermal does have long-established, well-proven, economical capability.

- I'm convinced humanity's energy future will largely be renewables.

- The ultra-deep plasma-torch borhole stuff might work. It's something of a longshot though. That said, I'd like to see related technology (e.g., the boring bits) and trial projects attempted.

Even the Habenero project wasn't an utter waste --- it was an experiment, just one that didn't pan out. It would have been far more worthwhile for it to have received coverage which portrayed it accurately, however.

Actually I don't think they will normally (maybe ever?) need to drill down tens of km, or even ten km, to get to enough heat.

I would like to see steam turbines displaced by something that doesn't need so much maintenance (high-pressure CO2?), so that geothermal could become more competitive with other renewables. But developing tech like that is no walk in the park. GE will do it when their market for steam turbines starts to fall off, faced with renewables.

>But disappointing that it ain't all that much at just 47 TW, compared to the sun's 200,000 TW

So 1 terrawatt = 47 terawatts

Reply on news.ycombinator.com