4 ms·

From a software / devils point of view Nuclear power is the Mainframe - and solar or wind are the nice new micro services. If we have a crack in our solar pane

From a software / devils point of view Nuclear power is the Mainframe - and solar or wind are the nice new micro services. If we have a crack in our solar panels, it ain't going to make the BBC news site.

mbell 7y ago
I wouldn't call solar 'micro', land use is one of its largest downsides. > If we have a crack in our solar panels, it ain't going to make the BBC news site. Actually cracking solar panels could be a rather large problem. e.g. Cadmium Telluride cells are gaining in popularity, if a tornado, large hail storm, or similar natural event hits a power plant built with such cells it could result in non-trivial release into the environment.
layoutIfNeeded 7y ago
Nuclear is like monolithic kernels: gets a lot of flak for ideological reasons, but it keeps powering the whole world. Solar is like microkernels: nice on paper but lacks the power for anything serious.
- baq 7y ago
  Eagerly waiting for the cloud for solar panels.
AnthonyMouse 7y ago
> From a software / devils point of view Nuclear power is the Mainframe - and solar or wind are the nice new micro services. The difference is that micro services don't all stop working at the same time when the sun goes down and the air is calm. Solar is great for running air conditioners. It generates when the demand is highest. But you still need something reliable to generate baseload. They're complementary, not replacements for each other. > If we have a crack in our solar panels, it ain't going to make the BBC news site. It won't, but maybe it should. Solar panels are full of toxic substances. As long as they're covered in glass and isolated from the environment that's no big deal, but break the glass and ignore it and you're leeching heavy metals and toxic chemicals into the environment. Things done competently are good, things done incompetently are bad. Nuclear is not different, solar is not different.
- einpoklum 7y ago
  Excess solar-based energy could be used to store energy that is later used in off-peak hours. Now, storage is a huge challenge (at least if you want it to be environmentally-friendly; otherwise - Tesla already makes a 300 MW-hour Li-Ion battery) - but it's coming along I think. About cracks in solar panels - those are: 1. Mostly localized (i.e. not a dust storm and such) 2. Doesn't produce a chain reaction - if a few panels are beginning to crack, you would isolate them and inspect the rest. Even if they've completely cracked it's not that bad.
  - AnthonyMouse 7y ago
    > Excess solar-based energy could be used to store energy that is later used in off-peak hours. Now, storage is a huge challenge (at least if you want it to be environmentally-friendly; otherwise - Tesla already makes a 300 MW-hour Li-Ion battery) - but it's coming along I think. "Excess solar energy" isn't really a thing. Due to the economics, there is no reason to even look at widespread storage until we already have enough solar capacity to satisfy the differential between daytime and nighttime demand. But just doing that is going to consume 100% of the available panel manufacturing capacity for at least a decade. To cover baseload as well isn't just a question of storage. First you would need to more than triple that amount of solar generating capacity. You'd need enough to satisfy not just the differential between the baseload and peak but the entire peak demand -- while charging the energy storage for nighttime use on top of that, including the conversion losses in both directions. Only then do you add the cost of storage, which currently isn't economical. It may become so in the future, or it may not, but that's ten or twenty years in any event before we can even start constructing it, much less finish. And there is a non-trivial chance that economical storage technology never materializes. > About cracks in solar panels - those are: > 1. Mostly localized (i.e. not a dust storm and such) Not really. In many cases it is a dust storm, or hail, or a tornado or hurricane or some other weather event. In other cases it's just time. You have a million panels and every year some percentage of them get cracked by wildlife or tree branches or weather or rogue humans. The damage accumulates. This is actually a significant issue for solar farms when they reach end of life. What happens then? Just abandoning them in situ would be a serious environmental problem, but the incentive to file for bankruptcy and do exactly that would exist once enough of them are damaged that the cleanup cost exceeds the value of the remaining panels. > 2. Doesn't produce a chain reaction - if a few panels are beginning to crack, you would isolate them and inspect the rest. Even if they've completely cracked it's not that bad. It's potentially worse than that, because solar installations tend to be more distributed and have less oversight, so instead of one big problem you have a million cumulative little problems. Which is actually harder to solve because a million little leaks are harder to plug than one big one, and you're adding coordination and distribution problems on top of the engineering problems.

mbell 7y ago

I wouldn't call solar 'micro', land use is one of its largest downsides. > If we have a crack in our solar panels, it ain't going to make the BBC news site. Actually cracking solar panels could be a rather large problem. e.g. Cadmium Telluride cells are gaining in popularity, if a tornado, large hail storm, or similar natural event hits a power plant built with such cells it could result in non-trivial release into the environment.

layoutIfNeeded 7y ago

Nuclear is like monolithic kernels: gets a lot of flak for ideological reasons, but it keeps powering the whole world. Solar is like microkernels: nice on paper but lacks the power for anything serious.

baq 7y ago
Eagerly waiting for the cloud for solar panels.

AnthonyMouse 7y ago

> From a software / devils point of view Nuclear power is the Mainframe - and solar or wind are the nice new micro services. The difference is that micro services don't all stop working at the same time when the sun goes down and the air is calm. Solar is great for running air conditioners. It generates when the demand is highest. But you still need something reliable to generate baseload. They're complementary, not replacements for each other. > If we have a crack in our solar panels, it ain't going to make the BBC news site. It won't, but maybe it should. Solar panels are full of toxic substances. As long as they're covered in glass and isolated from the environment that's no big deal, but break the glass and ignore it and you're leeching heavy metals and toxic chemicals into the environment. Things done competently are good, things done incompetently are bad. Nuclear is not different, solar is not different.

einpoklum 7y ago
Excess solar-based energy could be used to store energy that is later used in off-peak hours. Now, storage is a huge challenge (at least if you want it to be environmentally-friendly; otherwise - Tesla already makes a 300 MW-hour Li-Ion battery) - but it's coming along I think. About cracks in solar panels - those are: 1. Mostly localized (i.e. not a dust storm and such) 2. Doesn't produce a chain reaction - if a few panels are beginning to crack, you would isolate them and inspect the rest. Even if they've completely cracked it's not that bad.
- AnthonyMouse 7y ago
  > Excess solar-based energy could be used to store energy that is later used in off-peak hours. Now, storage is a huge challenge (at least if you want it to be environmentally-friendly; otherwise - Tesla already makes a 300 MW-hour Li-Ion battery) - but it's coming along I think. "Excess solar energy" isn't really a thing. Due to the economics, there is no reason to even look at widespread storage until we already have enough solar capacity to satisfy the differential between daytime and nighttime demand. But just doing that is going to consume 100% of the available panel manufacturing capacity for at least a decade. To cover baseload as well isn't just a question of storage. First you would need to more than triple that amount of solar generating capacity. You'd need enough to satisfy not just the differential between the baseload and peak but the entire peak demand -- while charging the energy storage for nighttime use on top of that, including the conversion losses in both directions. Only then do you add the cost of storage, which currently isn't economical. It may become so in the future, or it may not, but that's ten or twenty years in any event before we can even start constructing it, much less finish. And there is a non-trivial chance that economical storage technology never materializes. > About cracks in solar panels - those are: > 1. Mostly localized (i.e. not a dust storm and such) Not really. In many cases it is a dust storm, or hail, or a tornado or hurricane or some other weather event. In other cases it's just time. You have a million panels and every year some percentage of them get cracked by wildlife or tree branches or weather or rogue humans. The damage accumulates. This is actually a significant issue for solar farms when they reach end of life. What happens then? Just abandoning them in situ would be a serious environmental problem, but the incentive to file for bankruptcy and do exactly that would exist once enough of them are damaged that the cleanup cost exceeds the value of the remaining panels. > 2. Doesn't produce a chain reaction - if a few panels are beginning to crack, you would isolate them and inspect the rest. Even if they've completely cracked it's not that bad. It's potentially worse than that, because solar installations tend to be more distributed and have less oversight, so instead of one big problem you have a million cumulative little problems. Which is actually harder to solve because a million little leaks are harder to plug than one big one, and you're adding coordination and distribution problems on top of the engineering problems.