I strongly believe in the future of solar power in general because it's relatively low-tech and reliable. There are no moving parts (other than maybe turning an array to face the Sun while it moves across the sky but this isn't strictly required). So solar power can provide power in a lot of places where it's either impossible or infeasible to build infrastructure (eg remote places, war zones).

But once solar power source (or something else) is cheaper than burning oil, gas or coal you can use that power to make fossil fuelds by sequestering carbon from air. That's not that difficult but we don't generally do it because it makes little economic sense. But this method would make fossil fuelds essentially carbon neutral.

So there are vehicles that won't make sense to be replaced with EVs. Likewise, there are places where burning fuel may still make snese. Solar power may not make sense in jungles or hurricane zones so there may be continued use of generators. I can't speak to the specifics of Rarotongo. In general though, I think the technology and infrastructure will be truly transformative on a massive scale.

I was at a radio transmitter[1] on the West Coast of Australia the other day. It's the largest low frequency radio in the Southern Hemisphere, used by subs and military stuff.

The facility, which is 6 kilometres from a regular town, and has three diesel generators producing 18,000,000 watts of power, burns through 26,000 litres of diesel PER DAY. It's been doing that 24/7 since 1967. (There was a sign proudly stating as much)

We do a TON of stupid stuff, regularly.

[1] https://en.wikipedia.org/wiki/Naval_Communication_Station_Ha...

Now, how expensive - and how big - would an equivalent solar array be? Don't forget maintenance costs, including technicians.

Once all that is cheaper and just as reliable as their generators, only then will there be a discussion on replacement.

But in general, I agree that it should be done.

1. The inverter and other electrical systems do fail (surprisingly frequently).

2. Solar is very spread-out: maintaining a geographically large facility is not trivial (vegetation, water drainage, cables etc.)

3. Batteries also require maintenance, and are an inherent part of any independent PV installation (if the site relies on the grid, it is no longer independent).

For an easy point of comparison, there actually is an 18MW solar array under development in western Australia as part of phase 0 of the Yuri project [0] (the goal of that project is to produce hydrogen for ammonia production via electrolysis instead of steam methane reforming). This is apparently around 23 hectares (57 acres), at a cost of A$24-33m in capital expenditures, and A$0.5-1.5m in annual operating expenses [1].

For comparison, let's suppose diesel is A$2 / liter. 26,000 liters a day, for 365 days a year, is A$19 million, every year, just for fuel costs. Wait, you might argue, fuel prices are inflated this year. Okay, historically it might have been closer to A$1/L, for about A$10 million per year. Breakeven is therefore somewhere in the ballpark of 2-4 years, which is absurdly low for a project of this scale.

[0] https://gateway.icn.org.au/project/4938/yuri-phase-0

[1] Page 28-29 of https://arena.gov.au/assets/2020/11/engie-yara-renewable-hyd... -- capex estimate is A$70m. If you remove the capex of building the H2 plant, you're left with A$33 million. If you just consider the PV component (34% of the overall capex estimate), then you get A$24m.

Surely a 6km cable is the solution?

It's not like hospitals and data centres refuse grid connections and run entirely on the backups.