Thread by @mtrw85

THREAD! So, in a debate with @CJPomfret &@TheAusInstitute's @RDNS_TAI, I was asked to make a case that wind energy is likely more expensive than nuclear, if they have to deliver the same service. Here's a fun walk-through with rough calculations from my beautiful homeland!

Before focusing on wind, let's look at the big picture power for generation in Aus, right now! Just to get us in the mood, note that the mean wind has sort of died currently and the sun has set (red). So you'd need storage if you didn't have fossils (blue).

To give us an idea what wind is like, here's last month. It's the total over all of Australia. As you can see, it hovers around 2 GW of energy. For comparison, that's about what one biggish nuclear or coal power station makes. But coal sucks balls so we're talking nuclear today.

Now, to deliver dispatchable ("on demand, anytime") energy like nuclear, wind needs to exceed demand &/or use storage. A good test is to see if it can follow the shape of the demanded amount. To contribute its current ~5% dispatchably, it'd look something like this blue wiggle.

One way to do it is make more wind turbines, so the minimum generated is always more than demand. If we build more turbines, capacity moves approximately proportionally, so we essentially re-scale the Y-axis. Here's 1/2 capacity in red and 2x capacity in green to illustrate.

To do what the nuclear could via this over-production approach, wind would need quite a bit of scaling. Here the troughs are at 0.4 GW. To make, say, 10 GW means building 10/.4 = 25 times our current capacity. So ~2350 wind farms (!). Or four medium nuclear facilities.

So, maybe storage is a better solution. Here's a great way to picture how storage works in a grid ...

Storage is about energy capacity mostly, though most pop science pages report power capacity (the easiest bit!) We will look at the real challenge. Let's for sake of it imagine making Aus's WHOLE current wind capacity generate 1.5 GW of dispatchable energy (1 small nuclear plant)

To do it we store the red and release it as green. I will be generous and assume 100% efficient storage that never needs replacement. For reference, batteries could be 90% efficient and last a few decades. Grid hydrogen is like under 50% from memory. Pumped hydro is like 80ish.

These graphs may look a bit silly with me scribbling on them, so here's a real one showing SA's new battery storage working the same way. I also include a big picture SA power graph to make one point: even a huge battery stores basically nothing compared to what a grid uses.

To represent capacity we can just use an area on the graph, and ask how much we need. Well, let's assume the storage source "charges" up as fast as it possibly can. Then we just need to know the maximum it might need to produce before it needs to charge more, like the yellow bit.

What would that take? Let me again take a liberty you can fact check, and suggest molten silicon might be the best currently way to do grid scale storage. Let's say we get 100 (!) of these facilities and they deliver as advertised. Remember this is to do what 1 nuclear plant can.

So, 100x1GWh is 100GWh. Which is this much, pink. THat looks like it could just-about-but-not-quite fill that gap. We'll ignore issues like the form of the energy for now (in reality, I think some of it isn't electricity but heat).

Of course, there will be bigger gaps. to compare it to nuclear, you need to have enough that is fills the worst gaps that come up in the life of a nuclear grid. So, the types that happen once every few decades.

Now, there is optimisation to be done here, there is an optimal amount of expansion & storage. BUT to beat nuclear, the minimum under these generous assumptions is that you will be probably doubling the wind fleet at least once, and building 100 or more of these facilities.

So, that's a back-of-envelope sketch of why, even without getting into the nitpick zone of specific calculations, delivering what even a single nuclear plant can using wind alone is not going to even cost in the same ballpark. The arg extends to any combo of intermittent sources.

And this really is the challenge for the 100% renewable case. Wind/solar can SEEM to deliver exactly what fossils/nuclear/hydro do, but only when they have these things as backup. Once they get near 100%, the training wheels are off, and what I've described is the case!

Hope you've enjoyed the thread, if you did, please share :) These things are a passion of mine that I love learning about, analysing, and sharing, but they do take a while to put together so your support is appreciated :) @jlhulford1 @nuclearkatie @NuclearBootcamp etc.