Australia's nuclear energy debate

Last week I watched a nuclear energy debate between Simon Holmes à Court, the son of Australia's first billionaire, and the Centre for Independent Studies' (CIS) Aidan Morrison, which – while cordial – quickly devolved into a match to see who could best talk past the other person (for reference, that would be Holmes à Court).

Basically, Morrison spent his time pointing out flaws and missing details in the Albanese government's model (the Integrated System Plan or ISP), while Holmes à Court said none of that matters unless Morrison can produce an alternative model with nuclear power in it ("you haven't done the work to show people that there's a better plan").

Fair enough, I suppose; but then what was the point of the whole thing? During the debate, Holmes à Court certainly wasn't willing to put up the $500k he suggested such a model would cost to develop, despite his family's net worth being something like 1,000 times that much. You might think that someone who only wanted the best for the nation would be open to funding that kind of work. But I digress.

By far the most cringe moment came at the 42 minute mark, when Holmes à Court said that discussing the full system costs, rather than only the levelised cost of electricity (LCOE), was "an absolute red herring that you've [Morrison] introduced into the debate to confuse people on this, but is a key source of your confusion on the ISP, and please, let's sit down and have someone explain it to you in detail".

I don't watch many debates, but I suspect that kind of condescension is not a good way to win over neutrals; it certainly made me question Holmes à Court's sincerity. Even more so after I asked a large language model about it:

"Full system costs are a metric used to compare the cost-effectiveness of different technologies. They account for costs that occur beyond the power plant, such as the total costs of supplying electricity at a given load and security level.

The LCOE is used to compare different methods of electricity generation and for investment planning. It's the average revenue per unit of electricity generated that's required to cover the costs of building and operating a generating plant.

The LCOE doesn't consider all costs associated with a financial decision, and it can oversimplify the project context. The LCOE of different power plants can vary depending on the region or regulatory environment."

That... seems like an important distinction worthy of discussion, no? For example, a recent report by Frontier Economics added in a bunch of other costs – such as transforming the grid in WA and the NT, rooftop solar and home batteries, transmission costs – and found that the government's transition was more likely to cost $642 billion, not the oft-cited $122 billion (which, it turns out, is actually a net present value estimate, not what consumers will actually have to pay).

I also noted other contradictions in Holmes à Court's argument. For example, he dismissed nuclear because of Australia's relatively high construction costs, saying that we would never be able to build it as cheaply as China. But only a few minutes later those constraints miraculously vanished; if Australia goes all-in on renewables, says Holmes à Court, then because hydrogen can't be shipped and that's all we can do with the surplus power we'll need to keep the lights on (more on that in a bit), then we'll see an onshoring of minerals processing to Australia because we'll be the cheapest place in the world in which to transform things like iron ore and spodumene into green steel and batteries.

Except... we won't? Steel making certainly uses a lot of energy, but it's still only ~20% of the total production costs in most countries, and is even lower in China.

There are also geographical constraints. For instance, most of our renewable power generation will presumably be near where people live – i.e., primarily in the South East of the country – not where our iron ore and spodumene lithium are located, which would be the North West.