Aye, but op has misremembered. A Texas-sizes PV farm with 20% efficient cells, and assuming 25% duty cycle for night and winter, would produce ~35 terawatts.

http://www.wolframalpha.com/input/?i=texas%20area%20%2A%2010...

Worldwide power consumption is about 18 TW, electricity is about 3 TW.

And current PV looks like it’s good for 20+ years so you’re getting that power at about $3 trillion per year. The only reference I can find to current energy costs is this from Wikipedia:

"""In 2010, expenditures on energy totaled over US$6 trillion, or about 10% of the world gross domestic product (GDP)""" - https://en.m.wikipedia.org/wiki/World_energy_consumption

On that basis, you’d be getting roughly twice as many exajoules for roughly half as many USD.

Most PVs nowadays are guaranteed for 25 years and expected to last well north of 30 years. Some Chinese cos give 30 year warranties. BYD on its site also mention 50 year lifespans.

So we could halve our energy expenses ... but we aren't doing it? Why not?

We are doing it. PV is growing as fast as people can build factories and install panels.

(Also: Quarter of the expenses if you keep power use the same: It’s half cost for double the energy)

Ah, interesting. So are these PV workers in crazy high demand then? I feel like I haven't heard about the kind of wild growth in the PV industry that one would expect from such a significant endeavor?

Growth has been close to exponential since 1992, with an average doubling period of 2.2 years.

https://en.m.wikipedia.org/wiki/Growth_of_photovoltaics

(Someone should update the “Grid parity for solar PV around the world”, that map was last edited in 2015)

Also, you wouldn't have to power as many Texans. More for everyone else.

That's big, but so is the western hemisphere. 58 trillion USD to power the entire western hemisphere seems... Pretty reasonable? Assuming it depreciates over 30 years that's about 2 trillion per year, or "only" about double what the USA spends yearly on its military.

Spending 58 trillion on solar panels would probably induce some pretty good economies of scale, too. What kind of volume discounts do you get on a 15 digit purchase order?

And increase the base price of materials to outset gains but hopefully create a scale for materials that reduces price over time.

@mywittyname Nay, you MAKE the factories.

I assume at that point you just buy the factories.

At that scale you have to buy the factories that make the factories.

In 2010, expenditures on energy totaled over US$6 trillion, or about 10% of the world gross domestic product (GDP).

58T is the opposite of pretty reasonable.

Even if you could assume it wouldn't need any maintenance, which you can't, it's too much money for anything.

Not sure why you're being downvoted, you are exactly correct.

60T is 10x more than global energy expenditure in 2010, and that's just for the Western Hemisphere, which I'm fairly certain expends significantly less energy then the Eastern.

Edit: just checked, and I was correct. In 2010, Western Hemi energy demand was ~130 PBtu, while Easter Hemi demand was ~330 PBtu[1], or 2.5x more.

So at this same price point, it would be 210T to power the whole world with solar, or 50% more than global GDP in 2019 (but only using 2010 energy consumption numbers, so it's actually much more). And of course you'd need to keep adding more solar as demand went up.

[1] https://en.wikipedia.org/wiki/World_energy_consumption#/medi...

You wouldn't have to keep spending that amount every year though. A solar panel lasts for several decades and maintenance (especially at scale) is way less than a percent per year. I included 30 year depreciation in my calculations for a reason.

But is that maintenance lower than the existing maintenance burden?

Even then, this is still a wildly conservative estimate in terms of price. For example, you're also not taking into account any of the things that can't currently be easily electrified, such as passenger jets.

It'd need to be lower than the existing maintenance plus fuel burden of course, since sunshine is free but oil and coal are not. In any case, if you can't change the power source (as you say, passengers jets are pretty tricky to electrify) then you don't have to build solar panels for those so the upfront cost would go down.

Thinking about it more it would seem likely that as overall oil consumption falls, the relative cost of oil derivatives like jet fuel would go up because the advantages of scale decrease from what they are now. It's a pretty fascinating subject since solar prices seem to follow a Moore's law type price evolution at the moment while oil will get cheaper as well while demand decreases. Presumably there is a balance point somewhere?

> But is that maintenance lower than the existing maintenance burden?

While I haven't looked up the data yet, it seems plausible to assume that maintenance on photovoltaic electric generation is not only lower but significantly lower than thermal generation.

For example, having no moving parts in the generation process must be a large maintenance savings, though of course there must be unique costs associated with photovoltaics - cleaning, perhaps?

> Even then, this is still a wildly conservative estimate in terms of price.

I just did some quick calculations to check this statement and found it basically true - I estimate the cost of 100% PV solar generation at around $180tn 2019 USD for 30 years of capacity.

For reference, I used these figures:

173,340 TWh energy consumption in 2019 [1] * $35/MWh for utility-scale PV solar[2] * 1,000,000 MWh/TWh = $6.1t yearly energy cost for 100% PV energy production, or $182t over a 30-year lifetime.

This, of course, does not take into account the increase of energy consumption over that period which would raise costs, nor the economies of scale of this level of PV deployment which would surely lower costs, but as a BotE calculation it sounds about right and corresponds with your earlier estimate of $210tn for the same investment. Note also that the $35/MWh figure includes all operating expenses and amortized capital costs i.e. it takes all costs into account already.

However, as you imply with your first question regarding maintenance burden, the correct comparison is not "how much would it cost" but rather "how much would it cost relative to projected costs of energy" - and again per [2], utility PV solar is already cheaper than new utility thermal power generation. There is of course plenty of nuance when it comes to energy consumption - you point out, for example, that aviation will be a difficult sector to "electrify," a true enough statement in and of itself. However, it's pointed out in [3] that jet fuel represents 12% of transportation energy consumption and that transportation overall represents 25% of global energy consumption, implying that aviation only represents about 3% of global energy consumption.

Based on this, I speculate that aviation fuels can be produced in a 100% solar PV energy regime without increasing - and likely lowering - energy production costs above the current regime.

[1] https://ourworldindata.org/energy-production-consumption [2] https://www.lazard.com/perspective/lcoe2019 [3] https://www.maritime-executive.com/article/transport-uses-25...

I assume you wouldn't have to rebuild it every year.

I wouldn't call it small, I'd call impractical at best.

If thats small, how can you describe the size of nuclear power plants required to do the same job?

And power plants would be way smaller and aren't impractical.

Agreed. And covering that amount of land mass would have profound environmental impacts.

Solar efficiency has improved significantly since that stat as panels went from 13-14% to 18-20%. Which brings land area down to more like 10k square miles.

From wikipedia:

The plant will be able to store solar energy in the form of heated molten salt, allowing for production of electricity into the night. Phase 1 comes with a full- load molten salt storage capacity of 3 hours. Noor II, commissioned in 2018, and Noor III, commissioned in January 2019, store energy for up to eight hours.

Very cool.

https://en.wikipedia.org/wiki/Ouarzazate_Solar_Power_Station

In the context of powering an entire continent, it's still small.

I'm not at all familiar with this domain, but wouldn't this kind of farm be incredibly vulnerable to a foreign threat?

Indeed.

If the entire planet resided in TX, each person would get 1100 sq feet to themselves. That's a very large 1BR, good sized 2BR apt.