5 ms·

It's not that nuclear is unproven, it's that _renewables at this point are proven_. You can build huge (or tiny! it's super scalable) fleets of renewable energ

by rtpg 8y ago

It's not that nuclear is unproven, it's that _renewables at this point are proven_. You can build huge (or tiny! it's super scalable) fleets of renewable energy sources pretty economically, with a ~0% chance of making surrounding areas uninhabitable for 20 years if there's a failure. Yeah sure a truckload of uranium goes a long way but building the plants is really hard even when you look past the regulatory issues.

rurounijones 8y ago
Someone said roughly the same thing in another thread https://news.ycombinator.com/item?id=19167849 https://news.ycombinator.com/item?id=19167849 But as far as I know there is no proven large-scale storage system, which is a criticla part of renewables. The other commenter mentioned building batteries and then it is just a matter of scale which seems like an over-simplification to me but I cannot find much useful info on it. Personally I would love a 100% renewable setup but I would not bet the future of our civilization on it when nuclear is another option.
- darkmarmot 8y ago
  We do have a very simple large-scale storage system that's already in use across the world: pumping water uphill. It's just not as economical as rain when you factor in the efficiency losses from generating and transferring power twice.
  - rurounijones 8y ago
    And pumping water uphill is a regional solution that requires specific regional geography, to me that does not class as "scalable". I should be more specific in that to me "scalable" is a solution that can be deployed anywhere without require specific regional properties. Maybe you think this is unneeded and that a lot of region specific solutions can cover the storage requirements. However another posted did point out that 1 mile by 1 mile of batteries is apparently enough to store US energy requirements.
    - ShorsHammer 8y ago
      Unless there is big improvements in battery tech I don't see it as a solution. Lithium batteries have their own environmental problems and die after 10 years. Pumped hydro requires the geography, building new capacity tends to flood large areas and release huge amounts of greenhouse gases initially. Simply lifting weights on cranes or rail tracks is looking more efficient and can be adapted anywhere. It will probably win out in many cases.
      - usrusr 8y ago
        > Simply lifting weights on cranes or rail tracks is looking more efficient and can be adapted anywhere. The energy density of that kind of scheme is a joke. Thought experiment: a fully charged Tesla, how often could it climb a hill the size of that crane if it skips recuperation on the downhill leg? That's how much lower the energy density of a crane storage would be. Pumped storage works (where the geography allows it) because water is by far the cheapest and the most easily transported ballast and geographic height differences dwarf almost every human made structure. If you want something that scales everywhere, look no further than compressed air. It's usually ignored because of the big thermal losses, but if you have a direct application for coolant they are not that bad and even without, it serves as an almost trivial lower bound to the storage problem. We can calculate how much intermittent energy production we would need with compressed air to serve a power demand profile and everything else is just a possible improvement.
        ShorsHammer 8y ago
        Since when was energy density ever been a problem for grid storage? Even in the most crowded cities on earth it's not a constraint. Unless you are talking about portable applications it's all about round trip efficiency and cost, I'm not sure if you are getting confused about terms here or talking about an entirely different subject. Compressed air efficiency isn't great. Using it for cooling lowers the efficiency even more. Crane and rail gravity storage are pushing 80-90% round trip efficiency. Previous HN discussion: https://news.ycombinator.com/item?id=17789456 https://news.ycombinator.com/item?id=17789456
- dredmorbius 8y ago
  There are several solutions to supply-demand matching and buffering which either do or should work. A key is to think of this in terms of matching supply and demand rather than simply as storage. We've adapted over the course of a century or so to a regime of dispatchable supply energy, with little use of dispatchable demand. There's also been little consideration of major behavioural, social, economic, and land-use changes which will be prompted by changes to the energy regime. Much the same way as major impacts of internal combustion engines on land-use, construction, transport, and trade were almost wholly unanticipated, most discussion today is framed in terms of "how do we sustain present behaviours and activities under a novel energy regime" (if not quite so explicitly). The short answer is: you don't. Automobiles, rail, air transport, and powered shipping gave rise to suburban sprawl, transcontinental trade networks, same-day globe-spanning travel and light cargo, and transoceanic shipping centres, along with tremendous centralisation of activities in zones of maximum productivity (often, yes, through massive externalised costs). Little of those impacts was foreseen in the popular or academic literature of a century (or even half-century) ago. There's much of economics that's badly broken, but a part that's useful is the notion that behaviours do change tremendously in the face of changes to real and expressed costs. The Jevons Paradox cuts both ways: increased efficiency increases total use, whilst increased costs will decrease total use of some resource or factor. (Increasing efficiency is equivalent to saying "decreasing cost".) Addressing energy specifically: Expect to see far more dispatchable load, effectively, "making hay whilst the sun shines". High-load, but bufferable uses such as thermal heating (water, space, thermal storage), electrically-driven refinement (aluminium smelting, electric arc furnaces), reverse osmosis desalination, and the like, can if not "store electricity", then cache useful activity whilst supplies are abundant. Smaller industrial, commercial, residential, and possibly transport loads may also see time-shifting on a similar basis. For direct storage, pumped hydro, compressed air energy storage (CAES), grid-scale batteries (an area with frustratingly slow development, though some promise, especially with cheap-and-abundant if not highly-efficient electrolytes), are presently proven. There are a number of schemes which don't work particularly well -- flywheel storage doesn't seem useful for much besides replicating today's "spinning reserve", nor do supercapacitors look as if they'll offer much beyond grid-scale power conditioning. Two areas which offer tremendous promise and fairly high probability of success are grid-scale thermal energy storage with regeneration and electrically-based fuel synthesis appear at least on a back-of-the-envelope basis to provide national-scale grid-level storage capacity good for weeks (thermal) to millennia (fuel synthesis). The round-trip efficiencies are not great, but the simplicity, safety, and in the case of fuel synthesis, exceedingly long-duration storage, transportability, and utility of the derived medium are huge advantages. Thermal energy electrical storage is mostly used now in solar thermal generation plants, but could be utilised in other forms. Large insulated tanks of molten salt driving traditional steam turbine generation could offer weeks worth of grid storage for the US in a total storage capacity roughly of the magnitude of extant oil transport storage facilities in Oklahoma. Synthetic fuel generation, first suggested for nuclear power at Brookhaven National Laboratory in the 1960s and researched for over 50 years at Brookhaven, M.I.T., and the US Naval Research Laboratory, has yet to be proven at national scale, but the basic chemistry works, it's similar to coal-to-oil processes used by South Africa and Germany since World War II, produces direct analogues to current fossil fuels (methane through bunker oil) but is carbon-neutral as the carbon itself is sourced from current biosphere reserves, principally seawater. Otherwise: expect to see tremendous differences in how energy is used, in construction based around heating and cooling loads, lighting, transport, and other processes.
squirrelicus 8y ago
I think you're overstating the provenness of renewables. There's a lot of support infrastructure that hasn't been invented that is required to produce the roughly 3 terawatt average per day as well as the storage for the daily cycle. I also think your overstating the risk involved in nuclear power. Even fukushima was pretty tame, and unlike Japan the USA has lots of relatively uninhabited real estate we can put the plants on. Cost is one factor. Location is another. Reliability another. Risk another. Nuclear is a system of tradeoffs that is very hopeful for fighting emissions. Of course this is all moot since even if the USA and Europe went to zero emissions, we won't make a dent in climate change. The developing world must be addressed with at least as much urgency as the West must be addressed.
- ben_w 8y ago
  > There's a lot of support infrastructure that hasn't been invented that is required to produce the roughly 3 terawatt average per day as well as the storage for the daily cycle. On the contrary, two solutions already exist. Batteries are expensive, but (IIRC) still cheaper than nuclear, though more expensive than fossil fuels. And the losses of planet-scale grids are small compared to the cost-advantage that solar has. You could literally power your home at night from the sun hitting the other side of the planet. (That said, I have no idea what the Installation or maintenance or political costs are, only the efficiency).
  - pdonis 8y ago
    > the losses of planet-scale grids are small compared to the cost-advantage that solar has. You could literally power your home at night from the sun hitting the other side of the planet. You're comparing apples and widgets. The losses of planet-scale grids are way too large to make it even feasible to power your home at night from the other side of the planet: not enough power would be left by the time it got to you. The cost is irrelevant.
    - ben_w 8y ago
      “””the authors measured the corona loss of a 765kV, 3 phase, and bundled transmission line to be about 1.87kW/km in fair weather. This amounts to only about a 0.083% loss over a 1000km line. In bad weather, however, the authors measured the loss to be 84.3kW/km, or about a 3.7% loss.””” - http://large.stanford.edu/courses/2010/ph240/harting1/ http://large.stanford.edu/courses/2010/ph240/harting1/ 3.7% loss = (100-3.7)/100 multiplier per 1000km = 0.963 Half world circumstance = 20,000km -> 0.963^20 = ~0.47 multiplier “””New US Solar Record — 2.155 Cents Per kWh””” - https://cleantechnica.com/2018/06/14/new-us-solar-record-2-155-cents-per-kwh-400-mwh-of-energy-storage/ https://cleantechnica.com/2018/06/14/new-us-solar-record-2-1... -> 2.155 cents per 0.47 kWh = 4.5 cents per kWh from the opposite point of the planet, assuming the worst case studied in the first link on the entire route. It could be 80% losses (0.2 kWh receives for every 1 kWh produced) and still be effective both from ‘cost’ and ‘maximum possible power output’ perspectives.
666lumberjack 8y ago
Can you give an example of an industrialized first-world nation using Wind + Solar to provide baseload power on a consistent basis? Because if you can't then they're far from proven as anything more than a partial supplement to other energy sources. The economics of Solar/Wind get much worse as you scale past ~30% of the grid because utilization factor goes down and the need for energy storage goes way up. Right now grid-scale batteries are uneconomical and pumped hydro is only feasible in specific locations (not to mention its ecological impact) so the 'storage' backing renewable installations takes the form of big tanks of methane next to gas turbines. Even when conditions are good and it's not being burned, that methane has a tendency to leak and wipe out a fair proportion of the face-value greenhouse emission savings from swapping to renewables in the first place.
- olau 8y ago
  That's not correct. Denmark is currently is currently at something like 45% of electricity from wind, and going past 50% in a year or two IIRC. It's not base load, because you don't need base load - it's an obsolete model from a world of capital-intensive, constant output plants. What you need is to match the demand. Regarding long-term storage: Exactly. Tanks or caverns filled with methane generated from a renewable source is one possible solution. Of course, you need to have a firm grip on leaks.
  - 666lumberjack 8y ago
    Renewables aren't any better at load following than current nuclear technology, and unlike nuclear energy there's no potential for the development of proposed designs that do have load-following capability. Never mind the fact that intermittent renewables themselves are the reason the 'mostly base load with some load following' model is obsolete, precisely because they can't be relied upon to generate electricity on demand. Denmark is definitely the poster child of wind energy (and intermittent renewables in general), but it benefits enormously from utilizing the existing pumped hydro storage of Sweden and Norway, which is an option the vast majority of countries don't have. The economics would look at lot worse if they were having the build out batteries to provide that storage capacity instead of taking advantage of what's already there. Just saying 'you need to have a firm grip on leaks' is trivializing a significant engineering problem. Existing natural gas plants have a 1-9% leak rate, and given than leaking fuel is a complete waste of money we can safely assume that doing better than that is non-trivial. Given that in the case of carbon-neutral 'synthetic' methane you have the possibility of leaks at the manufacturing plant as well as in transmission/storage, it's reasonable to expect a figure closer to the higher end of that spectrum, which makes lifetime emissions from Wind/Solar sources significantly worse than Nuclear on a per-GWh basis.

by rtpg 8y ago

rurounijones 8y ago

Someone said roughly the same thing in another thread https://news.ycombinator.com/item?id=19167849 https://news.ycombinator.com/item?id=19167849 But as far as I know there is no proven large-scale storage system, which is a criticla part of renewables. The other commenter mentioned building batteries and then it is just a matter of scale which seems like an over-simplification to me but I cannot find much useful info on it. Personally I would love a 100% renewable setup but I would not bet the future of our civilization on it when nuclear is another option.

darkmarmot 8y ago
We do have a very simple large-scale storage system that's already in use across the world: pumping water uphill. It's just not as economical as rain when you factor in the efficiency losses from generating and transferring power twice.
- rurounijones 8y ago
  And pumping water uphill is a regional solution that requires specific regional geography, to me that does not class as "scalable". I should be more specific in that to me "scalable" is a solution that can be deployed anywhere without require specific regional properties. Maybe you think this is unneeded and that a lot of region specific solutions can cover the storage requirements. However another posted did point out that 1 mile by 1 mile of batteries is apparently enough to store US energy requirements.
  - ShorsHammer 8y ago
    Unless there is big improvements in battery tech I don't see it as a solution. Lithium batteries have their own environmental problems and die after 10 years. Pumped hydro requires the geography, building new capacity tends to flood large areas and release huge amounts of greenhouse gases initially. Simply lifting weights on cranes or rail tracks is looking more efficient and can be adapted anywhere. It will probably win out in many cases.
    - usrusr 8y ago
      > Simply lifting weights on cranes or rail tracks is looking more efficient and can be adapted anywhere. The energy density of that kind of scheme is a joke. Thought experiment: a fully charged Tesla, how often could it climb a hill the size of that crane if it skips recuperation on the downhill leg? That's how much lower the energy density of a crane storage would be. Pumped storage works (where the geography allows it) because water is by far the cheapest and the most easily transported ballast and geographic height differences dwarf almost every human made structure. If you want something that scales everywhere, look no further than compressed air. It's usually ignored because of the big thermal losses, but if you have a direct application for coolant they are not that bad and even without, it serves as an almost trivial lower bound to the storage problem. We can calculate how much intermittent energy production we would need with compressed air to serve a power demand profile and everything else is just a possible improvement.
      - ShorsHammer 8y ago
        Since when was energy density ever been a problem for grid storage? Even in the most crowded cities on earth it's not a constraint. Unless you are talking about portable applications it's all about round trip efficiency and cost, I'm not sure if you are getting confused about terms here or talking about an entirely different subject. Compressed air efficiency isn't great. Using it for cooling lowers the efficiency even more. Crane and rail gravity storage are pushing 80-90% round trip efficiency. Previous HN discussion: https://news.ycombinator.com/item?id=17789456 https://news.ycombinator.com/item?id=17789456
dredmorbius 8y ago
There are several solutions to supply-demand matching and buffering which either do or should work. A key is to think of this in terms of matching supply and demand rather than simply as storage. We've adapted over the course of a century or so to a regime of dispatchable supply energy, with little use of dispatchable demand. There's also been little consideration of major behavioural, social, economic, and land-use changes which will be prompted by changes to the energy regime. Much the same way as major impacts of internal combustion engines on land-use, construction, transport, and trade were almost wholly unanticipated, most discussion today is framed in terms of "how do we sustain present behaviours and activities under a novel energy regime" (if not quite so explicitly). The short answer is: you don't. Automobiles, rail, air transport, and powered shipping gave rise to suburban sprawl, transcontinental trade networks, same-day globe-spanning travel and light cargo, and transoceanic shipping centres, along with tremendous centralisation of activities in zones of maximum productivity (often, yes, through massive externalised costs). Little of those impacts was foreseen in the popular or academic literature of a century (or even half-century) ago. There's much of economics that's badly broken, but a part that's useful is the notion that behaviours do change tremendously in the face of changes to real and expressed costs. The Jevons Paradox cuts both ways: increased efficiency increases total use, whilst increased costs will decrease total use of some resource or factor. (Increasing efficiency is equivalent to saying "decreasing cost".) Addressing energy specifically: Expect to see far more dispatchable load, effectively, "making hay whilst the sun shines". High-load, but bufferable uses such as thermal heating (water, space, thermal storage), electrically-driven refinement (aluminium smelting, electric arc furnaces), reverse osmosis desalination, and the like, can if not "store electricity", then cache useful activity whilst supplies are abundant. Smaller industrial, commercial, residential, and possibly transport loads may also see time-shifting on a similar basis. For direct storage, pumped hydro, compressed air energy storage (CAES), grid-scale batteries (an area with frustratingly slow development, though some promise, especially with cheap-and-abundant if not highly-efficient electrolytes), are presently proven. There are a number of schemes which don't work particularly well -- flywheel storage doesn't seem useful for much besides replicating today's "spinning reserve", nor do supercapacitors look as if they'll offer much beyond grid-scale power conditioning. Two areas which offer tremendous promise and fairly high probability of success are grid-scale thermal energy storage with regeneration and electrically-based fuel synthesis appear at least on a back-of-the-envelope basis to provide national-scale grid-level storage capacity good for weeks (thermal) to millennia (fuel synthesis). The round-trip efficiencies are not great, but the simplicity, safety, and in the case of fuel synthesis, exceedingly long-duration storage, transportability, and utility of the derived medium are huge advantages. Thermal energy electrical storage is mostly used now in solar thermal generation plants, but could be utilised in other forms. Large insulated tanks of molten salt driving traditional steam turbine generation could offer weeks worth of grid storage for the US in a total storage capacity roughly of the magnitude of extant oil transport storage facilities in Oklahoma. Synthetic fuel generation, first suggested for nuclear power at Brookhaven National Laboratory in the 1960s and researched for over 50 years at Brookhaven, M.I.T., and the US Naval Research Laboratory, has yet to be proven at national scale, but the basic chemistry works, it's similar to coal-to-oil processes used by South Africa and Germany since World War II, produces direct analogues to current fossil fuels (methane through bunker oil) but is carbon-neutral as the carbon itself is sourced from current biosphere reserves, principally seawater. Otherwise: expect to see tremendous differences in how energy is used, in construction based around heating and cooling loads, lighting, transport, and other processes.

squirrelicus 8y ago

I think you're overstating the provenness of renewables. There's a lot of support infrastructure that hasn't been invented that is required to produce the roughly 3 terawatt average per day as well as the storage for the daily cycle. I also think your overstating the risk involved in nuclear power. Even fukushima was pretty tame, and unlike Japan the USA has lots of relatively uninhabited real estate we can put the plants on. Cost is one factor. Location is another. Reliability another. Risk another. Nuclear is a system of tradeoffs that is very hopeful for fighting emissions. Of course this is all moot since even if the USA and Europe went to zero emissions, we won't make a dent in climate change. The developing world must be addressed with at least as much urgency as the West must be addressed.

ben_w 8y ago
> There's a lot of support infrastructure that hasn't been invented that is required to produce the roughly 3 terawatt average per day as well as the storage for the daily cycle. On the contrary, two solutions already exist. Batteries are expensive, but (IIRC) still cheaper than nuclear, though more expensive than fossil fuels. And the losses of planet-scale grids are small compared to the cost-advantage that solar has. You could literally power your home at night from the sun hitting the other side of the planet. (That said, I have no idea what the Installation or maintenance or political costs are, only the efficiency).
- pdonis 8y ago
  > the losses of planet-scale grids are small compared to the cost-advantage that solar has. You could literally power your home at night from the sun hitting the other side of the planet. You're comparing apples and widgets. The losses of planet-scale grids are way too large to make it even feasible to power your home at night from the other side of the planet: not enough power would be left by the time it got to you. The cost is irrelevant.
  - ben_w 8y ago
    “””the authors measured the corona loss of a 765kV, 3 phase, and bundled transmission line to be about 1.87kW/km in fair weather. This amounts to only about a 0.083% loss over a 1000km line. In bad weather, however, the authors measured the loss to be 84.3kW/km, or about a 3.7% loss.””” - http://large.stanford.edu/courses/2010/ph240/harting1/ http://large.stanford.edu/courses/2010/ph240/harting1/ 3.7% loss = (100-3.7)/100 multiplier per 1000km = 0.963 Half world circumstance = 20,000km -> 0.963^20 = ~0.47 multiplier “””New US Solar Record — 2.155 Cents Per kWh””” - https://cleantechnica.com/2018/06/14/new-us-solar-record-2-155-cents-per-kwh-400-mwh-of-energy-storage/ https://cleantechnica.com/2018/06/14/new-us-solar-record-2-1... -> 2.155 cents per 0.47 kWh = 4.5 cents per kWh from the opposite point of the planet, assuming the worst case studied in the first link on the entire route. It could be 80% losses (0.2 kWh receives for every 1 kWh produced) and still be effective both from ‘cost’ and ‘maximum possible power output’ perspectives.

666lumberjack 8y ago

Can you give an example of an industrialized first-world nation using Wind + Solar to provide baseload power on a consistent basis? Because if you can't then they're far from proven as anything more than a partial supplement to other energy sources. The economics of Solar/Wind get much worse as you scale past ~30% of the grid because utilization factor goes down and the need for energy storage goes way up. Right now grid-scale batteries are uneconomical and pumped hydro is only feasible in specific locations (not to mention its ecological impact) so the 'storage' backing renewable installations takes the form of big tanks of methane next to gas turbines. Even when conditions are good and it's not being burned, that methane has a tendency to leak and wipe out a fair proportion of the face-value greenhouse emission savings from swapping to renewables in the first place.

olau 8y ago
That's not correct. Denmark is currently is currently at something like 45% of electricity from wind, and going past 50% in a year or two IIRC. It's not base load, because you don't need base load - it's an obsolete model from a world of capital-intensive, constant output plants. What you need is to match the demand. Regarding long-term storage: Exactly. Tanks or caverns filled with methane generated from a renewable source is one possible solution. Of course, you need to have a firm grip on leaks.
- 666lumberjack 8y ago
  Renewables aren't any better at load following than current nuclear technology, and unlike nuclear energy there's no potential for the development of proposed designs that do have load-following capability. Never mind the fact that intermittent renewables themselves are the reason the 'mostly base load with some load following' model is obsolete, precisely because they can't be relied upon to generate electricity on demand. Denmark is definitely the poster child of wind energy (and intermittent renewables in general), but it benefits enormously from utilizing the existing pumped hydro storage of Sweden and Norway, which is an option the vast majority of countries don't have. The economics would look at lot worse if they were having the build out batteries to provide that storage capacity instead of taking advantage of what's already there. Just saying 'you need to have a firm grip on leaks' is trivializing a significant engineering problem. Existing natural gas plants have a 1-9% leak rate, and given than leaking fuel is a complete waste of money we can safely assume that doing better than that is non-trivial. Given that in the case of carbon-neutral 'synthetic' methane you have the possibility of leaks at the manufacturing plant as well as in transmission/storage, it's reasonable to expect a figure closer to the higher end of that spectrum, which makes lifetime emissions from Wind/Solar sources significantly worse than Nuclear on a per-GWh basis.