If your looking to buy a generator and tie it into the house grid for X$ and desire not to because your installing solar then that's real savings. For survivalists who want 1+ years of gas X gets rather large.

Even if you just have regular power outages, or are a data center, hospital, etc that is required to gave full power off grid for days solar becomes far more attractive. Solar is also rather atractive to the military as logistics get complex and expencive.

The examples in your second paragraph are not made-up scenarios; they're real needs caused by faulty power grids or large-scale disaster-management requirements. For those it makes sense to compare the costs of different off-grid power sources.

Your first paragraph, and the post I replied to, aren't like that. Is someone looking to buy a generator because they just want one, or "something may happen", or do they actually suffer from power outages? As for survivalists, most of what they do is based entirely on their beliefs and world-view rather than verifiable facts, so they're a perfect example of what I'm talking about. (I'm not saying they shouldn't do what they're doing; so long as they're not hurting anyone I think they should be free to believe what they want and act accordingly. I'd just like them to be honest, with themselves and others, about why they're preparing to be self-sufficient.)

BTW, I'm pretty familiar with the faulty power grid scenario. Even in my very densely populated neighborhood all of the houses on my street and one or two others used to loose power during every big storm, because for years the electric company just patched the lines that kept getting knocked down by tree movement instead of fixing the problem. Some of my neighbors had generators, and it became common to have a network of extension cords criss-crossing the street for a few days after a storm. Solar would have been pretty useless, since it's usually not sunny during big storms.

I recently came back from a year in Puerto Rico with a similar situation. The utility company didn't do proper maintenance and the electrical system uptime was about 93% every month.

With the grid mostly being up, but not being reliable, the most logical solution was a gas generator to provide power during the outages. All the gas stations had backup generators, so the general operating procedure was to just keep a 5 gallon gas can handy. If the power was out for more than 12 hours, we just went and got more gas. The frequent outages meant that the gas was used often enough that it wouldn't spoil.

I'm moving up to upstate New York soon and my experience in PR has made me wary of multi-day outages during blizzards, so I'll likely replicate the setup.

If I ever live in a place with natural gas lines but worry about power outages, I'll likely get a natural gas generator. Even during power outages, natural gas lines stay pressurized because the gas distribution system burns the gas itself to power the equipment they use to boost the pressure for the system.

While I don't think it's a reasonable need, people really do build under ground shelters etc. Also, solar works on cloudy days if you can see there is light that can be turned into electricity. Worst case where there is enough cluds it's basically dark you get ~10%. What stops working is concentrating solar aka those big plants with lots of mirrors.

PS: Even just moonlight can provide very tiny ~1/20,000th, but measurable amounts of power.

For my particular experience, that still wouldn't have helped. First, the storms that knocked out power mostly occurred at night and had 100% cloud cover, so we're talking about total darkness. There was also really heavy rain or snow, which I'm sure makes the panels less effective even when there's light hitting them.

10% or less power wouldn't be enough, I think. The goal would be to keep the refrigerator and some lights running, and during the winter my heating system needed power for the electric ignition on the gas-burning boiler. (Why those things don't have a backup manual spark ignitor, like a gas barbecue, is beyond me.) During the storm this would all be without power, though I guess solar would help during the day once the storm passed, so long as the panels weren't covered with snow.

10% is generally the extreme low end and not something your going to see repeated for several days.

If you want days of off grid use your going to have battery's anyway. If your willing to cut back on energy that much then you might not need to even clear the snow from your panels. But, there is a wide range of easy solutions to that from manual clearing with a long pole, heating elements, or even just a steep mounting angle.

One major upside is the system should be automatic so you can leave your house unattended through a major storm. However, from a cost perspective solar hot water systems are generally worth it long before solar panels in the north.

Unless you're running that datacenter or hospital only part of the day, you're still going to have a fuel-based generator in your scenarios.

Battery's work ok for regular short term power outages, but going 2-7 days after a storm is not that uncommon every few years in many areas. At which point solar for charging in the day + battery backup is a good solution vs grid + generating capacity.

Basically, installation costs with solar are high, but they don't increase much as you scale the system. Battery's let you scale further, reducing price per watt. You also need an inverted for battery's which you can reuse with solar.

Can you point us to a hospital or commercial datacenter that has solar and batteries and no fuel-based generator?

You would need solar generating capacity well in excess of demand (to serve the daytime demand AND fully charge the batteries on an overcast day to address a multi-day outage). That seems far less practical than a mechanical generator as backup.

Storing a year of gas is just stupidity, even for a crazy survivalist.

Less insane use cases for generators are a week or two, mostly to keep heat, well pumps, and refrigeration up. A generator for that purpose is about $3k installed, and goes up with output.

The off grid use case is a fringe thing, because most if not all states require grid connections for residential construction.