[cdeonier]

Cheap drilling would be a large boon for geothermal, considering the cost of surveying/exploring/drilling is > 50% of the cost of the development of a geothermal site.

I don't understand the articles goal of 300C target, though. While some types of geothermal plants do require temperatures that high, binary cycle power plants can use lower temperatures (130C) [1], which seems to open up more area for geothermal development since we expect most gradients between the surface and bottom of the crust to be ~2.5-3.1C / 100M. A lower temperature requirement would in turn allow you to drill less deep, which could consequently also decrease drilling costs.

Another thing the article doesn't mention: another interesting approach (aside from improving the technology, like drill bits) is with financing innovation. There have been / are government programs to de-risk the exploration/drilling cost by reimbursing the costs of drilling (80% for failed wells, for example) which also likely adds well data that could better characterize the underlying geothermal resources in regions (which would allow more accurate future development).

Really glad to see a deeper dive on geothermal though; its non-intermittency is a valuable characteristic separating it from other renewables that we're currently favoring (solar/wind). Because we generally break down energy generation to LCOE, it omits advantages like uptime of the renewable resource.

[1] https://www.energy.gov/eere/geothermal/electricity-generatio...

[Animats]

This is a good article.

The big breakthrough seems to be making drill bits out of a composite material formed from diamond and tungsten carbide.[1] One of their bits lasted through 25km of drilling. (Not one hole, re-used for multiple shallow holes.) That's encouraging. The geothermal people only need to go down 10km. Being able to do much of the job without backing out the drill string, one pipe section at a time, to change the bit is what seems to yield the cost estimates in the original article.

The next problem is to get everything at the down-hole end up to that level of reliability. Which is why the author talks about seal problems in mud-powered drilling motors. For the geothermal application, they just want to drill straight down, so they don't need all the fancy stuff used for slant and horizontal drilling.

So there remain some grungy, hard, and important problems to solve, like a seal material that will work better at high temperatures. Such things exist.[2]

This is encouraging.

The article points out that this isn't like hunting for oil and gas pockets; if you have roughly the correct overall geology, there will be hot rock down there anywhere you drill. This upsets some financial models, where drilling the first well in a new area is more like a VC-funded high-risk high return project. You're really drilling for the valuable info that oil or gas is there, not for the oil or gas from the exploration well. Deep geothermal is going to be dull, boring (literally), usually successful, and profitable over a long period but not in the short term. Great for regulated utilities.

[1] https://palmerbit.com/

[2] https://allsealsinc.com/HighTemperatureGaskets.html

[jillesvangurp]

The 300 degrees is needed to get enough steam pressure to drive a turbine. You need the temperature gradient basically to get that. It's all about efficiencies. A lower efficiency basically means you need to pump more water through, which means more drilling, which raises the cost.

Heat exchange pumps work with much lower temperature gradients which is great for heating a building or some water since you don't need to drill that deep. But it's not very efficient for generating electricity. There actually are some companies that can use heated water in your boiler as a battery and generate electricity from it but that is more from the point of view of using the energy you are storing anyway instead of letting it cool down. So a lower efficiency is acceptable for that.

The open question mark for geothermal is if the cost of drilling will ever be low enough to compete with solar and wind + batteries. Solar and wind are a lot cheaper per kwh but of course intermittent. There are various ways of fixing that that basically involve using some form of battery. You can think of geothermal as a battery where the fully charged battery simply is our planet. Nice if you can get to it but not necessarily cheap enough compared to other ways to store energy. Getting to it involves expensive drilling projects and operating a lot of plumbing to get energy out of it.

An example of a battery that is pretty cheap is a thermal mass based batteries. It is basically the same material (i.e. rocks) plus some insulator. Given enough mass, you can store quite large amounts of energy for very long and there are some companies starting to do exactly that. Several companies are working on those. It's all going to boil down to cost per kwh in the end. wind and solar converging on about a cent per kwh. Batteries tend to be more expensive but still cheaper than burning gas/coal. Geothermal sits somewhere in between. It could be cheaper in some places long term. But then batteries are also getting cheaper.

[briffle]

My old University powers, heats, and cools itself with Geothermal wells that are at 195F, not sure about the 300C either. (and clears snow/ice from sidewalks and outdoor staaircases) The college also sells extra power to the hospital next door. (it makes around 2MW with a binary cycle plant) https://urbanecologycmu.wordpress.com/2016/11/01/geothermal-...

[Accujack]

I think the 300C target is to support the article's assertion that geothermal could replace e.g. nuclear plants. Geothermal for heating can work well with a lower approach temp, but industrial processes/power generation needs a higher differential.

[cdeonier]

Could I get your help understanding your statement "industrial processes/power generation needs a higher differential"? Why does geothermal need a higher differential for power generation?

Power generation is already accomplished with lower heat cycles (e.g., binary plants mentioned earlier would probably use a rankine cycle to deal with the low heat), though we'd expect those power plants to have less nameplate capacity than something like a double flash-steam plant.

I think you're correct you'd get more efficiency with higher gradients, but I don't understand what's limiting about the lower temperatures. Is it economics?

[orolle]

The higher the differential the higher the efficiency of heat to electricity transformation. If I remember correctly it's a big efficiency gain between 200C and 300C. From economic side of things, more bang for the buck.