[ben_w]

> our current battery tech can’t improve fast enough to supplant fossil fuels in the time frames needed.

I disagree. The tech itself already good enough to supplant the majority of cases, which in turn gives us more time for the things that remain (such as long-haul aircraft).

That said, I may be a little on the optimistic side about how much warming the ecosphere can take. If it's already too hot, then yes, naturally you are correct.

> Interesting that you mention fusion though considering fission is available today and provides a substantial amount of power (not to mention actually reduces the amount of fossil fuels whereas solar has a negligible impact on fossil fuels and at best is only absorbing energy growth).

That's not what the graphs show: https://ourworldindata.org/electricity-mix

• Coal: down since 2012

• Gas: close enough to steady since 2012

• Nuclear: down since early 2000s

• Wind and solar: up

Looks to me like gas mostly replaced oil (since the late 90s); and that wind+solar is displacing nuclear (since the former became big enough to show up on a graph).

[pyrale]

> That's not what the graphs show

The graphs you provided show that coal and usage are still growing in absolute numbers. They're only going down in perentages. So, aside from oil (which is mostly still there), nothing was displaced.

The only significant thing we learn is that we've doubled our electricity usage since 2000. The share of low carbon electricity generation barely moved since 1985. Renewables just helped avoid it crumble due to hydro not being scalable.

[XorNot]

If batteries were literally free today, they still wouldn't be good enough due to the cost in support electronics required to put them on the grid (which is currently about price parity with batteries per watt).

EDIT: and those electronics also degrade - a lifespan of 20 years would be reasonable at scale.

[svnt]

So you’re charging twice for the grid tie? Because the reason for installing batteries would be 1) to not grid-tie, or 2) to grid-tie along with a renewable energy source, which if it is a commercial site, is already grid-tied.

[XorNot]

It's the same thing: the inverters and support electronics needed for batteries currently cost, per watt, about the same as LiFePO4.

You can't run "bare" LiFePO4: you're either forming a grid, or you're connecting to one. Both involve BMSes and inverters.

[jaggederest]

But that's already cheaper than building new coal or natural gas plants. The cost per watt of solar plus battery, right now, is the cheapest available form of power. I don't understand why you don't think that it's going to replace a majority of primary and secondary energy usage on the planet relatively quickly.

I'd bet a dollar that, in 50 years time, nearly all energy usage is going to be primitive biofuels or solar-PV-origin.

[XorNot]

The sticker price of any given power source is irrelevant: the question is how much does it return on the investment.

If building a solar installation is cheaper per kW then building a gas generator, that literally doesn't matter if the only times the solar installation generates power is when power prices are negative.

[jaggederest]

Including batteries. It's a peaking plant in a can. As far as I can tell the all-in non-land costs are cheaper, full stop.

I also did the analysis for new nuclear under a relaxed regulatory regime (i.e. substantially cheaper and faster than now) and there's no way it wins. For the price of a gigawatt of nuclear, you can get 5 gigawatts of solar that's online next year, plus half a gigawatt of battery.

I could be wrong, I'm just an armchair economist on this stuff, but I just don't see how it makes any economic sense to build anything but solar unless you're located somewhere remote and arctic (i.e. Åland or something)

[svnt]

That’s a cost on a renewables installation though, it doesn’t need to be loaded on the batteries. You aren’t installing bare solar panels either.

[XorNot]

The hardware to manage batteries and the hardware to manage solar panels is different (but sometimes can be packaged into the same box, or with one-off benefits like hybrid inverters), with batteries being considerably more complicated. You would struggle to ruin a solar panel with a bad inverter. You can completely destroy a battery by over-discharging it once.

A solar panel produces energy. A battery only stores it (and loses, round trip about 8% in the process) - which is to say, batteries are solely arbitrage instruments.

[svnt]

You started out loading the whole cost for a grid tie. Now you are talking about a BMS. A BMS is typically included in the cost of batteries.

Batteries are extremely functional in many installations, and even if you’re not using, only selling, arbitrage can work well. This is especially true if you get paid to accept the commodity in one time window and can get others to pay you to take it later.