Comment by g8oz | Hacker News Reader

Comment by g8oz | Hacker News Reader

Parent Full thread

Further to your point - South Korea built 4 reactors for the UAE at the Barakah site. Took 11 years from deal to first grid connection and cost $25 billion for 5.3 GW of nameplate capacity. It's expected to cost another $20 billion for operation, maintenance and fuel

https://en.m.wikipedia.org/wiki/Barakah_nuclear_power_plant supply.

Contrast that to the Al Dhafra PV project also in the UAE - $1 billion for 2 GW of nameplate capacity. One can apply whatever discount is wanted to account for solar's shortcomings and it still comes out ahead when planning an energy generation portfolio. Further away from the equator a similar case can be made for wind.

The Barakah reactors will be generating 5.3 GWh from 6 pm to 8am which of course solar can't do. In that time it will deliver 14 hours X 5.3 GW ~= 75 GWh of electricity.

Capital costs for 4 hour utility scale lithium storage are expected to be $250 million per GWh by 2025. https://www.nrel.gov/docs/fy21osti/79236.pdf

75 GWh x $250 million = ~$19 billion to get those gigawatt hours from storage.

The nameplate capacity of the solar plant to generate 75 GWh for night time use in the hours from 8 am to 6pm is 7.5 GW. To prove my point let's double that to 15 GW.

The cost of that 15 GW solar plant, based on the Al Dhafra PV plant would be $7.5 billion. Another 5 GW solar plant to deliver the same day time GWh that the reactors do...let's double that to 10 GW so another $5 billion.

So my back of the napkin calculations say that the cost to replace the Barakah reactors with solar + storage are: - $19 billion for Li-ion storage - $7.5 billion for a solar plant to charge that storage for nighttime use. - $5 billion for a solar plant to deliver (more than) equivalent daytime use.

$31.5 billion in total. Very much competitive with the nuclear option taking into account a) time to completion / general project risk b) the falling cost of storage plus the development of new options like sodium ion storage c) radioactive waste risks, solvable as they are d) security concerns - Houthi rebels fired a few missiles at the Barakah site a while back. e) construction and operational workforce skill requirements

To make the costs more comparable, in the report he mentions, they estimate 2.5% per year of the installed battery capacity cost as the ongoing maintenance costs. And, the batteries have an 85% efficiency rating meaning they'd need 92 gwh of installed battery capacity to meet the 75 gwh demand, so tack on an additional $4.25 billion so the battery cost which is now $23 billion.

Based on the lifetimes of these South Korean plants(https://en.wikipedia.org/wiki/Nuclear_power_in_South_Korea), I'm just going to use 40 years as the lifetime for these calculations as I couldn't find any numbers for how long the UAE expects that plant to be operating regarding that operational costs of $20 billion which gives us:

Maintenance costs for solar of: 40 years x $23 billion x 2.5% = $23 billion

So now we're at a total current cost of $46 +7.5 = $53.5 billion for Solar vs. $45 billion for Nuclear.

Edit: Forgot the actual cost of the solar array that G80z said was $7.5 billion

The amount of solar/wind paired storage drops significantly if you take into account wind energy's anti correlation with solar, demand shaping and overproduction.

It isnt just a matter of getting 2GW of solar and getting enough batteries to supply 2GW through the night.

Worth pointing out here that nobody is planning to do large scale 12-hour storage with lithium ion batteries.

Even leaving out better battery chemistries for the moment, you can use pumped hydro (even with seawater - some of the UAE’s highest mountains are near the coast) and thermal storage (you might think you don’t need it in such a hot place but you’ve still got to bake bread etc etc).

Word on the street is that locations for pumped hydro are, for a variety of reasons that vary by locale, pretty much maxed out already.

Pumped hydro doesn't require mountains, you can also dig an underground reservoir.

Pumped hydro requires a large volume of fluid and a height difference.

Valleys are useful as they provide a large volume that can be capped with relative ease.

An underground reservoir requires a large volume to be excavated (or a pre existing cave system) and a deeper chamber again for the turbines to powered by the fall .. to yet another deeper underground reservoir.

I'm not saying this is impossible, I stating that as an engineering venture (actual civil engineering not software "engineering") there are some real challenges and greater ones than using existing topography.

You're hand waving a lot away with your comment.

To be honest I'm pretty doubtful the economics of excavating a reservoir are going to work. It needs to be very big and very deep, which is going to be very expensive unless you're planning to go all Edward Teller (don't go all Edward Teller).

You're mixing it up with hydroelectric dams which need a river. Pumped hydro needs two close bodies of water at different heights. That geography is very very common.

> That geography is very very common.

Could you provide some more info for this? My understanding was a lot of the sites that could do this, have done. If that's not the case it'd be good to see what the real picture is.

There's no point if you don't have energy to store. Only 1% of potential sites need to be developed

https://re100.eng.anu.edu.au/global/

Probably won't happen because batteries will be even cheaper by the time much of it is developed.

That link is what makes me concerned when talking on these topics. Far too much certainty and trust seem to be derived from and placed in such studies.

In this case, from the study:

> None of the PHES sites discussed in this study have been the subject of geological, hydrological, environmental, heritage and other studies, and it is not known whether any particular site would be suitable. The commercial feasibility of developing these sites is unknown.

> makes me concerned

There it is. Always so concerned" about whatever imaginary downside alternatives have, but completely credulous about nuclear which has failed over and over again.

You're claiming without any evidence whatsoever that at least 99% of geographically appropriate sites are unusable.

> Always so concerned" about whatever imaginary downside alternatives have

It's not a downside. It's a problem with the study.

> completely credulous about nuclear which has failed over and over again

Can you cite where I've done this?

> You're claiming without any evidence whatsoever that at least 99% of geographically appropriate sites are unusable.

And where I've done this?

yes and no. we've put dams in the obvious places, but we're mostly using them as generation rather than storage. switching them to be run as batteries will massively expand the amount of grid storage.

Pumped hydro is storage by definition, is it not?

A dam in a river can either be used as generation (if you always run it) or as storage if you vary the output. Most pumped hydro is already storage, but most hydro currently isn't pumped.

And it’s worth pointing out here that conventional hydro is being optimised in some places for a world with a lot of wind and solar in the grid.

In a nutshell, you put in more generators to maximise your peak generation capacity, and operate them only when required.

Ah, understood. But this isn't to do with renewable power storage at this point, just using artificial lakes as separate peak power storage?

Outside of the enormous engineering pitfalls of dealing with seawater, you would even get a slight boon in energy storage because it is 2-3% more dense than pure water.

The math here seems pretty well reasoned. But it assumes that the price doesn't change as the amount of materials increases. Can an organization just write a $10b check and get the batteries? Does anyone just have the stock of them sitting around?

I'd have to imagine not. And that it might take as long to get all the batteries built and installed as it does to build a nuke plant.

If you cut a $10 billion check, I think you can negotiate yourself to the front of the line. Also, we are talking about the UAE - $30 billion for energy is a significantly better investment than a large number of Softbank bets.

Total global battery production seems to be 200-300GWh/year. If you want to buy approx 30% of total yearly output it's very unlikely you can negotiate your way to the head of the line. In fact, buying that much might raise global prices for batteries.

If I want enough solar panels and batteries to make my house off grid I can just write a check and the stuff shows up in a few weeks. But that doesn't scale upwards without limit. Eventually the limits on global supply kick in no matter how much you want it.

Nobody said the installation had to happen overnight. A five year plan would be impressive and still feasible.

Reply on news.ycombinator.com