This is a great summary. Advocates of renewables really need to focus on the storage issue. Solving this at scale will be key. Until then it’s all about nukes.
There is no "storage issue". It just has not been built out yet. It has not been built out yet because that would be stupid use of capital before there is generating capacity to charge it from.
a kilogram of gasoline gives you about 13 kWatt-hrs,
a kilogram of coal a little less, about 8 kWatt-hrs,
a kilogram of lithium ion battery can store only 0.2 kWatt-hrs, and,
a kilogram of water at 1 km altitude is just 0.0027 kWatt-hrs.
On the other hand,
a kilogram of Uranium 235 gives you 24000 kWatt-hrs!! There’s a lot of energy in Uranium 235!
-----
So, barring antimatter, U235 is the ULTIMATE BATTERY.
Most fusion research seems to be focused on fusing the lightest elements, usually isotopes of hydrogen: deuterium and tritium and occasionally lithium to produce energy.
I wonder how much research has been done on the fusion of heavy elements such as lead (Pb)? If it were possible to reverse the fission of U235, i.e., fuse the fission end products such as Pb back into U235 in an energy efficient manner, and with a very low carbon footprint [1], that would solve the energy storage problem.
[1] Also from the article: "CO2 emissions resulting from current storage technologies range from 104 to 407 kg per Mega Watt-hour of delivered energy. Compare this with coal, which releases almost a 1000 kg of CO2 per Mega Watt-hour produced. " This is the CO2 emissions created in the construction and maintenance of the energy storage facility averaged over the total energy stored and delivered during its lifetime.
Nuclear energy has so much potential. Pretty sad people are scared of it. It's clean, it's energy dense, and we can make a lot of it. France gets most of it's energy from Nuclear plants.
Nuclear is dead not because people are afraid of it. Nuclear is dead because it is way more expensive than alternatives. Always was, and getting moreso as renewables cost continues on down.
In the foreseeable future, nukes will be mothballed long short of their design life just because nobody will want to pay enough for their power to enable continuing to operate them. Geothermal might suffer the same fate.
When a nuke is mothballed early, every kWh it ever produced will suddenly be found to have cost much more than had previously been reckoned.
Nuclear as base load is rather price efficient. And with enough units it can deliver at that price every hour of the year. Not as cheap as renewables, but still very affordable.
> It's clean, it's energy dense, and we can make a lot of it. France gets most of it's energy from Nuclear plants.
It's not clean, since the nuclear industry is also polluting the environment. It's not safe, since there is a risk of nuclear weapons, it's also limited (for example one needs water for the production of nuclear electricity - which in dry summers - like right now - can be a problem).
Worst:
Parts of the nuclear energy production is largely unsolved (nuclear waste storage, clean reprocessing, ...).
It's expensive, needs lots of government money and creates huge centralized structures.
It is sad that proliferation of material for nuclear weapons is that scary. Not that I'm arguing that it isn't. But the nuclear waste that's radioactive for 200,000 years doesn't have to be. Breeder reactors and fuel reprocessing can reduce that to hundreds of years. Yeah that's still longer than a couple human lifetime generations, but nowhere near trying to put up warning signs that will continue to scare away people through the rise and fall of several civiliations.
Would have liked for her to call out NRGV as the actual investment scam it is, rather than just pathetically inadequate.
Would have preferred for mineshaft gravity storage to be mentioned, which is not a scam, in place of NRGV.
Would have liked for her to cite ammonia along with hydrogen as chemical storage media. And to mention recent, radical improvements in hydrogen electrolysis efficiency.
Would have liked for pumped hydro not to be described as if it were necessarily dependent on rivers. Or on hills (i.e. mention deep underground cavities as storage volumes).
Would have liked for variant battery tech to have been mentioned, particularly iron, zinc-bromide, and molten-metal chemistries.
But I am happy to see mention of very substantial compressed-air storage systems in production use.
Would have liked for her to mention undersea buoyancy storage as the dark horse. But I understand why she would not be aware of it.
Disappointed that she brought up nukes without mentioning that they are wholly unable to compete with renewables, on cost.
Finally, it would have been very helpful for her to have noted that for now and well into the future, money is overwhelmingly better spent on building out renewable generating capacity than on storage.
> money is overwhelmingly better spent on building out renewable generating capacity than on storage.
Yes, the world uses about 18 TW (continous). If we wanted to provide all that with PV at 20% capacity factor and with a lifetime of 20 years, we need to be producing 5 TW of PV panels per year.*
Currently we're making around one percent of that.
* Actually we don't need all that. The 18 TW is based on "primary energy supply" which ignores the two-thirds energy losses with fossil fuel based supply. In terms of delivered, useful power, we only probably need 8-ish TW currently.
We need to build way more PV factories. But building the wrong kind would be counterproductive. If perovskites end up way better, it would be wasteful to have built out too many new silicon cell factories.
While building out storage now would be a mistake, we should be building factories for storage like crazy. Every kind. We don't know which storage methods will win, but we can't afford to wait and find out.
> Disappointed that she brought up nukes without mentioning that they are wholly unable to compete with renewables, on cost.
Is this true? I mean per average watt I imagine solar might be cheaper, but is it still cheaper when you factor in the cost of the energy storage you need to have a purely solar/wind grid?
I also wonder if NREs can stay cheaper once we actually try to replace the entire world electricity production considering the higher material usage than nuclear if concrete and steel production cannot keep up with demand.
(well, in the end whichever we do is better than doing nothing)
There are different ways to cook the numbers. In particular, there are different kinds of costs.
In nukes, capex (capital expenditure) absolutely dominates. That has to be amortized over all energy produced over the working life of the plant. Opex (operating expense) in nukes is also high in comparison to renewables', dominated by maintenance: overhauling steam turbines, inspections, repairs; and 24/7 staff and security. Fuel cost is low compared to coal, but far from negligible. There are hidden costs usually not counted, such as liability insurance always paid directly by taxpayers. Paying that out of revenue would price nukes' power completely out of the market, assuming any insurance company could afford to cover it at all. And there is decommissioning cost, a substantial fraction of capex.
If the plant is mothballed early because it cannot produce at a competitive price, its cost for all the kWh already delivered suddenly jumps because its capex can be amortized over only those kWh. Similarly, if it is operated only at night, its per-kWh cost doubles because it costs about the same per day whatever its output.
Usually existing nukes are compared to renewables with the nukes' capex, liability, and decommissioning costs artificially pinned to zero, but counting the full capex of the renewables, because those have to be built before they can produce. Renewables' opex is very close to zero.
Maintaining a big steam turbine is remarkably expensive. Nuke plants are built with spares because they spend so much time down for overhauling.
How can you claim nuclear cannot compete on costs? We have no idea how much any of these things you're talking about will cost, because they either don't exist, exist only in research labs, or have never been deployed at the necessary scale.
> But then there’s the dunkelflaute and its evil brother, cold dunkelflaute. ... It’s a shame there aren’t any umlauts in the word, otherwise it’d make a great name for a metal band.
That made me laugh.
Sabine is absolutely right that storage is the key, and it's a presently unsolved problem. The technologies she lists mostly can't scale up to world scale: pumped hydro and CAES need particular geography/geology, flywheeels and falling blocks have too low energy storage density, and lithium is needed for mobile applications. Thermal storage can work for durations of a day or so, but it's very capital intensive and has all the problems of thermal engines. Not really confidence-inspiring.
There are a couple of infant/ancient technologies that Sabine doesn't mention. Iron is plentiful and rechargeable iron-based batteries could supply energy at week-long duration. Ammonia is under active research as an energy storage medium. That could supply seasonal storage.
We have well over a century of experience with both materials, and they are already produced at the necessary scale for the rapid scale-up we need in energy storage. Both storage technologies are near commercial deployment.
I also think that Sabine may be focusing a little too much on efficiency. This is because I expect the price of PV production to drop fairly quickly to low single cents per watt, so PV plus storage could end up in the low double-digit cents per watt.[1] At these prices we could perhaps our relax efficiency requirements a little.
Synthesis of methane from air-captured carbon is one possibility if we do that. That has the advantage of using existing storage and distribution structures, already present at the required scale.
The recent ESA proposal for space-based solar[2], though, is just sad.
The fact that it even exists shows that European officials have not yet faced up to the reality of their situation, and they are still expecting the technology fairy to wave her magic wand to rescue them by lifting a million tonnes of PV satellite into orbit in less than a decade, more or less for free. Such passivity is very disappointing.
1. The path to this is via PV cells printed on flexible plastic film laid directly on the ground, with the strips laid side by side and bonded together. (Think greenhouse plastic strips hundreds of meters long, and with hundred-meter widths). No panel glass covers or aluminum frames, no galvanised steel stand mounts bedded in concrete, no inter-panel wiring, avoiding looming bottlenecks in glass, aluminum, and copper production. Minimal site prep and construction labor, and good land-use efficiency. (Legacy panel farms are only around 50% covered by panels because of shading and the need for access to fix wiring problems or replace panels with broken glass.)
Erthos is half-way to this installation design, but still uses legacy panels. Perovskites are this year at the point where roll-printed PV cells are feasible.
Except, storage is a solved problem. Different solutions will be used in different places. Both compressed-air and pumped hydro are nowhere near as dependent on local details as is often imagined.
There will be no need for seasonal storage at all, because anyplace that needs more energy than they have stored, and cannot book enough from transmission lines, may simply order a shipment of ammonia from any of many producers in the tropics.
But it is correct that orbital solar will not happen, and ESA should be ashamed.
What if we don't need anything fancy? We can reduce our fossil fuel dependency by more than 90% if we switch to renewables, and just use natural gas plants when they don't produce enough energy. No need for any new storage. The current natural gas storage will be many times more than enough.
But isn't natural gas bad? Doesn't it mean we still emit CO2? We can offset that with trees. And even if we don't, I don't think there's any climate scientist who will say if you reduce emissions by 100% by 2050 we'll all be fine, but if we reduce by 90% we'll be doomed.
Sure. How many natural gas plants do you reckon we will need?
Following your idea, it seems like we will need to replace almost all existing base load with a stable options like nat gas.
But then why bother investing in these unreliable “sustainable” sources at all?
So we can collectively pat ourselves on the back while paying much more for electricity?
At some point, perhaps we should start asking why we are replacing base load with intermittent sources when we have no good ways to store energy at scale.
That's the dirty secret. Wind/Solar might be price competitive at some $/MWh but you need build MUCH more capacity to make up for how unreliable they are.
Each utility needs to purchase generation capacity in advance and it has to be enough to cover estimated peak usage. This is a regulatory requirement. I find it interesting that if the utility chooses to reserve wind generation, they can only apply 7% of the nameplate capacity (!)
According to that math, if we're going to replace current coal/nuke/nat gas base load with wind and expect some amount of reliability, we will need to build 14x as much capacity. That is just madness.
Interestingly, regulated utilities actually LOVE this because they make money for their shareholders by having some capital expenditure base amount. The Public Utilities Commission then allows them to earn some return on that base. I believe the allowed return is usually something like 8-10%. The more they spend on capacity and the less reliable the source is, the more money they make.
Pretty perverse incentives, but you're a luddite science denier if you at all question this scheme.
SO, if you have to waaaaay over build capacity because your sources are unreliable, the utility has to invest more to build that out. Their return on investment in generation then comes from YOU via much higher rates.
> Wind and solar are quite unreliable. That's the dirty secret.
It's not a secret, that's what most people say. But is it true?
Take a look at the graphs in [1] with monthly electricity generation by source in the US. You will notice that renewable electricity generation is a bit jagged, but not a lot. The ratio between the highs and the lows may be 1.3, but it's certainly not 7 or 15 as you imply.
The thing about 7% of the nameplace capacity counting towards peak availability is a red herring. Power companies can keep natural gas plants. Those are very cheap (about $1000/kw of nameplate capacity, on par with solar, but obviously without intermittency issues). They don't need to run them all the time, 5-10% of the time will be enough. Overbuilding renewables by a huge factor to supposedly be allowed to earn 8-10% is a nice conspiracy theory, but assumes they'll find people to sell to at a profit. When you generate many times what people consume, you'll eventually need to pay others to take the electricity from you (yes, negative prices are a thing in the power markets).
France is (by far) the world leader when it comes to the fraction of nuclear-produced electricity and always produces around 9% of its electricity thanks to fossil fuel, because building enough reactors to satisfy peaks and to load-follow would be way too expensive (as most would be idling most of the time).
9% of a year is approximately 33 days, therefore we know how to cope with a Dunkelflaut.
https://www.statistiques.developpement-durable.gouv.fr/editi...
Article failed to mention that wind/solar storage can be usefully teamed-up with the another (renewable) energy source, the heat in the ground ... which is nearby many of our houses.
Today's heat pumps can work well for many down to outdoor temperatures of about 5 degrees F (-15C). They'll bring down the cost of using the other sources. [https://www.youtube.com/watch?v=MFEHFsO-XSI]
36 comments
[ 5.6 ms ] story [ 91.8 ms ] thread-----
a kilogram of gasoline gives you about 13 kWatt-hrs,
a kilogram of coal a little less, about 8 kWatt-hrs,
a kilogram of lithium ion battery can store only 0.2 kWatt-hrs, and,
a kilogram of water at 1 km altitude is just 0.0027 kWatt-hrs.
On the other hand,
a kilogram of Uranium 235 gives you 24000 kWatt-hrs!! There’s a lot of energy in Uranium 235!
-----
So, barring antimatter, U235 is the ULTIMATE BATTERY.
Most fusion research seems to be focused on fusing the lightest elements, usually isotopes of hydrogen: deuterium and tritium and occasionally lithium to produce energy.
I wonder how much research has been done on the fusion of heavy elements such as lead (Pb)? If it were possible to reverse the fission of U235, i.e., fuse the fission end products such as Pb back into U235 in an energy efficient manner, and with a very low carbon footprint [1], that would solve the energy storage problem.
[1] Also from the article: "CO2 emissions resulting from current storage technologies range from 104 to 407 kg per Mega Watt-hour of delivered energy. Compare this with coal, which releases almost a 1000 kg of CO2 per Mega Watt-hour produced. " This is the CO2 emissions created in the construction and maintenance of the energy storage facility averaged over the total energy stored and delivered during its lifetime.
In the foreseeable future, nukes will be mothballed long short of their design life just because nobody will want to pay enough for their power to enable continuing to operate them. Geothermal might suffer the same fate.
When a nuke is mothballed early, every kWh it ever produced will suddenly be found to have cost much more than had previously been reckoned.
At the moment, renewables still can't compete on pricing with 70s nuclear designs...
b) are you aware that half of the nuclear fleet in France isn't online?
https://www.brusselstimes.com/belgium-news/274261/controlled...
> It's clean, it's energy dense, and we can make a lot of it. France gets most of it's energy from Nuclear plants.
It's not clean, since the nuclear industry is also polluting the environment. It's not safe, since there is a risk of nuclear weapons, it's also limited (for example one needs water for the production of nuclear electricity - which in dry summers - like right now - can be a problem).
Worst:
Parts of the nuclear energy production is largely unsolved (nuclear waste storage, clean reprocessing, ...).
It's expensive, needs lots of government money and creates huge centralized structures.
Had this money been spent on the nuclear grid, the gris would be in a much better state now.
https://www.reuters.com/business/energy/edf-hopeful-end-sigh...
With more than half of the French nuclear fleet offline, nuclear in this energy crisis is just only contributing to the problems.
Would have liked for her to call out NRGV as the actual investment scam it is, rather than just pathetically inadequate.
Would have preferred for mineshaft gravity storage to be mentioned, which is not a scam, in place of NRGV.
Would have liked for her to cite ammonia along with hydrogen as chemical storage media. And to mention recent, radical improvements in hydrogen electrolysis efficiency.
Would have liked for pumped hydro not to be described as if it were necessarily dependent on rivers. Or on hills (i.e. mention deep underground cavities as storage volumes).
Would have liked for variant battery tech to have been mentioned, particularly iron, zinc-bromide, and molten-metal chemistries.
But I am happy to see mention of very substantial compressed-air storage systems in production use.
Would have liked for her to mention undersea buoyancy storage as the dark horse. But I understand why she would not be aware of it.
Disappointed that she brought up nukes without mentioning that they are wholly unable to compete with renewables, on cost.
Finally, it would have been very helpful for her to have noted that for now and well into the future, money is overwhelmingly better spent on building out renewable generating capacity than on storage.
Yes, the world uses about 18 TW (continous). If we wanted to provide all that with PV at 20% capacity factor and with a lifetime of 20 years, we need to be producing 5 TW of PV panels per year.*
Currently we're making around one percent of that.
* Actually we don't need all that. The 18 TW is based on "primary energy supply" which ignores the two-thirds energy losses with fossil fuel based supply. In terms of delivered, useful power, we only probably need 8-ish TW currently.
While building out storage now would be a mistake, we should be building factories for storage like crazy. Every kind. We don't know which storage methods will win, but we can't afford to wait and find out.
Is this true? I mean per average watt I imagine solar might be cheaper, but is it still cheaper when you factor in the cost of the energy storage you need to have a purely solar/wind grid?
I'd love to see some numbers on this.
(well, in the end whichever we do is better than doing nothing)
In nukes, capex (capital expenditure) absolutely dominates. That has to be amortized over all energy produced over the working life of the plant. Opex (operating expense) in nukes is also high in comparison to renewables', dominated by maintenance: overhauling steam turbines, inspections, repairs; and 24/7 staff and security. Fuel cost is low compared to coal, but far from negligible. There are hidden costs usually not counted, such as liability insurance always paid directly by taxpayers. Paying that out of revenue would price nukes' power completely out of the market, assuming any insurance company could afford to cover it at all. And there is decommissioning cost, a substantial fraction of capex.
If the plant is mothballed early because it cannot produce at a competitive price, its cost for all the kWh already delivered suddenly jumps because its capex can be amortized over only those kWh. Similarly, if it is operated only at night, its per-kWh cost doubles because it costs about the same per day whatever its output.
Usually existing nukes are compared to renewables with the nukes' capex, liability, and decommissioning costs artificially pinned to zero, but counting the full capex of the renewables, because those have to be built before they can produce. Renewables' opex is very close to zero.
Maintaining a big steam turbine is remarkably expensive. Nuke plants are built with spares because they spend so much time down for overhauling.
That made me laugh.
Sabine is absolutely right that storage is the key, and it's a presently unsolved problem. The technologies she lists mostly can't scale up to world scale: pumped hydro and CAES need particular geography/geology, flywheeels and falling blocks have too low energy storage density, and lithium is needed for mobile applications. Thermal storage can work for durations of a day or so, but it's very capital intensive and has all the problems of thermal engines. Not really confidence-inspiring.
There are a couple of infant/ancient technologies that Sabine doesn't mention. Iron is plentiful and rechargeable iron-based batteries could supply energy at week-long duration. Ammonia is under active research as an energy storage medium. That could supply seasonal storage.
We have well over a century of experience with both materials, and they are already produced at the necessary scale for the rapid scale-up we need in energy storage. Both storage technologies are near commercial deployment.
I also think that Sabine may be focusing a little too much on efficiency. This is because I expect the price of PV production to drop fairly quickly to low single cents per watt, so PV plus storage could end up in the low double-digit cents per watt.[1] At these prices we could perhaps our relax efficiency requirements a little.
Synthesis of methane from air-captured carbon is one possibility if we do that. That has the advantage of using existing storage and distribution structures, already present at the required scale.
The recent ESA proposal for space-based solar[2], though, is just sad.
The fact that it even exists shows that European officials have not yet faced up to the reality of their situation, and they are still expecting the technology fairy to wave her magic wand to rescue them by lifting a million tonnes of PV satellite into orbit in less than a decade, more or less for free. Such passivity is very disappointing.
1. The path to this is via PV cells printed on flexible plastic film laid directly on the ground, with the strips laid side by side and bonded together. (Think greenhouse plastic strips hundreds of meters long, and with hundred-meter widths). No panel glass covers or aluminum frames, no galvanised steel stand mounts bedded in concrete, no inter-panel wiring, avoiding looming bottlenecks in glass, aluminum, and copper production. Minimal site prep and construction labor, and good land-use efficiency. (Legacy panel farms are only around 50% covered by panels because of shading and the need for access to fix wiring problems or replace panels with broken glass.)
Erthos is half-way to this installation design, but still uses legacy panels. Perovskites are this year at the point where roll-printed PV cells are feasible.
2. https://arstechnica.com/science/2022/08/european-space-chief...
There will be no need for seasonal storage at all, because anyplace that needs more energy than they have stored, and cannot book enough from transmission lines, may simply order a shipment of ammonia from any of many producers in the tropics.
But it is correct that orbital solar will not happen, and ESA should be ashamed.
But isn't natural gas bad? Doesn't it mean we still emit CO2? We can offset that with trees. And even if we don't, I don't think there's any climate scientist who will say if you reduce emissions by 100% by 2050 we'll all be fine, but if we reduce by 90% we'll be doomed.
Following your idea, it seems like we will need to replace almost all existing base load with a stable options like nat gas.
But then why bother investing in these unreliable “sustainable” sources at all?
So we can collectively pat ourselves on the back while paying much more for electricity?
At some point, perhaps we should start asking why we are replacing base load with intermittent sources when we have no good ways to store energy at scale.
Many fewer than we already have.
Not sure I understand your other questions. What unreliable "sustainable" sources? Why paying much more for electricity?
That's the dirty secret. Wind/Solar might be price competitive at some $/MWh but you need build MUCH more capacity to make up for how unreliable they are.
Each utility needs to purchase generation capacity in advance and it has to be enough to cover estimated peak usage. This is a regulatory requirement. I find it interesting that if the utility chooses to reserve wind generation, they can only apply 7% of the nameplate capacity (!)
According to that math, if we're going to replace current coal/nuke/nat gas base load with wind and expect some amount of reliability, we will need to build 14x as much capacity. That is just madness.
Interestingly, regulated utilities actually LOVE this because they make money for their shareholders by having some capital expenditure base amount. The Public Utilities Commission then allows them to earn some return on that base. I believe the allowed return is usually something like 8-10%. The more they spend on capacity and the less reliable the source is, the more money they make.
Pretty perverse incentives, but you're a luddite science denier if you at all question this scheme.
SO, if you have to waaaaay over build capacity because your sources are unreliable, the utility has to invest more to build that out. Their return on investment in generation then comes from YOU via much higher rates.
It's not a secret, that's what most people say. But is it true?
Take a look at the graphs in [1] with monthly electricity generation by source in the US. You will notice that renewable electricity generation is a bit jagged, but not a lot. The ratio between the highs and the lows may be 1.3, but it's certainly not 7 or 15 as you imply.
The thing about 7% of the nameplace capacity counting towards peak availability is a red herring. Power companies can keep natural gas plants. Those are very cheap (about $1000/kw of nameplate capacity, on par with solar, but obviously without intermittency issues). They don't need to run them all the time, 5-10% of the time will be enough. Overbuilding renewables by a huge factor to supposedly be allowed to earn 8-10% is a nice conspiracy theory, but assumes they'll find people to sell to at a profit. When you generate many times what people consume, you'll eventually need to pay others to take the electricity from you (yes, negative prices are a thing in the power markets).
[1] https://www.eia.gov/totalenergy/data/monthly/pdf/sec7_4.pdf
Moreover a continental mix (wind, solar, geothermal...) reduces intermittency. Wind: https://www.imperial.ac.uk/news/180592/european-cooperation-...
Moreover modern gas turbines are able to burn (up to 100% of) hydrogen, therefore feeding them green hydrogen (obtained during over-production periods) will lower emissions. https://global.kawasaki.com/news_211209-2e.pdf https://www.get-h2.de/en/project-lingen/
Moreover storage has many gorillas in the room, mainly V2G https://en.wikipedia.org/wiki/Vehicle-to-grid
Today's heat pumps can work well for many down to outdoor temperatures of about 5 degrees F (-15C). They'll bring down the cost of using the other sources. [https://www.youtube.com/watch?v=MFEHFsO-XSI]