I'm not sure what the top poster's "plan" is, but if we wanted to switch to significant amounts of solar+wind, that would require batteries costing thousands of trillions. Practically impossible.

However if he meant planting trees and switching the world to nuclear, that might be doable for a few trillion.

Every government project goes over budget, so it's probably much higher than that.

Li-ion batteries cost ~$145/kWh (new), can last >4,000 cycles. That’s <$0.03625/kWh when adjusting for product lifetime.

World electricity production is ~5 TW, assume that you need to store half of it to cover nighttime (seasonal adjustments exist, but I think that’s more a problem for production than storage when the storage itself is effectively a stockpileable consumable), 2.5 TW-years of battery production per year @ $0.03625/kWh = $729.4 billion/year worldwide.

And that’s even ignoring things like a worldwide power grid as an alternative to batteries, as I agree that government projects are often over budget: batteries can be done at a household scale and every kWh helps.

I really should write up a long form blog post about this, these conversations keep circling around the same points.

That's still a pretty pessimistic view of what the costs would actually be like. We've got pure renewable generation costs that are now pushing $0.015 LCoE. While high efficiency batteries (90% efficient round trip) might cost around your estimate, it would actually be much cheaper to use other forms of energy storage that might be less efficient, but then overprovision on generation to make up for the loss in efficiency.

Yes. My calculations assume the wind does not blow at night, that night-time power use is the same as daytime power use even though most people will be asleep, and that storage costs will not continue to decline despite anticipated improvements.

It also has the imperfect assumption that storage becomes totally worthless after 4000 cycles instead of gradually degrading, but that degradation starts at the first use and I don’t know the curve so I can’t integrate it to find the total lifetime stored energy, which might be higher or lower than the simplified model I’m giving here.

Taking your numbers for granted, that would be just the ongoing replacement costs, not the capital costs.

Current capital costs are about $200/kWh. Daily global energy use is about 400 TWh. That would be an 80000 trillion dollar battery.

The number is so large we just can't build it, so it's actually infinite. It also doesn't take into account land costs, because these batteries will take huge amounts of space. It's also only for 1 day, and there will surely be multiple days where wind and sun are below 20% output, so you'll need to accept periodic blackouts.

1. It’s $145/kWh

2. You’re counting all energy use, I’m just counting electricity

3. $200/kWh * 400 TWh is $80 Tr not $80,000 Tr

4. That’s not “just one day”, it is >4,000 days because the battery can be recharged that many times

5. Even if you forced everyone to accept blackouts until this would be completed (won’t happen), they would occur at night.

You're right about #3 - it would be hundreds of trillions, not thousands.

6. Your 400 TWh number is for 24 hours, not “simultaneous no wind and nighttime”, so even if you include all power and not just electricity (even though power intended for non-electrical use can be stored in other things than batteries) you’re now talking $40 Tr, worldwide, spread over however many years you want to build this collection of batteries.

Except that’s $29 Tr when using the cheapest currently available batteries that I’ve seen.

7. The lifetime cost of the battery and a PV cell is cheaper than the fuel in a coal plant even when the cost of the coal plant itself has been written off, so each incremental step of that combination is still profitable compared to that particularly fossil fuel.

I’m not sure about the cost of wind or other fossil fuels, but the combined cost of PV+battery that anything else has to beat is 5¢/kWh.

EDIT:

8. Land costs. The energy density of LiIon batteries is “250-693 Wh/L”. A high efficiency 1 m^2 sun-tracking PV panel with the lowest energy density battery would produce up to 400 W * 12 hours = 4800 Wh = 19.2 litres = a layer behind the PV panel of 1.92 cm thickness. Why is this supposed to be a criticism?

> that would require batteries costing thousands of trillions.

Hilariously disconnected from reality. That would be tens of petawatt-hours. The world uses tens of petawatt-hours of electricity annually.

> That would require batteries costing thousands of trillions.

Overprovisioning is a much cheaper solution.