4 ms·

Now compare wind and solar and existing hydro and 2-5% w2e to just wind instead of just wind vs. just solar. At a mix appropriate for the region (for example so

Now compare wind and solar and existing hydro and 2-5% w2e to just wind instead of just wind vs. just solar. At a mix appropriate for the region (for example some regions get 800kWh/kWp/yr, others get 2400, wind ranges between 2000 and 5000), and include reasonable predictons of demand curves. Wind and solar are negatively correlated. Adding both reduces storage and makes the system more robust https://en.m.wikipedia.org/wiki/Negative_relationship https://en.m.wikipedia.org/wiki/Negative_relationship

robomartin 4y ago
> At a mix appropriate for the region We don't need to think regionally. We can move power thousands of miles. Yes, of course, it's easier over the ocean, however, we have the means to do so over land as well. For reference, a project from UK is moving power from approximately 4000 km away (Morocco). In other words, if wind --as I believe is the case-- has significant advantages over solar, we can locate wind wherever it makes sense and then move the power as far and wide as needed. We could spider out to various storage facilities and distribute to towns and cities. The point is, power is easy to move. We already do it. > Now compare wind and solar and existing hydro and 2-5% w2e to just wind instead of just wind vs. just solar I don't understand why you are asking for this comparison. My point is that grid-scale solar is a bad idea, by itself or mixed with something else, including wind. I gave most of the reasons in support of this conclusion in the comment you responded to. With solar we are starting with a technology that does not produce any power half the time, every day, forever. Tesla's Master Plan Part 3 (MPP3) figure 12 shows a close-up chart of a few days in July in the US. I reproduce it here for convenience: https://i.imgur.com/HTxWL6o.png https://i.imgur.com/HTxWL6o.png The variability of solar over time is massive. Not so for wind. particularly if located optimally and geographically distributed. In the US you cannot do any better with solar than that picture. Clouds, rain, snow and dirt will turn this: https://i.imgur.com/aNnbmDp.png https://i.imgur.com/aNnbmDp.png Into this nightmare: https://i.imgur.com/breTHQd.png https://i.imgur.com/breTHQd.png And so, reliability, capacity factor, availability, or whatever measure or term one might care to use for solar is always significantly inferior to that of wind. There's data on this. My question is rooted in optimization at many levels (list is not exaustive): - Energy production per installed GW - Land utilization - Ecological impact - Operation and maintenance costs - Storage requirements - Strategic supply chain factors - Scale of future waste/recycling problem Looking back at MPP3, table 6, we can ask a fundamental question: How much installed grid-sale capacity do we have to build? No new nuclear or hydro. The table, is reduced to onshore and offshore wind and solar. The total for those technologies is 5,087 GW. That's what we have to build. That's the challenge. My question is simple: We have to build 5,087 GW generation capacity. Given the optimization criteria list presented above. What combination of solar and wind would be optimal? Looking at the numbers, I cannot find justification for building almost any grid-scale solar. Let me give you a simple example of this... Solar requires at least four times more storage than wind to reach the same or similar reliability. I'll assume batteries. Well, that means at least four times more mining, processing, factories, CO2 emissions, damage to the ecosystem, transportation, installation and maintenance costs, mean time between failure, waste, recycling costs and effects, etc. And that's just one of my comparison vectors. Isn't that enough to at least question the decision? No? OK. The land-use ratio isn't something that can be ignored. The same 5,087 GW capacity would use about 2.5% of California if it was wind. Solar would require 30% of the State. Just think about what that implies in terms of every single factor that goes into building and maintaining such systems. This is why I argue grid-scale solar isn't making sense to me. Why use 12 times the land and over four times the batteries when you don't have to?
- Schroedingersat 4y ago
  ... You are willfully failing to understand the concept of correlation. Sun and wind are anticorrelated in most regions. Adding both reduces the storage needed. It doesn't increase it. Condescendingly stating that solar alone requires 12 hours of batteries does not make it any less true that solar + wind is superior to wind. Transmission costs money and raw materials and depending on it entirely (as an all wind or all nuclear solution would) makes the whole system brittle. Seeing as you are positing zero cost for transmission you can just put all of the solar in the least cloudy areas with 30% capacity factor and uncorrelated weather. Given the assumption of no transmission costs, try this tool: https://model.energy https://model.energy Using real costs for real projects that actually exist, and "dispatchable 1" to model existing hydro, you can solve for the optimal mix: https://model.energy/?results=3491f4124100114a1c3ba16966390326#solve https://model.energy/?results=3491f4124100114a1c3ba169663903... It almost always involves solar, because PV is the cheapest energy source by at least a factor of two over about 70% of the inhabited world. The smaller the region and the less long distance transmission, the more correlated the wind and the more solar is optimal. https://model.energy/?results=732e5306df0bde74aa22f57433e5883a#solve https://model.energy/?results=732e5306df0bde74aa22f57433e588... https://model.energy/?results=e0560182e98eef88fa715ec2a0b118b2#solve https://model.energy/?results=e0560182e98eef88fa715ec2a0b118... Plus tilting solar (which this tool does not model) in some regions matches the CF of wind. Some transmission, some solar, some wind, and some other/hydro will always be a better toolbox to have than just wind and transmission. Throwing solar away entirely because of an irrational response to the isolated results of a specific residential install with no transmission is just as crazy as trying to shoe-horn more LWRs than it is possible to fuel even once into the solution. Also you're not putting enough constraints on your land use estimates to make the numbers comparable in any coherent way. A high coverage fixed tilt system at low latitude has a coverage ratio of about 90% or 180W (nameplate) per m^2. A single axis system has a lower coverage ratio around 100W/m^2 but a higher capacity factor. You can't go using median CF if you're assuming transmission is free (what was that about how stupid people who assume averages for everything are?), you'd be putting all your solar in the southwest where capacity factors are 30% for 50% coverage ratios and the *net* energy per unit area would be higher than what you're assuming for nameplate. In either case your figures for land use are not consistent with your assumptions. Finally wind isn't infinite. You can't just go and find another TW of world class wind resource. There is plenty if used sensibly, but you can't magic more out of nothing.

robomartin 4y ago

> At a mix appropriate for the region We don't need to think regionally. We can move power thousands of miles. Yes, of course, it's easier over the ocean, however, we have the means to do so over land as well. For reference, a project from UK is moving power from approximately 4000 km away (Morocco). In other words, if wind --as I believe is the case-- has significant advantages over solar, we can locate wind wherever it makes sense and then move the power as far and wide as needed. We could spider out to various storage facilities and distribute to towns and cities. The point is, power is easy to move. We already do it. > Now compare wind and solar and existing hydro and 2-5% w2e to just wind instead of just wind vs. just solar I don't understand why you are asking for this comparison. My point is that grid-scale solar is a bad idea, by itself or mixed with something else, including wind. I gave most of the reasons in support of this conclusion in the comment you responded to. With solar we are starting with a technology that does not produce any power half the time, every day, forever. Tesla's Master Plan Part 3 (MPP3) figure 12 shows a close-up chart of a few days in July in the US. I reproduce it here for convenience: https://i.imgur.com/HTxWL6o.png https://i.imgur.com/HTxWL6o.png The variability of solar over time is massive. Not so for wind. particularly if located optimally and geographically distributed. In the US you cannot do any better with solar than that picture. Clouds, rain, snow and dirt will turn this: https://i.imgur.com/aNnbmDp.png https://i.imgur.com/aNnbmDp.png Into this nightmare: https://i.imgur.com/breTHQd.png https://i.imgur.com/breTHQd.png And so, reliability, capacity factor, availability, or whatever measure or term one might care to use for solar is always significantly inferior to that of wind. There's data on this. My question is rooted in optimization at many levels (list is not exaustive): - Energy production per installed GW - Land utilization - Ecological impact - Operation and maintenance costs - Storage requirements - Strategic supply chain factors - Scale of future waste/recycling problem Looking back at MPP3, table 6, we can ask a fundamental question: How much installed grid-sale capacity do we have to build? No new nuclear or hydro. The table, is reduced to onshore and offshore wind and solar. The total for those technologies is 5,087 GW. That's what we have to build. That's the challenge. My question is simple: We have to build 5,087 GW generation capacity. Given the optimization criteria list presented above. What combination of solar and wind would be optimal? Looking at the numbers, I cannot find justification for building almost any grid-scale solar. Let me give you a simple example of this... Solar requires at least four times more storage than wind to reach the same or similar reliability. I'll assume batteries. Well, that means at least four times more mining, processing, factories, CO2 emissions, damage to the ecosystem, transportation, installation and maintenance costs, mean time between failure, waste, recycling costs and effects, etc. And that's just one of my comparison vectors. Isn't that enough to at least question the decision? No? OK. The land-use ratio isn't something that can be ignored. The same 5,087 GW capacity would use about 2.5% of California if it was wind. Solar would require 30% of the State. Just think about what that implies in terms of every single factor that goes into building and maintaining such systems. This is why I argue grid-scale solar isn't making sense to me. Why use 12 times the land and over four times the batteries when you don't have to?

Schroedingersat 4y ago
... You are willfully failing to understand the concept of correlation. Sun and wind are anticorrelated in most regions. Adding both reduces the storage needed. It doesn't increase it. Condescendingly stating that solar alone requires 12 hours of batteries does not make it any less true that solar + wind is superior to wind. Transmission costs money and raw materials and depending on it entirely (as an all wind or all nuclear solution would) makes the whole system brittle. Seeing as you are positing zero cost for transmission you can just put all of the solar in the least cloudy areas with 30% capacity factor and uncorrelated weather. Given the assumption of no transmission costs, try this tool: https://model.energy https://model.energy Using real costs for real projects that actually exist, and "dispatchable 1" to model existing hydro, you can solve for the optimal mix: https://model.energy/?results=3491f4124100114a1c3ba16966390326#solve https://model.energy/?results=3491f4124100114a1c3ba169663903... It almost always involves solar, because PV is the cheapest energy source by at least a factor of two over about 70% of the inhabited world. The smaller the region and the less long distance transmission, the more correlated the wind and the more solar is optimal. https://model.energy/?results=732e5306df0bde74aa22f57433e5883a#solve https://model.energy/?results=732e5306df0bde74aa22f57433e588... https://model.energy/?results=e0560182e98eef88fa715ec2a0b118b2#solve https://model.energy/?results=e0560182e98eef88fa715ec2a0b118... Plus tilting solar (which this tool does not model) in some regions matches the CF of wind. Some transmission, some solar, some wind, and some other/hydro will always be a better toolbox to have than just wind and transmission. Throwing solar away entirely because of an irrational response to the isolated results of a specific residential install with no transmission is just as crazy as trying to shoe-horn more LWRs than it is possible to fuel even once into the solution. Also you're not putting enough constraints on your land use estimates to make the numbers comparable in any coherent way. A high coverage fixed tilt system at low latitude has a coverage ratio of about 90% or 180W (nameplate) per m^2. A single axis system has a lower coverage ratio around 100W/m^2 but a higher capacity factor. You can't go using median CF if you're assuming transmission is free (what was that about how stupid people who assume averages for everything are?), you'd be putting all your solar in the southwest where capacity factors are 30% for 50% coverage ratios and the *net* energy per unit area would be higher than what you're assuming for nameplate. In either case your figures for land use are not consistent with your assumptions. Finally wind isn't infinite. You can't just go and find another TW of world class wind resource. There is plenty if used sensibly, but you can't magic more out of nothing.