Here's some rough numbers, hopefully i haven't made any mistakes:

Total US energy generation 2021: 4 trillion kwh

Average power generation: 457Gw

Cost of Topaz solar farm: $2.5bn Potential output: 550 Mw Cost per Mw $4.5m

Spend a trillion on similar capacity, you get 220Gw potential output.

But that's assuming you get 550Mw average. In fact, it produces about 1282Gwh output a year, so average is more like 145Mw.

So at a rough estimate, you need more like 10 trillion to meet US electricity demand. And that assumes you have unlimited battery storage for free.

And of course electricity is a small minority of total energy use.

Incidentally, I think it's amazing you get to about 10% of total electricity needs with only $1tr, i.e. around 4% of US national output.

https://en.wikipedia.org/wiki/Topaz_Solar_Farm https://www.eia.gov/energyexplained/electricity/electricity-...

```python trillion_dollars = 1e12

mw = 550

cost_farm = 2.5e9

cost_dollar_per_mw = cost_farm/mw

num_mw = trillion_dollars/cost_dollar_per_mw

num_gw = num_mw/1000

num_gw

kwh = 4e12

mw = kw/1000

gw = mw/1000

gw ```

Don’t forget that, 1 trillion can get you lots of leverage and cost benefits In installation and solar panel production. I’ll do it for 1 trillion.

> Cost per MW: $4.5m

That's quite high. Here in Germany they assume 530-800€ per kW_peak for a utility-scale ground-mounted system (2021) [0]. You can add additional CAPEX for inflation and expansion of the grid, but <$1500 per kW_peak (aka $1.5m per MW) should be quite possible. Especially if you include scaling effects.

If you add batteries for night-time balancing, LCOE roughly doubles for now.

Economically, no other energy source will beat non-winter day-time PV in the foreseeable future. Imo, as long as other plants are running during those times, investing in PV is a no-brainer.

[0] https://www.ise.fraunhofer.de/en/publications/studies/cost-o...

No you can't because solar isn't a continious source and fusion is.

No amount of investment into solar power can make the sun magically shine 24 hours a day.

Of course you can invest enough and make solar equivalent to continuous power.

Solar + reverse hydropower, solar + synthetic fuel manufacture, solar + appropriate sized batteries, solar cranes-with-weights, etc.

Solar is so cheap that all those are plausible. Fusion is still far from ready.

I suggest we do not lay all of our eggs in one basket. Climate change can be best countered if we use the sweet spot of all current available and anticipated technologies, including photovoltaic, long & short term storage solutions, and fusion. There is no need to bet on one technology at this stage.

There is a market need.

Solar panels are cheap and the market appears to be fine with unreliable power[0] if it's cheap enough.

[0] Unreliable, but predictable that is. You can estimate with decent confidence what the weather will be like tomorrow and you can definitely tell what it's going to be a few hours from now, so energy auctions a day ahead are feasible.

Do you have week long blackouts currently? In any given year you can count on a week where a panel only produces 10%. It just has to be cloudy or rainy. However much solar we add, even with storage to get through a night, needs some other source to get through the cloudy week, unless we are ok with blackouts.

A panel. Panels spread over hundreds of thousands of square kilometres produce power averaged over that area, so they're much more consistent and predictable.

HVDC lines help with this a lot. It has gotten to a point where there are serious plans to build a long, undersea HVDC line from the UK to Morocco which, get this, is poised to cost less than the equivalent(GWh delivered annually) Hinkley Point C nuclear power plant:

https://xlinks.co/morocco-uk-power-project/

That's how cheap the combination of solar and HVDC is.

Glad it will work there. There has been local opposition to new transmission lines in the USA (see Maine recently) so when folks say “just build powerlines” all I can think of is the difficulty of the land assembly and permitting. Technically it is easy, it is all the humans in the way that make it hard.

I mean, you could literally build a ring of solar power collectors and microwave transmitters in orbit. It would always have 50% coverage.

Or just have a bunch of power lines running east-west.

Energy storage is not an unsolved issue.

Batteries

<during the day>. $1T doesn't buy nearly enough batteries for the storage we would need.

We should be installing more solar & wind, but that alone is not a solution to our problems.

Fusion can give us 10x our energy need. Imagine a world where energy is simply not a blocker for anything.

Solar can't do that without massive batteries that don't exist. Even then it's not viable in areas where there isn't sufficient sunlight.

The batteries do exist, but it’s better to use them in cars. If we start diverting our battery supply for grid storage it will drive up the price of EVs.