>Solar panels need to be replaced after 30 years max, but most probably will be replaced much earlier because the operators want to pocket the improved efficiency of newer panels. Reactors will need maintenance but it is normal to keep them running much longer. Also, what about the human labor? To maintain such large solar array you'd need hundreds of people. A modern nuclear powerplant with 5+ reactors is maintained by few dozens of people.
To just pick at one of your "statistics": E.g. Flamanville, a reactor plant with 2 reactors (and 1 being built) employed 670 people (not a "few dozens")[1]. I would argue the solar panels are the ones needing much less people. The ones I know don't need any kind of big servicing. But in those sizes this is pure speculation from my side.
> To replace just a single nuclear reactor you need so many solar panels that the deaths caused by the transportation will be significant - accidents, pollution + you have to count the damaged roads that need to be replaced - again pollution and accidents. And high-voltage electricity is not exactly safe - you would need so much electrical work that there simply must be a lot of death from accidents too, on this scale we're getting into the area where it's statistics and not really something to simply prevent.
Electronics-wise I don't see why there is so much more (and more dangerous) work involved in wiring up solar panels compared to constructing a nuclear plant. But I don't have statistics on that. I also wanted to know if your "deaths caused by the transportation will be significant" makes any sense. For that I tried to estimate the kind of size a solar park would have be to replace a reactor plant, e.g. Flamanville. Flamanville produces 14TWh per year. I took [2] as a reference(2.14km², 117GWh) and it came out to 256km² of area to produce the same amount of energy per year with solar in the flat part of germany. Storage is missing there. This is a huge amount of space, but also nothing that will produce "significant accidents" on its own. Constructing reactors and transporting fuel rods and nuclear waste is also a huge act with tons of material involved.
> It will never be possible to phase out fossil fuels in transport and industrial uses without nuclear power - solar just doesn't scale this much, it's perfect for individual houses but you just can't power thousands of large container ships with it, nor blast furnaces etc - and definitely not the uncountable amount of trucks and cars on roads.
Maybe direct batteries are not a solution for transport and industrial uses. But green h2 would be. This then has the fortunate aspect that it can be produced anywhere and independent of the need, e.g. in the Sahara and only when the sun shines. In that way solar can replace fossil fuels and nuclear power. Regarding cars and regular power use (btw germany covers ~40% of its energy with renewables already, taking this times two or three is not an issue of "scale"), lets just take the solar example from above and try to do 150% of germany's current power needs in full solar and see where that gets us. Germany needed 624 TWh in its peak year (2007, [3]), lets be generous and say we need 1000TWh in the future. This comes down to 18300 km². Germany has an area of 358000km². So our (very generous) solar capacity would need about an area of 5%. How much is 5%? According to [4], citys and streets (etc.) sum up to 14.5%, Agriculture has about 50%, Forests 30%. Since we can "simply" put the panels on top of the first category and don't have to replace existing areas most of the time, this looks pretty feasible to me, even with exaggerated demands in napkin calculations. So maybe it does not scale well, but it does scale.
And all of this does not take into account that solar is not the only option. Power from wind is a great way to mix since wind is strong mostly when there is no/less sun.
To just pick at one of your "statistics": E.g. Flamanville, a reactor plant with 2 reactors (and 1 being built) employed 670 people (not a "few dozens")[1]. I would argue the solar panels are the ones needing much less people. The ones I know don't need any kind of big servicing. But in those sizes this is pure speculation from my side.
> To replace just a single nuclear reactor you need so many solar panels that the deaths caused by the transportation will be significant - accidents, pollution + you have to count the damaged roads that need to be replaced - again pollution and accidents. And high-voltage electricity is not exactly safe - you would need so much electrical work that there simply must be a lot of death from accidents too, on this scale we're getting into the area where it's statistics and not really something to simply prevent.
Electronics-wise I don't see why there is so much more (and more dangerous) work involved in wiring up solar panels compared to constructing a nuclear plant. But I don't have statistics on that. I also wanted to know if your "deaths caused by the transportation will be significant" makes any sense. For that I tried to estimate the kind of size a solar park would have be to replace a reactor plant, e.g. Flamanville. Flamanville produces 14TWh per year. I took [2] as a reference(2.14km², 117GWh) and it came out to 256km² of area to produce the same amount of energy per year with solar in the flat part of germany. Storage is missing there. This is a huge amount of space, but also nothing that will produce "significant accidents" on its own. Constructing reactors and transporting fuel rods and nuclear waste is also a huge act with tons of material involved.
> It will never be possible to phase out fossil fuels in transport and industrial uses without nuclear power - solar just doesn't scale this much, it's perfect for individual houses but you just can't power thousands of large container ships with it, nor blast furnaces etc - and definitely not the uncountable amount of trucks and cars on roads.
Maybe direct batteries are not a solution for transport and industrial uses. But green h2 would be. This then has the fortunate aspect that it can be produced anywhere and independent of the need, e.g. in the Sahara and only when the sun shines. In that way solar can replace fossil fuels and nuclear power. Regarding cars and regular power use (btw germany covers ~40% of its energy with renewables already, taking this times two or three is not an issue of "scale"), lets just take the solar example from above and try to do 150% of germany's current power needs in full solar and see where that gets us. Germany needed 624 TWh in its peak year (2007, [3]), lets be generous and say we need 1000TWh in the future. This comes down to 18300 km². Germany has an area of 358000km². So our (very generous) solar capacity would need about an area of 5%. How much is 5%? According to [4], citys and streets (etc.) sum up to 14.5%, Agriculture has about 50%, Forests 30%. Since we can "simply" put the panels on top of the first category and don't have to replace existing areas most of the time, this looks pretty feasible to me, even with exaggerated demands in napkin calculations. So maybe it does not scale well, but it does scale.
And all of this does not take into account that solar is not the only option. Power from wind is a great way to mix since wind is strong mostly when there is no/less sun.
[1] https://en.wikipedia.org/wiki/Flamanville_Nuclear_Power_Plan...
[2] https://en.wikipedia.org/wiki/Templin_Solar_Park (https://de.wikipedia.org/wiki/Solarpark_Templin_%E2%80%93_Gr... has the energy statistics regarding GWh in 2017)
[3] https://www.umweltbundesamt.de/daten/energie/stromverbrauch (german, but the graph should be readable)
[4] https://www.umweltbundesamt.de/daten/flaeche-boden-land-oeko... (also german)