If you read the article, this is not a tech from Google. They just bought a small geothermal plant for their datacenter from a startup ([1]). As long as the plant actually works (very much likely), it will be operational for many-many years.
3.5MW is a start but a fraction of 1 datacenter's load. I read that California's Long Valley has more than adequate geothermal resources to power the entire state, so I wonder if locating datacenters in Nevada or eastern California will become popular. There are already generating facilities there, but only a few hundred megawatts in total.
That's a good idea, assuming that these datacenters would use a close-loop water cooling, as opposed to evaporation techniques, which waste enormous amount of water. See the previous discussion about datacenters in dry areas: https://news.ycombinator.com/item?id=37732550
1/2 of the energy we consume ends up as heat. 1/2 of that is for industrial use.
In our glorious renewable energy future, heat from data centers will be harnessed to recharge thermal batteries, run district heat (homes, buildings, pools), supply industrial heat, etc.
I'm trying to say that waste heat could be used downstream, eg for heating homes or generating steam. Since heat is the end use of 1/2 of energy, this kind of conservation is low hanging fruit.
IIRC, Google, Microsoft, others have already committed to pilot projects for capturing heat from their data centers. Transforming an expense into a revenue source. Woot.
That'll be a chicken / egg problem tbh, but I've seen windmill park projects get co-funded by local governments and capacity bought up by the likes of Microsoft for their datacenters ashore.
That is: Local governments should set up infrastructure for datacenters and other energy intensive industry, developers should make plans to build geothermal plants, and big tech companies should help fund the construction of these plants.
That said, the big tech companies have billions just sitting around, they can create their own energy companies and power plants.
I think the geothermal people would have built more by now if they weren't blocked by CEQA/NEPA etc. It took them decades to build generators near Mammoth Lakes, just to grind through all the lawsuits.
>I wonder if locating datacenters in Nevada or Eastern California will become popular.
I believe that Fervo (the company Google partnered with) has open roles in Nevada, so it may be something that they're considering. Nevada also has the benefit of having cheap land once you move outside of places like Reno and Vegas because the majority of the state is arid/desert.
Horizontal drilling with a closed loop system sounds a lot like Eavor's system (https://www.eavor.com/). Eavor was a hot company here in Alberta a while ago but I haven't heard anything since.
I wonder what the cost is to generate 3.5MW vs other forms of energy production? Also, how would this scale, as Elon says, a prototype is easy, production scale is very hard.
Note that Fervo Energy (the geothermal start-up that Google partnered with) is founded by an ex oil/gas drilling engineer who decided to take all the lessons learned by oil/gas of horizontal drilling and hydraulic fracturing over the last 20 years and apply to geothermal.
Turns out all the assumptions for the economics of geothermal were based on 20+ year old data. With the advances applied from oil/gas, geothermal is squarely back in the economic region of renewables.
It took them 72 days to drill their injection well to 11220 ft. 59 days to drill their production well to same depth. So from their 1st well to their 2nd, they've already improved 18%. Oil/gas drillers can do 11220 ft wells within a couple days.
Sure 3.5 MW isn't a ton but this was literally a proof of concept. They can expand by orders of magnitude by drilling hundreds of these injector/producer doublets. They could probably get 2-3x more power per well after optimization. They can easily control production by altering the injection rate. I forget the exact delay but basically, if they ramp up/down injection, they see the same exact response on the production side a couple hours later. They expect to be able to combine this with solar/wind to get a firm 24 hour renewable base load.
Note that they do monitor induced sesmicity and have green/yellow/red light system for when to stop if necessary. Again, this was for the proof of concept with the Department of Energy.
Edit: with the exception of the high temps (376F max recorded), the drilling is all very standard oil/gas style that happens all the time across the US. 376F isn't unheard of in deepwater offshore but it's damn hot.
Thank you for this, lots to be optimistic about! Forgive my ignorance, but...
> They expect to be able to combine this with solar/wind to get a firm 24 hour renewable base load.
Why would they need solar/wind to get a renewable base load? I know solar/wind can't provide that base load by themselves (at least not in one region), but I expected that geothermal would basically be constant, or at least fluctuate by only small percentages throughout a day? Does solar/wind really bring anything to the table in terms of base load here, or is it just that geothermal would provide the base load and solar/wind could be used for seasonal usage on top?
Geothermal heat reservoirs are renewable but not unlimited: heat is taken out of the rock by the working fluid and results in "thermal drawdown" until heat is conducted from adjacent areas. I would assume this is worse for hot dry rock enhanced geothermal systems like this one, where you don't have hydrothermal circulation moving hot water through the reservoir. Just for some graphs: https://www.osti.gov/servlets/purl/888907 . So perhaps only running your geothermal plant at night extends the lifespan of your well.
Base load is an obsolete concept that nobody needs at all. It applies to some coal and nuclear power plants that take days to startup and cannot throttle back. You accepted this because they are very cheap to run.
Renewables have their issues, but they are different than base load.
Base load is decidedly not an obsolete concept. Renewables go to zero, and building sufficient peaking capacity to overcome a total loss of grid power would be preposterously expensive. Renewables spread over a large enough geography could, hypothetically, overcome variability, but nowhere in the world has developed the grid capacity (or even renewable generation capacity) to substantiate that strategy, and the odds of the US developing that kind of grid capacity are nil. There will always be a need for substantial base load generation.
What you described is NOT what baseload is about though.
Once again, baseload refers to power plants that deliver exactly the same power 24x7, but only after a couple week startup process. That is not what renewables deliver. It is in fact worthless with renewables because when renewables are at their peak time of day they already provide 100% of power and baseload forces more power on top of that.
Baseload is a useless concept except in the context of coal powerplants that don't exist. Anything modern either follows the load without problems or has an unpredictable nature that means you don't care about averages.
IIRC there have been some experiments with using geothermal plants as peaker plants since you can pump more water down and cap the output so you can ramp output up quickly later. This could work well with the variability of renewables to help balance out the grid, a role traditionally held by coal / gas.
> Why would they need solar/wind to get a renewable base load? I know solar/wind can't provide that base load by themselves (at least not in one region), but I expected that geothermal would basically be constant, or at least fluctuate by only small percentages throughout a day? Does solar/wind really bring anything to the table in terms of base load here, or is it just that geothermal would provide the base load and solar/wind could be used for seasonal usage on top?
Think of geothermal as being a drop-in replacement for natural gas turbines today. On wind/solar heavy grids, natural gas turbines spin up when wind/solar go down. The most reliable returns on investment in geothermal will come from targeting gas turbines, because that will result in maximizing the utilization of the geothermal system. Building geothermal capacity to meet variable demand is very expensive. Relatively to that, building solar/wind is not.
Utility solar and onshore wind on the other hand are incredibly cheap, especially when you don't have to store the energy. At this moment (~09:45 PDT on November 28, 2023), solar and wind are providing 61% of electric demand in California, with no storage involved [1].
In the case that geothermal starts taking over baseload from natural gas and nuclear, wind and solar will continue their downward price trend and still be favored when their supply and demand are time-correlated, and will also be used for very dispatchable end-uses (i.e. EV charging, water heating).
They will also be coupled with storage to provide demand-following peaker resources (since batteries can respond extremely quickly to demand spikes).
If geothermal gets so cheap that there is little difference in price between it and solar/wind+storage, then it will probably indeed take over a lot of current wind/solar demand, but ever increasing demand from variable loads will continue to soak up wind/solar supply, especially as the transition away from fossil fuels accelerates.
Finally, geothermal has much higher upfront/fixed costs (due to the necessity of drilling and the machinery for the turbine), so while better than nuclear, it's still harder to use for distributed generation than solar. Distributed generation (especially behind the meter) provides resilience to distribution outages.
The best thing about this technology is all the existing systems. Experienced workforce, an established manufacturing and supply chain, global projects. And it all comes from the oil and gas industry that will decline with green power, so there can be transition instead of a huge disruption.
And it's not just the inconvenience of disruption. The reason oil companies are putting up resistance against clean energy isn't that they hate the environment. It's that they love money. They don't want someone to declare their goose's golden eggs obsolete.
If someone came along and said "we can teach your golden goose to lay a different type of egg which is also golden", they might get behind that. And they might tell the politicians that they pay for to get behind it too.
Transitioning the entire world's energy infrastructure is still going to be very challenging and very expensive, but it would be less difficult if we could at least get all sides to agree that it's something we want to do.
Advanced geothermal energy (AGE) completes the puzzle. Wind, solar, others hit 80% of our needs. AGE fulfills the final 20%.
Additionally, by serving as heat reservior, AGE will maximize solar and wind. Instead of curtailing, Fervo will store their excess output as heat in the ground.
Can be located anywhere there's accessible heat. Even urban areas. Surface level operation is just a well head and a modest building.
Scales "horizontally". Meaning just drill more wells. Versus larger or more complex wells. This means Fervo has jumped onto the cost-learning curve (Wright's Law) by doing the same thing repeatedly. Instead of every project being bespoke.
All that R&D investment in the much hated fracking industry unlocked this achievement.
Are there concerns for earthquakes similar to how "fracking" used by the oil and gas industry has caused earthquakes? From the article:
>To generate geoethermal energy there, Fervo had to drill two horizontal wells through which it pumps water. Fervo pushes cold water through fractures in the rock, which heats it up so it can generate steam back at the surface.
Edit: Fervo energy is using the same exact techniques used by oil/gas for the hydraulic fracturing so yes the risk exists, but they didn't run into any issues. Seismic risk is typically highest near water disposal wells, which continuously inject water at high rates (>10000 bbl/d). These geothermal wells were fractured once, and whatever is injected is produced a few hours later (no net increase of volume after the initial frac job).
"Prior to starting the operations phase of this project, a preliminary screening assessment was performed to evaluate the potential risk of induced seismicity at the project site. We followed the guidelines outlined by the US Department of Energy’s Protocol for Addressing Induced Seismicity Associated with Enhanced Geothermal Systems (Majer et al. 2012). Based on our analysis, we developed and implemented a Traffic Light System (TLS) to mitigate the risk of potential induced seismicity associated with the project. The TLS system was in effect throughout all major phases of the project from July 2020 through June 2023, including drilling, completions, stimulation, and well testing. All observed seismicity fell within the Green TLS protocol, with no significant impact or incident throughout the full duration of the
project."
Someone correct me if this is wrong, but I believe human induced quakes can only have any serious power if they release tension that is already present.
Which means making a quake happen earlier than it would, not creating one in a seismically calm area.
This is great to see. Enhanced geothermal is a grossly underrated technology. We're dumping billions into fusion energy research and wind and solar subsidies but it seems like we're really close to having "geothermal anywhere" we just need a little more investment.
(1) If fusion goes our energy problems are pretty much resolved.
(2) Solar and wind are so cheap right now and will continue to get cheaper that we won't need oil and gas for electricity production except for specific use-cases. Combine that with decreasing energy storage costs and better grid management - its going well.
(3) Nuclear is in a renaissance and can supply base load as we ramp it back up.
Unfortunately geothermal, tidal and biofuels will be reserved to niche technologies. Too late, too inefficient or too expensive.
1. And if I win the lottery my financial needs are taken care of but I'm not planning on that
2. You are overestimating solar, wind, and energy storage and underestimating our energy needs, doubly so with transportation moving to electricity. Renewables still arent replacing fossil fuels despite all the hype. [1]. Moving off fossil fuels is going to require getting energy from whatever we can and having a constant consistent energy source is really important. People keep mentioning how cheap wind and solar is now but the true cost wont be realized until governments quit subsidizing them and start taxing [2].
I think it depends on what our goals are. If we are just trying to replace current electricity capacity with renewables, the I think solar/wind/etc. + batteries are fine and can be done relatively quickly (though keep in mind new coal plants are still being built!). If our goal is to completely replace all fossil fuels, then I think it will take a lot longer (I think like 100 years) and requires serious investment in battery production such as described in Tesla Master Plan Part 3. However if the goal is all of the above, but also start removing CO2 from the air, I think only something like a fusion breakthrough will make it work. We will need to generate like and order of magnitude more energy to do this. Also we have to realize that while renewables are growing, so are fossil fuels! See https://ourworldindata.org/grapher/electricity-prod-source-s... our total energy usage is increasing, so in order to replace all fossil fuels, you have to not only take up the new demand, you have to eat into the existing, almost like a double exponential! (not sure if the right term)
Regarding (1), it's not that simple. It depends on what form fusion takes. Fission was once thought to be the future of energy generation, unlimited clean energy for all, but now we realize the (justified) regulatory, maintenance, safety, and build costs are so incredible it is usually more economical to build gas/coal plants or solar/wind.
If we achieve fusion production in the form of modern tokamaks with insane regulatory, maintenance, and build costs similar or worse than fission, we will be in the same situation so the real goal is to achieve fusion without all of those added costs. Only then could fusion realize the result we all dream of and that seems unlikely for many decades.
The best hope we have atm is one of the fusion startups working on alternative methods of fusion has a breakthrough. Some have made claims of breakthroughs and also claim net energy gain in a few years, but as history has taught us, that's unlikely to occur.
As much as I want and hope for the unlimited, clean fusion energy we all dream of, the realist in me recognizes the challenges we face and that we probably still have a long way to go. In which case, it's probably good to put money in other efforts like geothermal and not put all our eggs in one basket.
Interesting. I wonder if there are any fracking related risks with ground water contamination (breaking up the rocks that separate naturally occurring zones).
57 comments
[ 0.19 ms ] story [ 144 ms ] thread1. https://fervoenergy.com/
The only guarantee this plant survives Google’s next manic-depressive episode is if it can manage to drive ad revenue or violate user data privacy.
[0] https://cloud.google.com/blog/products/infrastructure/learn-...
In our glorious renewable energy future, heat from data centers will be harnessed to recharge thermal batteries, run district heat (homes, buildings, pools), supply industrial heat, etc.
100% of the energy used by a datacenter becomes heat.
I'm trying to say that waste heat could be used downstream, eg for heating homes or generating steam. Since heat is the end use of 1/2 of energy, this kind of conservation is low hanging fruit.
IIRC, Google, Microsoft, others have already committed to pilot projects for capturing heat from their data centers. Transforming an expense into a revenue source. Woot.
That is: Local governments should set up infrastructure for datacenters and other energy intensive industry, developers should make plans to build geothermal plants, and big tech companies should help fund the construction of these plants.
That said, the big tech companies have billions just sitting around, they can create their own energy companies and power plants.
They do?
https://finance.yahoo.com/quote/MSFT/balance-sheet/
https://finance.yahoo.com/quote/GOOGL/balance-sheet/
https://finance.yahoo.com/quote/AAPL/balance-sheet/
Numbers pulled from quarterly (not annual) breakdown.
I believe that Fervo (the company Google partnered with) has open roles in Nevada, so it may be something that they're considering. Nevada also has the benefit of having cheap land once you move outside of places like Reno and Vegas because the majority of the state is arid/desert.
But you've gotta start somewhere, I suppose.
Turns out all the assumptions for the economics of geothermal were based on 20+ year old data. With the advances applied from oil/gas, geothermal is squarely back in the economic region of renewables.
It took them 72 days to drill their injection well to 11220 ft. 59 days to drill their production well to same depth. So from their 1st well to their 2nd, they've already improved 18%. Oil/gas drillers can do 11220 ft wells within a couple days.
Sure 3.5 MW isn't a ton but this was literally a proof of concept. They can expand by orders of magnitude by drilling hundreds of these injector/producer doublets. They could probably get 2-3x more power per well after optimization. They can easily control production by altering the injection rate. I forget the exact delay but basically, if they ramp up/down injection, they see the same exact response on the production side a couple hours later. They expect to be able to combine this with solar/wind to get a firm 24 hour renewable base load.
16 page Fervo white paper with tons of data (pressure, temperatures, geologic setting, frac plans, etc) - https://eartharxiv.org/repository/object/5704/download/11142...
Note that they do monitor induced sesmicity and have green/yellow/red light system for when to stop if necessary. Again, this was for the proof of concept with the Department of Energy.
Edit: with the exception of the high temps (376F max recorded), the drilling is all very standard oil/gas style that happens all the time across the US. 376F isn't unheard of in deepwater offshore but it's damn hot.
> They expect to be able to combine this with solar/wind to get a firm 24 hour renewable base load.
Why would they need solar/wind to get a renewable base load? I know solar/wind can't provide that base load by themselves (at least not in one region), but I expected that geothermal would basically be constant, or at least fluctuate by only small percentages throughout a day? Does solar/wind really bring anything to the table in terms of base load here, or is it just that geothermal would provide the base load and solar/wind could be used for seasonal usage on top?
edit: and when the sun isn't shining and wind isn't blowing
Renewables have their issues, but they are different than base load.
Once again, baseload refers to power plants that deliver exactly the same power 24x7, but only after a couple week startup process. That is not what renewables deliver. It is in fact worthless with renewables because when renewables are at their peak time of day they already provide 100% of power and baseload forces more power on top of that.
...Here's your baseload: https://energyeducation.ca/encyclopedia/Baseload_power#:~:te....
Also, I guess: what in your mind is the difference between what baseload is, and what baseload is about?
https://www.volts.wtf/p/enhanced-geothermal-power-is-finally
Think of geothermal as being a drop-in replacement for natural gas turbines today. On wind/solar heavy grids, natural gas turbines spin up when wind/solar go down. The most reliable returns on investment in geothermal will come from targeting gas turbines, because that will result in maximizing the utilization of the geothermal system. Building geothermal capacity to meet variable demand is very expensive. Relatively to that, building solar/wind is not.
Utility solar and onshore wind on the other hand are incredibly cheap, especially when you don't have to store the energy. At this moment (~09:45 PDT on November 28, 2023), solar and wind are providing 61% of electric demand in California, with no storage involved [1].
In the case that geothermal starts taking over baseload from natural gas and nuclear, wind and solar will continue their downward price trend and still be favored when their supply and demand are time-correlated, and will also be used for very dispatchable end-uses (i.e. EV charging, water heating).
They will also be coupled with storage to provide demand-following peaker resources (since batteries can respond extremely quickly to demand spikes).
If geothermal gets so cheap that there is little difference in price between it and solar/wind+storage, then it will probably indeed take over a lot of current wind/solar demand, but ever increasing demand from variable loads will continue to soak up wind/solar supply, especially as the transition away from fossil fuels accelerates.
Finally, geothermal has much higher upfront/fixed costs (due to the necessity of drilling and the machinery for the turbine), so while better than nuclear, it's still harder to use for distributed generation than solar. Distributed generation (especially behind the meter) provides resilience to distribution outages.
1. https://www.caiso.com/TodaysOutlook/Pages/supply.aspx
If someone came along and said "we can teach your golden goose to lay a different type of egg which is also golden", they might get behind that. And they might tell the politicians that they pay for to get behind it too.
Transitioning the entire world's energy infrastructure is still going to be very challenging and very expensive, but it would be less difficult if we could at least get all sides to agree that it's something we want to do.
Both audio and transcript:
"Enhanced geothermal power is finally a reality" [2023-07-21] https://www.volts.wtf/p/enhanced-geothermal-power-is-finally
My key takeaways:
Advanced geothermal energy (AGE) completes the puzzle. Wind, solar, others hit 80% of our needs. AGE fulfills the final 20%.
Additionally, by serving as heat reservior, AGE will maximize solar and wind. Instead of curtailing, Fervo will store their excess output as heat in the ground.
Can be located anywhere there's accessible heat. Even urban areas. Surface level operation is just a well head and a modest building.
Scales "horizontally". Meaning just drill more wells. Versus larger or more complex wells. This means Fervo has jumped onto the cost-learning curve (Wright's Law) by doing the same thing repeatedly. Instead of every project being bespoke.
All that R&D investment in the much hated fracking industry unlocked this achievement.
>To generate geoethermal energy there, Fervo had to drill two horizontal wells through which it pumps water. Fervo pushes cold water through fractures in the rock, which heats it up so it can generate steam back at the surface.
It sounds like it might be a possibility.
From the white paper (https://eartharxiv.org/repository/object/5704/download/11142...) -
"Prior to starting the operations phase of this project, a preliminary screening assessment was performed to evaluate the potential risk of induced seismicity at the project site. We followed the guidelines outlined by the US Department of Energy’s Protocol for Addressing Induced Seismicity Associated with Enhanced Geothermal Systems (Majer et al. 2012). Based on our analysis, we developed and implemented a Traffic Light System (TLS) to mitigate the risk of potential induced seismicity associated with the project. The TLS system was in effect throughout all major phases of the project from July 2020 through June 2023, including drilling, completions, stimulation, and well testing. All observed seismicity fell within the Green TLS protocol, with no significant impact or incident throughout the full duration of the project."
Which means making a quake happen earlier than it would, not creating one in a seismically calm area.
https://www.thelocal.de/20170818/this-historic-german-town-i...
Unfortunately geothermal, tidal and biofuels will be reserved to niche technologies. Too late, too inefficient or too expensive.
[1] https://www.eia.gov/energyexplained/us-energy-facts/ [2] https://www.reuters.com/business/energy/us-doubles-renewable...
> the true cost wont be realized until governments quit subsidizing them and start taxing
This is true for a lot of things, including for oil and gas if subsidized.
If we achieve fusion production in the form of modern tokamaks with insane regulatory, maintenance, and build costs similar or worse than fission, we will be in the same situation so the real goal is to achieve fusion without all of those added costs. Only then could fusion realize the result we all dream of and that seems unlikely for many decades.
The best hope we have atm is one of the fusion startups working on alternative methods of fusion has a breakthrough. Some have made claims of breakthroughs and also claim net energy gain in a few years, but as history has taught us, that's unlikely to occur.
As much as I want and hope for the unlimited, clean fusion energy we all dream of, the realist in me recognizes the challenges we face and that we probably still have a long way to go. In which case, it's probably good to put money in other efforts like geothermal and not put all our eggs in one basket.