Baseload generation is useless in 2025. It's in the name; it's called "base load", not "base generation".
Base generation was a cost optimization. Planners noticed that load never dropped below a specific level, and that cheapest power was from a plant designed to run 100% of the time rather than one designed to turn on and off frequently. So they could reduce cost by building a mix of base and peaker generation plants.
In 2025, that's no longer the case. The cheapest power is solar & wind, which produces power intermittently. And the next cheapest power is dispatchable.
To take advantage of this cheap intermittent power, we need a way to provide power when the sun isn't shining and the wind isn't blowing. Which is provided by storage and/or peaker plants.
That's what we need. If added non-dispatchable power to that mix than we're displacing cheap solar/wind with more expensive mix, and still not eliminating the need for further storage/peaker plants.
If non-dispatchable power is significantly cheaper than storage and/or peaker power than it's useful in a modern grid. That's not the case in 2025. The next cheapest power is natural gas, and it's dispatchable. If you restrict to clean options, storage & geographical diversity is cheaper than other options. Batteries for short term storage and pumped hydro for long term storage.
It's nuclear fission. It's always been nuclear fission (well, at least since the '50s) and it will continue to be until we commercialize fusion reactors. Everything else is nice to have but it's like NIH syndrome.
Geothermal is fission, and wind, solar, and batteries are fusion at a distance. In both cases, the failure scenarios are benign vs traditional fission generation. It's fine to keep striving for fusion humans control, but the problem (global electrification and transition to low carbon generation) is already solved with the tech we have today. It took the world 68 years to achieve the first 1TW of solar PV. The next 1TW took 2 years. Globally, ~760GW of solar PV is deployed per year (as of this comment), and will at some point hit ~1TW/year of deployment between now and 2030.
Geothermal is a great fit for dispatchable power to replace coal and fossil gas today (where able); batteries are almost cheaper than the cost to ship them, but geothermal would also help solve for seasonal deltas in demand vs supply ("diurnal storage").
Failure scenario in modern fission reactors is also benign. Reactors are designed to lock down to prevent any leaks.
We deploy solar PV capacity, this doesn't mean we actually get that much power from the deployments. Nuclear fission provides reliable, baseload power, and doesn't require huge battery arrays to compensate for the sun setting or winds calming.
Things are more expensive when we keep reinventing the wheel and trying to do new things instead of just reusing proven designs. Remember that solar power also used to cost wheelbarrows of cash back in the day. When you do something repeatedly, it becomes less expensive over time.
Nuclear is actually the leader in waste management. No other energy source has as complete a story. Eg what happens to solar panels when they EOL in 25 years? They go into landfills and leach toxic chemicals into the ground. These chemicals, like lead and cadmium are toxic forever. They have no 'half-life' in which their toxicity reduces.
Solar panels are recycled at almost 100% of total materials. Redwood Materials (founded by Tesla's former CTO) has already established a supply chain to ingest and recycle EV and stationary storage batteries at scale. The problem is that the hardware is lasting longer than expected, and meaningful recycling volume does not yet exist.
Conversely, ~95,000 metric tons of nuclear waste in the US does not have permanent storage or recycling solutions, as of this comment, and there is no plan for long term storage or recycling. Nuclear generation is experiencing a negative learning curve; we keep spending more the more we attempt to build it.
(solar PV panels have a 25-30 year service life, at which point they will still produce power at ~80-85% initial rating, batteries have a 15-20 year service life, with sodium ion chemistries estimated to have up to 50 year service life assuming once daily cycling)
Enough renewables are deployed annually to replace the global nuclear fission fleet, year after year, even when accounting for capacity factor derating (to make a like for like comparison). The race is over, and renewables (with batteries) won. If you can find someone unsophisticated to invest in a fission reactor that takes billions of dollars and 10-15 years to build, more power to you. There will be no need for it by 2035-2040 when it prepares to send its first kwh to the grid.
(and to stay on topic for this thread, geothermal is a component of this when geothermal potential exists, cost is competitive, and dispatachability is a requirement to push out fossil generation in concert with renewables, hydro, legacy nuclear, battery storage discharge, and demand response)
> If you can find someone unsophisticated to invest in a fission reactor that takes billions of dollars and 10-15 years to build
Unsophisticated investors like the Chinese government? 'Nearly every Chinese nuclear project that has entered service since 2010 has achieved construction in 7 years or less.'
They dabble in nuclear, but it is not their focus. China can do what the developed world cannot because they are a command economy with less expensive labor, which will only last for a bit longer due to their structural demographics. Unless the developed world no longer has labor regulations, developed world wages, and capital based allocation systems, my statement stands with regards to investment. If capital and labor does not matter, certainly, anything is possible (Paraoh demanding pyramids, for example).
Your citation comes from an organization with biased intentions.
China built more solar power in the last 8 months than all the nuclear power built in the entire world in the entire history of human civilisation. And even if you adjust for utilisation rate to compare against nuclear utilisation China built more solar power generated per hour than all the nuclear power currently in operation generate in an hour - and did so in 12-18 months - https://bsky.app/profile/climatenews.bsky.social/post/3lggqu... - January 23, 2025
Ive been very pro nuclear my whole life, but a part of me is disheartened by the mega projects that commercial fission deployments have become (even if the reasons are bad) that’s a problem that nerfs traditional fission. If nuclear remains both political, extremely bureaucratic and requires public investment, it just won’t be the solution, and not because the tech or physics is bad, but the decision makers & investors can no longer organize large infrastructure projects effectively (except maybe China). This is not unique to nuclear.
Having smaller scale local power generation, whether it’s SMRs, solar, wind or geothermal, there’s a huge advantage in terms of economy, investment, and politics.
Nuclear has broad bipartisan support, and the Trump administration is heavily into it, so I wouldn't count it out just yet. If the various Green parties of the Western world ever come into power though, we are cooked.
I've always been curious why a cost-effective widespread implementation of geothermal energy has never been considered a holy grail of energy production, at least not in the public debate. Much of the discussion is so focussed on nuclear fusion, which seems so much harder and less likely to be reliable.
Drilling is one of those things which used to be extremely expensive but has very gradually come down in price. Thanks, ironically, to the oil industry. It's unsexy because there's no "silver bullet" waiting in the wings.
It's also quite hard to find suitably hot rocks suitably close to the surface.
Focusing on fusion .. I think that's a legacy of 60s SF, when the fission revolution was still promising "energy too cheap to meter".
To be fair, that promise of fission made sense from a purely scientific and mathematical perspective, before running into the practical realities of how its externalities would interact with real-world politics. Fission is expensive because in the real world it turns out we care quite a lot about proper waste management, non-proliferation, and preventing meltdowns.
In a world where anyone could just YOLO any reactor into production with minimal red tape, consequences be damned, fission energy would actually be extremely cheap. Hence the optimism around fusion. The promise of fusion is an actualization of last century's idealistic conception of fission. It can be a silver bullet for all intents and purposes, at least once it's established with a mature supply chain.
Many others here have talked about the difficulties of geothermal, which doesn't really get to the heart of my question: why the lack of hype around breaking down those difficulties? I appreciate that you took the time to comment on why it isn't so sexy, the SF argument probably has a lot to do with it.
The problems are that rock isn't a good conductor of heat, so once you've cooled a bit down, you have to wait for it to warm up. Warming only happens very slowly at the rate of < 50mW / m² which limits the amount of power you can get out.
The worst earthquake that was induced that way was 3.5, but given that one of the quakes happened in an area that had a catastrophic earthquake in the Middle Ages, some caution might be warranted: https://en.wikipedia.org/wiki/1356_Basel_earthquake
Because unless you sit on top of a volcano, amount of renewable geothermal energy is minuscule. In most places on Earth it's somewhere around 40 mW/m2 (i.e. accounting for conversion losses you need to capture heat from ~500 m2 to renewably power one LED light bulb!). In other words, in most places geothermal plant acts more like a limited battery powered by hot rock, so unless drilling is extremely cheap, it does not make economic sense compared to other energy sources.
> In most places on Earth it's somewhere around 40 mW/m2 (i.e. accounting for conversion losses you need to capture heat from ~500 m2 to renewably power one LED light bulb!)
Ground-source heat pumps extract about 1000 times more power from ground loops, where does the difference come from?
I think it mainly depends on how easy it is to access that energy. I went to Tuscany last year and to my surprise there were geothermal plants everywhere. I have never heard about these plants beforehand, but here they are in Italy quietly powering the countryside and heating greenhouses to grow basil all year around.
Probably because not everywhere on earth has the same easy access that Iceland has. The article mentions this:
> There aren’t gates of Hell just anywhere. A kilometre below ground in Kamchatka is considerably hotter than a kilometre below ground in Kansas. There is also readily accessible geothermal energy in Kenya (where it provides almost fifty per cent of the country’s energy), New Zealand (about twenty per cent), and the Philippines (about fifteen per cent)—all volcanic areas along tectonic rifts. But in less Hadean landscapes the costs and uncertainties of drilling deep in search of sufficient heat have curtailed development.
There is a crazy amount of energy available everywhere but it is not in the interest of the very powerful very wealthy existing players. This isn't some grand CONSPIRACY. For example oil companies may construct energy investment portfolios that would quite sensibly acquire promising energy related research. They do a simple cost benefit analysis then chose to modestly further research it or shelve it. They turn it into valuable pieces of paper that accumulate value over time. What is there for them not to like about it?
I like how David Hamel put it: We live in this thin sliver on the surface of the planet where it is reasonably peaceful. This is the tranquility! It's a good thing! If you go up or down by a mere few miles there is so much energy it kills you.
Instead of drilling deep, there is also an intersting case for storing cheap solar energy as hat in piles of dirts in the summer to power turbines in the winter: https://austinvernon.site/blog/standardthermal.html
People across the road from have geothermal, driven by a 1.5m-deep pond right near their house. Their heat never costs more than $100 a month in the winter.
That's a different "geothermal" - the correct name is "ground source heat pump" or in your neighbor's case, a pond-source heat pump. Those exploit the temperature stability that occurs some small numbers or meters subsurface for heating in the winter and cooling in the summer.
"Geothermal energy" involves drilling down to hot rock to tap intense heat to run a turbine that produces electricity.
It always has been. Our problem is switching over existing infrastructure without asinine complainers ruining the revolution. We can't even declare total victory with LED bulbs over incandescent. The war to have solar plants over more coal is falling back to coal thanks mostly to AI. Pushback on geothermal will arrive, I guarantee it.
> We can't even declare total victory with LED bulbs over incandescent.
The LED bulbs I have access to (whatever's in the aisles at Home Depot, Costco, etc.) fail much more frequently than the incandescent bulbs I used to buy, and produce an uglier light that is less warm even on the softest/warmest color settings.
My suspicion is that incandescents were at the "end" of their product lifecycle (high quality available for cheap) and LEDs are nearing the middle (medium quality available for cheap), and that I should buy more expensive LED bulbs, but I still think that there are valid "complaints" against the state of widespread LED lighting. I hope these complaints become invalid within a decade, but for now I still miss the experience of buildings lit by incandescent light.
The other thing with AI--the LED revolution was led on this idea that we all need to work as hard as we can to save energy, but now apparently with AI that's no longer the case, and while I understand that this is just due to which political cabals have control of the regulatory machinery at any given time, it's still frustrating.
The light color they call "daytime" is around 5000K, so I expected it to look like being outside in the sun; but instead I got a cold blueish vibe. The problem? Not enough power! I got the equivalent of a moonlit room.
So I got this 180W LED lamp (that's actual 180W, not 180W equivalent) [1]. It's so bright I couldn't see for 5 minutes. I put two in my office on desk lamps. The room now looks like being outside, without the "ugly blue" tint, even though the product says it's 6000K. The days of my SAD suffering are over!
Maybe buy your bulbs somewhere else? I'm yet to replace any of the LED bulbs I've bought over the past 15 years and honestly can't even remember the last time a bulb failed.
Actually, since posting this I've vaguely remembered a previous discussion on here about differences between LED bulbs sold in the US and those sold in UK/EU so maybe that explains it.
> The LED bulbs I have access to (whatever's in the aisles at Home Depot, Costco, etc.) fail much more frequently than the incandescent bulbs I used to buy, and produce an uglier light that is less warm even on the softest/warmest color settings.
LED lamps work just fine, you just need to pay more attention when you’re buying them. Philips makes decent LED lamps.
Make sure you’re buying lamps with 90+ CRI, that will help with the quality of light. 2700K is a good color temp for indoor living room/dining room/bedroom lighting, 3500-4000K for kitchen/garage/task lighting.
You also need to buy special lamps if you put them in an enclosed fixture, look for ‘enclosed fixture’ rated lamps. Regular LED lamps will overheat in an enclosed fixture.
There is an enormous push to build and power data centers in the DC / Northern Virginia region, and there's legislation in West Virginia right now requiring all coal-fired power plants to operate at at least 69% capacity at all times to support it.
> “West Virginia has numerous coal plants that have powered this country for decades. We need these plants to remain operational,” [WV Governor] Morrisey said. “… We will never turn our backs on our existing coal plants and we will work with the federal government to pursue new coal-fired generation.”
The only way new coal plants get built from today on is with massive lifetime subsidies, because they are uncompetitive. Ie, if they get built it’s for dumb politics not economics
> The war to have solar plants over more coal is falling back to coal thanks mostly to AI.
Also, due to solar not panning out at scale.[1]
More seriously, coal is just cheaper and, with incentives being removed for green energy, it's the cheapest and fastest option to deploy. It's dead simple and well understood reliable power.
I see yow it can read that way but it isn't what I said. Coal plants exist, either shuttered or running low loads due to financial incentives (not favoring them).
Studies show solar is cheaper but businesses continue to choose coal. I think the entity who's entire existence depends on them making the correct financial choice is a much better indicator of economic reality than a study made by people who have zero stake (at best) in the game.
I'm all for green energy but I also don't think people are stupid.
The example you chose is of a mirror based Solar system, which yes, is an obsolete technology.
Direct solar continues to be installed at greater amounts every year and coal is economically uncompetitive with basic anything (which is why it is collapsing), and especially against natural gas.
You're exactly right and it raises a question for me. Why do energy generation topics bring people out of the woodworks who cite some very idiosycratic one-off and use it to make out-of-proportion declarations about the utility of a given technology? This is the second one I've seen suggesting solar is doomed when they mean mirrors.
On twitter I saw someone claim PV is useless for heat because non-PV solar water heating is just so much more efficient. Not even true (I think it's a approximately a wash, different advantages in different applications), but very strangely in the weeds on a specific topic. Much too narrow a factual context to substantiate general level claims about solar as an energy writ large.
I think for whatever reason the missing the forest for the trees trap is really potent in energy discussions.
at some point we will figure out that because we took some much energy out of earths core that it stops spinning and causes the magnetic field to collapse ;-)
Not really how that works. Also earths core is being heated from nuclear decay and tidal effects. It’s getting 10’s or TW worth of heat until the sun expands and eats the earth. https://en.wikipedia.org/wiki/Earth's_internal_heat_budget
Baseload generation is useless in 2025. It's in the name; it's called "base load", not "base generation".
Base generation was a cost optimization. Planners noticed that load never dropped below a specific level, and that cheapest power was from a plant designed to run 100% of the time rather than one designed to turn on and off frequently. So they could reduce cost by building a mix of base and peaker generation plants.
In 2025, that's no longer the case. The cheapest power is solar & wind, which produces power intermittently. And the next cheapest power is dispatchable.
To take advantage of this cheap intermittent power, we need a way to provide power when the sun isn't shining and the wind isn't blowing. Which is provided by storage and/or peaker plants.
That's what we need. If added non-dispatchable power to that mix than we're displacing cheap solar/wind with more expensive mix, and still not eliminating the need for further storage/peaker plants.
If non-dispatchable power is significantly cheaper than storage and/or peaker power than it's useful in a modern grid. That's not the case in 2025. The next cheapest power is natural gas, and it's dispatchable. If you restrict to clean options, storage & geographical diversity is cheaper than other options. Batteries for short term storage and pumped hydro for long term storage.
It's nuclear fission. It's always been nuclear fission (well, at least since the '50s) and it will continue to be until we commercialize fusion reactors. Everything else is nice to have but it's like NIH syndrome.
Geothermal is fission, and wind, solar, and batteries are fusion at a distance. In both cases, the failure scenarios are benign vs traditional fission generation. It's fine to keep striving for fusion humans control, but the problem (global electrification and transition to low carbon generation) is already solved with the tech we have today. It took the world 68 years to achieve the first 1TW of solar PV. The next 1TW took 2 years. Globally, ~760GW of solar PV is deployed per year (as of this comment), and will at some point hit ~1TW/year of deployment between now and 2030.
Geothermal is a great fit for dispatchable power to replace coal and fossil gas today (where able); batteries are almost cheaper than the cost to ship them, but geothermal would also help solve for seasonal deltas in demand vs supply ("diurnal storage").
https://reneweconomy.com.au/it-took-68-years-for-the-world-t...
https://ember-energy.org/data/2030-global-renewable-target-t...
I also love geothermal for district heating in latitudes that call for it; flooded legacy mines appear to be a potential solution for that use case.
Flooded UK coalmines could provide low-carbon cheap heat 'for generations' - https://news.ycombinator.com/item?id=45860049 - November 2025
Failure scenario in modern fission reactors is also benign. Reactors are designed to lock down to prevent any leaks.
We deploy solar PV capacity, this doesn't mean we actually get that much power from the deployments. Nuclear fission provides reliable, baseload power, and doesn't require huge battery arrays to compensate for the sun setting or winds calming.
Nuclear is great, but it does require wheelbarrows of cash, and we don’t have a solution for waste products.
Things are more expensive when we keep reinventing the wheel and trying to do new things instead of just reusing proven designs. Remember that solar power also used to cost wheelbarrows of cash back in the day. When you do something repeatedly, it becomes less expensive over time.
Nuclear is actually the leader in waste management. No other energy source has as complete a story. Eg what happens to solar panels when they EOL in 25 years? They go into landfills and leach toxic chemicals into the ground. These chemicals, like lead and cadmium are toxic forever. They have no 'half-life' in which their toxicity reduces.
Solar panels are recycled at almost 100% of total materials. Redwood Materials (founded by Tesla's former CTO) has already established a supply chain to ingest and recycle EV and stationary storage batteries at scale. The problem is that the hardware is lasting longer than expected, and meaningful recycling volume does not yet exist.
Conversely, ~95,000 metric tons of nuclear waste in the US does not have permanent storage or recycling solutions, as of this comment, and there is no plan for long term storage or recycling. Nuclear generation is experiencing a negative learning curve; we keep spending more the more we attempt to build it.
(solar PV panels have a 25-30 year service life, at which point they will still produce power at ~80-85% initial rating, batteries have a 15-20 year service life, with sodium ion chemistries estimated to have up to 50 year service life assuming once daily cycling)
https://www.epa.gov/hw/solar-panel-recycling
https://www.energy.gov/eere/solar/articles/beyond-recycling-...
https://e360.yale.edu/features/solar-energy-panels-recycling
https://www.cnbc.com/2025/11/09/nuclear-power-energy-radioac...
https://www.gao.gov/nuclear-waste-disposal
https://decarbonization.visualcapitalist.com/visualizing-all...
(nuclear power accounts for about 10% of electricity generation globally, as of this comment)
Enough renewables are deployed annually to replace the global nuclear fission fleet, year after year, even when accounting for capacity factor derating (to make a like for like comparison). The race is over, and renewables (with batteries) won. If you can find someone unsophisticated to invest in a fission reactor that takes billions of dollars and 10-15 years to build, more power to you. There will be no need for it by 2035-2040 when it prepares to send its first kwh to the grid.
(and to stay on topic for this thread, geothermal is a component of this when geothermal potential exists, cost is competitive, and dispatachability is a requirement to push out fossil generation in concert with renewables, hydro, legacy nuclear, battery storage discharge, and demand response)
https://www.google.com/search?q=baseload+is+a+myth
https://cleantechnica.com/2025/11/15/coal-killing-sodium-ion...
https://ember-energy.org/latest-insights/q3-global-power-rep...
https://ember-energy.org/latest-insights/solar-electricity-e...
https://ember-energy.org/latest-insights/solar-electricity-e...
https://world-nuclear.org/information-library/economic-aspec...
https://www.lazard.com/research-insights/levelized-cost-of-e...
https://ourworldindata.org/grapher/solar-pv-prices
https://ourworldindata.org/battery-price-decline
https://ourworldindata.org/data-insights/solar-panel-prices-...
https://news.ycombinator.com/item?id=44513185 (lfp battery storage cost citation in 2025)
> If you can find someone unsophisticated to invest in a fission reactor that takes billions of dollars and 10-15 years to build
Unsophisticated investors like the Chinese government? 'Nearly every Chinese nuclear project that has entered service since 2010 has achieved construction in 7 years or less.'
https://thebreakthrough.org/issues/energy/chinas-impressive-...
They dabble in nuclear, but it is not their focus. China can do what the developed world cannot because they are a command economy with less expensive labor, which will only last for a bit longer due to their structural demographics. Unless the developed world no longer has labor regulations, developed world wages, and capital based allocation systems, my statement stands with regards to investment. If capital and labor does not matter, certainly, anything is possible (Paraoh demanding pyramids, for example).
Your citation comes from an organization with biased intentions.
https://en.wikipedia.org/wiki/Breakthrough_Institute
Can China Break Nuclear Power’s Cost Curse—and What Can the US Learn? - https://rooseveltinstitute.org/blog/can-china-break-nuclear-... - September 17th, 2025
China built more solar power in the last 8 months than all the nuclear power built in the entire world in the entire history of human civilisation. And even if you adjust for utilisation rate to compare against nuclear utilisation China built more solar power generated per hour than all the nuclear power currently in operation generate in an hour - and did so in 12-18 months - https://bsky.app/profile/climatenews.bsky.social/post/3lggqu... - January 23, 2025
China is installing the wind and solar equivalent of five large nuclear power stations per week - https://www.abc.net.au/news/science/2024-07-16/chinas-renewa... - July 15th, 2024
Nuclear Continues To Lag Far Behind Renewables In China Deployments - https://cleantechnica.com/2024/01/12/nuclear-continues-to-la... - January 12th, 2024
Nuclear Energy & Free Market Capitalism Aren’t Compatible - https://cleantechnica.com/2023/11/06/nuclear-energy-free-mar... - November 6th, 2023
Ive been very pro nuclear my whole life, but a part of me is disheartened by the mega projects that commercial fission deployments have become (even if the reasons are bad) that’s a problem that nerfs traditional fission. If nuclear remains both political, extremely bureaucratic and requires public investment, it just won’t be the solution, and not because the tech or physics is bad, but the decision makers & investors can no longer organize large infrastructure projects effectively (except maybe China). This is not unique to nuclear.
Having smaller scale local power generation, whether it’s SMRs, solar, wind or geothermal, there’s a huge advantage in terms of economy, investment, and politics.
Nuclear has broad bipartisan support, and the Trump administration is heavily into it, so I wouldn't count it out just yet. If the various Green parties of the Western world ever come into power though, we are cooked.
I've always been curious why a cost-effective widespread implementation of geothermal energy has never been considered a holy grail of energy production, at least not in the public debate. Much of the discussion is so focussed on nuclear fusion, which seems so much harder and less likely to be reliable.
Drilling is one of those things which used to be extremely expensive but has very gradually come down in price. Thanks, ironically, to the oil industry. It's unsexy because there's no "silver bullet" waiting in the wings.
It's also quite hard to find suitably hot rocks suitably close to the surface.
Focusing on fusion .. I think that's a legacy of 60s SF, when the fission revolution was still promising "energy too cheap to meter".
To be fair, that promise of fission made sense from a purely scientific and mathematical perspective, before running into the practical realities of how its externalities would interact with real-world politics. Fission is expensive because in the real world it turns out we care quite a lot about proper waste management, non-proliferation, and preventing meltdowns.
In a world where anyone could just YOLO any reactor into production with minimal red tape, consequences be damned, fission energy would actually be extremely cheap. Hence the optimism around fusion. The promise of fusion is an actualization of last century's idealistic conception of fission. It can be a silver bullet for all intents and purposes, at least once it's established with a mature supply chain.
Plasma drilling is a recent development that looks promising for unlocking deeper wells for geothermal.
Remains to be seen, it has serious trouble with water getting into the borehole.
Many others here have talked about the difficulties of geothermal, which doesn't really get to the heart of my question: why the lack of hype around breaking down those difficulties? I appreciate that you took the time to comment on why it isn't so sexy, the SF argument probably has a lot to do with it.
Why not is explained in David McKay's book Sustainable Energy - without the hot air
https://www.withouthotair.com/c16/page_96.shtml
The problems are that rock isn't a good conductor of heat, so once you've cooled a bit down, you have to wait for it to warm up. Warming only happens very slowly at the rate of < 50mW / m² which limits the amount of power you can get out.
That's a really good point, and it does take the wind out of the sails. Isn't it possible that eventually we could drill deeper?
There have been numerous trials that had to be stopped because of the triggered earthquakes... Geothermal is not so easy.
Was going to mention that point: http://www.seismo.ethz.ch/en/knowledge/geothermal-energy-ear...
The worst earthquake that was induced that way was 3.5, but given that one of the quakes happened in an area that had a catastrophic earthquake in the Middle Ages, some caution might be warranted: https://en.wikipedia.org/wiki/1356_Basel_earthquake
Because unless you sit on top of a volcano, amount of renewable geothermal energy is minuscule. In most places on Earth it's somewhere around 40 mW/m2 (i.e. accounting for conversion losses you need to capture heat from ~500 m2 to renewably power one LED light bulb!). In other words, in most places geothermal plant acts more like a limited battery powered by hot rock, so unless drilling is extremely cheap, it does not make economic sense compared to other energy sources.
> In most places on Earth it's somewhere around 40 mW/m2 (i.e. accounting for conversion losses you need to capture heat from ~500 m2 to renewably power one LED light bulb!)
Ground-source heat pumps extract about 1000 times more power from ground loops, where does the difference come from?
I think OP meant technology for drilling becoming cheaper rather than the near-surface availability of it.
I think it mainly depends on how easy it is to access that energy. I went to Tuscany last year and to my surprise there were geothermal plants everywhere. I have never heard about these plants beforehand, but here they are in Italy quietly powering the countryside and heating greenhouses to grow basil all year around.
Probably because not everywhere on earth has the same easy access that Iceland has. The article mentions this:
> There aren’t gates of Hell just anywhere. A kilometre below ground in Kamchatka is considerably hotter than a kilometre below ground in Kansas. There is also readily accessible geothermal energy in Kenya (where it provides almost fifty per cent of the country’s energy), New Zealand (about twenty per cent), and the Philippines (about fifteen per cent)—all volcanic areas along tectonic rifts. But in less Hadean landscapes the costs and uncertainties of drilling deep in search of sufficient heat have curtailed development.
There is a crazy amount of energy available everywhere but it is not in the interest of the very powerful very wealthy existing players. This isn't some grand CONSPIRACY. For example oil companies may construct energy investment portfolios that would quite sensibly acquire promising energy related research. They do a simple cost benefit analysis then chose to modestly further research it or shelve it. They turn it into valuable pieces of paper that accumulate value over time. What is there for them not to like about it?
I like how David Hamel put it: We live in this thin sliver on the surface of the planet where it is reasonably peaceful. This is the tranquility! It's a good thing! If you go up or down by a mere few miles there is so much energy it kills you.
https://archive.today/CR2KL
Instead of drilling deep, there is also an intersting case for storing cheap solar energy as hat in piles of dirts in the summer to power turbines in the winter: https://austinvernon.site/blog/standardthermal.html
The Berlin one also had some promising news today: https://www.adlershof.de/en/news/successful-tests-on-undergr...
We have to see if and when any of them goes into production, but the technology seems very interesting
People across the road from have geothermal, driven by a 1.5m-deep pond right near their house. Their heat never costs more than $100 a month in the winter.
That's a different "geothermal" - the correct name is "ground source heat pump" or in your neighbor's case, a pond-source heat pump. Those exploit the temperature stability that occurs some small numbers or meters subsurface for heating in the winter and cooling in the summer.
"Geothermal energy" involves drilling down to hot rock to tap intense heat to run a turbine that produces electricity.
Drill baby drill!
Seriously this would be such a dream!
Turns out that the best battery is literally 10 feet away* - and you don't even need to charge it!
*if you want to make steam its a few thousand, but for heating and cooling its literally just 10 feet!
The problem is how do you remove the incumbents. Oil lobby is pretty strong. Imagine what would happen to car lobby once we have teleportation.
It always has been. Our problem is switching over existing infrastructure without asinine complainers ruining the revolution. We can't even declare total victory with LED bulbs over incandescent. The war to have solar plants over more coal is falling back to coal thanks mostly to AI. Pushback on geothermal will arrive, I guarantee it.
> We can't even declare total victory with LED bulbs over incandescent.
The LED bulbs I have access to (whatever's in the aisles at Home Depot, Costco, etc.) fail much more frequently than the incandescent bulbs I used to buy, and produce an uglier light that is less warm even on the softest/warmest color settings.
My suspicion is that incandescents were at the "end" of their product lifecycle (high quality available for cheap) and LEDs are nearing the middle (medium quality available for cheap), and that I should buy more expensive LED bulbs, but I still think that there are valid "complaints" against the state of widespread LED lighting. I hope these complaints become invalid within a decade, but for now I still miss the experience of buildings lit by incandescent light.
The other thing with AI--the LED revolution was led on this idea that we all need to work as hard as we can to save energy, but now apparently with AI that's no longer the case, and while I understand that this is just due to which political cabals have control of the regulatory machinery at any given time, it's still frustrating.
> uglier light that is less warm
I figured out why this happens.
The light color they call "daytime" is around 5000K, so I expected it to look like being outside in the sun; but instead I got a cold blueish vibe. The problem? Not enough power! I got the equivalent of a moonlit room.
So I got this 180W LED lamp (that's actual 180W, not 180W equivalent) [1]. It's so bright I couldn't see for 5 minutes. I put two in my office on desk lamps. The room now looks like being outside, without the "ugly blue" tint, even though the product says it's 6000K. The days of my SAD suffering are over!
[1] https://www.amazon.com/dp/B0962X573M
Maybe buy your bulbs somewhere else? I'm yet to replace any of the LED bulbs I've bought over the past 15 years and honestly can't even remember the last time a bulb failed.
Actually, since posting this I've vaguely remembered a previous discussion on here about differences between LED bulbs sold in the US and those sold in UK/EU so maybe that explains it.
> The LED bulbs I have access to (whatever's in the aisles at Home Depot, Costco, etc.) fail much more frequently than the incandescent bulbs I used to buy, and produce an uglier light that is less warm even on the softest/warmest color settings.
LED lamps work just fine, you just need to pay more attention when you’re buying them. Philips makes decent LED lamps.
Make sure you’re buying lamps with 90+ CRI, that will help with the quality of light. 2700K is a good color temp for indoor living room/dining room/bedroom lighting, 3500-4000K for kitchen/garage/task lighting.
You also need to buy special lamps if you put them in an enclosed fixture, look for ‘enclosed fixture’ rated lamps. Regular LED lamps will overheat in an enclosed fixture.
> falling back to coal thanks mostly to AI
citation needed
There is an enormous push to build and power data centers in the DC / Northern Virginia region, and there's legislation in West Virginia right now requiring all coal-fired power plants to operate at at least 69% capacity at all times to support it.
> “West Virginia has numerous coal plants that have powered this country for decades. We need these plants to remain operational,” [WV Governor] Morrisey said. “… We will never turn our backs on our existing coal plants and we will work with the federal government to pursue new coal-fired generation.”
https://westvirginiawatch.com/2025/09/11/morrisey-shares-new...
https://wvpublic.org/story/energy-environment/data-center-bi...
https://www.wvlegislature.gov/Bill_Status/bills_text.cfm?bil...
The only way new coal plants get built from today on is with massive lifetime subsidies, because they are uncompetitive. Ie, if they get built it’s for dumb politics not economics
That sounds like they want to subsidize the coal industry. AI is just the excuse.
https://fortune.com/2025/08/31/ai-data-center-boom-old-coal-...
> The war to have solar plants over more coal is falling back to coal thanks mostly to AI.
Also, due to solar not panning out at scale.[1]
More seriously, coal is just cheaper and, with incentives being removed for green energy, it's the cheapest and fastest option to deploy. It's dead simple and well understood reliable power.
[1]https://apnews.com/article/california-solar-energy-ivanpah-b...
Wild take. New coal is not cheaper. There have been no new coal plants built in the US since 2013.
That solar plant you linked is an obsolete experimental technology. Obsolete because regular PV became so much cheaper.
> New coal is not cheaper.
I see yow it can read that way but it isn't what I said. Coal plants exist, either shuttered or running low loads due to financial incentives (not favoring them).
Studies show solar is cheaper but businesses continue to choose coal. I think the entity who's entire existence depends on them making the correct financial choice is a much better indicator of economic reality than a study made by people who have zero stake (at best) in the game.
I'm all for green energy but I also don't think people are stupid.
The example you chose is of a mirror based Solar system, which yes, is an obsolete technology.
Direct solar continues to be installed at greater amounts every year and coal is economically uncompetitive with basic anything (which is why it is collapsing), and especially against natural gas.
You're exactly right and it raises a question for me. Why do energy generation topics bring people out of the woodworks who cite some very idiosycratic one-off and use it to make out-of-proportion declarations about the utility of a given technology? This is the second one I've seen suggesting solar is doomed when they mean mirrors.
On twitter I saw someone claim PV is useless for heat because non-PV solar water heating is just so much more efficient. Not even true (I think it's a approximately a wash, different advantages in different applications), but very strangely in the weeds on a specific topic. Much too narrow a factual context to substantiate general level claims about solar as an energy writ large.
I think for whatever reason the missing the forest for the trees trap is really potent in energy discussions.
at some point we will figure out that because we took some much energy out of earths core that it stops spinning and causes the magnetic field to collapse ;-)
Not really how that works. Also earths core is being heated from nuclear decay and tidal effects. It’s getting 10’s or TW worth of heat until the sun expands and eats the earth. https://en.wikipedia.org/wiki/Earth's_internal_heat_budget
The world's total energy consumption (most of which is fossil fuels) is currently estimated at 620 exajoules, or 17TWh / year.
Assuming zero growth in energy consumption (hello AI), extracting even half of that seems like it would be consequential.
Thanks for that wikipedia link, it's fascinating!
I think they meant that as a joke.