Will New Nuclear Technology Replace Today’s PWRs?

Fast breeder reactors use highly enriched fuel. They achieve extreme burn rates, using something like 100 times less uranium than a simple PWR. But these theoretical fuel savings don’t come close to offsetting the extra upfront cost of such plants. The BN-800, currently operating in Russia, cost just about 2.8 billion dollars to build adjusted for inflation 4. It produces 789MW of net electricity, ending up at a cost of $3550 per kW. And considering the costs of processing and production for MOX fuel, I doubt that you’d end up with major fuel savings to begin with.

Rather than better fuel, reactors using heavy water are able to use cheaper fuel. But there is one downside: the cost of the heavy water itself. How little can we get away with using? A CANDU reactor, the most-common model, needs a kilogram of heavy water for each kilowatt of capacity 5. That adds 900$ in costs per kW immediately 6. If heavy water could cut fuel costs by 40% or more, then it’d surely be worth the extra investment. In practice, it seems like CANDU reactors just don’t manage to achieve the level of fuel savings necessary for them to prove superior economically.

Talking about cutting out the moderator completely, in a fast reactor that’s using better fuel, or having heavy water as a moderator in order to use incredibly cheap raw uranium, is only one half of the equation. A reactor needs a coolant to actually turn the energy from the core into electricity in the generator.

But it’s incredibly difficult to find a better coolant than water. Water is close to free, and it has an extreme thermal capacity. These factors make it one of the best possible materials to use as a coolant for a thermal power plant. A different, better coolant would have 2 advantages: lower pressures, and higher temperatures.

The higher temperatures would really only have the advantage of letting us save on fuel. After all, the higher the temperature difference in our thermodynamic cycle, the higher the efficiency. And the APR-1400, the PWR that we’re wanting to replace, has an efficiency of 36.5% 7, give or take. A magical coolant, that somehow let it turn 100% of its thermal energy into electric energy, would of course cut fuel costs by just under 65%.

Here, the low running costs of nuclear energy bite back. We’re neither paying more for more refined fuel, nor investing extra in the ability to use unrefined uranium. We’re still stuck on the same reactor core to make energy for the rest of the power plant. Our spending on fuel remains the same as it is for the APR-1400, an amazingly low number of just $0.011 per kWh in today’s dollars. That’s roughly $97 for a full year, as much as the interest on investing an extra $2425. A magical, 100% efficient coolant would cut that number two-thirds or so to $855, a savings of $1570. I doubt that we’d be able to make our reactor that much more efficient, that cheaply. But let’s go through some options to check.

Here’s an easy one to check. Supercritical steam reactors are used for coal plants. They’re amazingly efficient, the current world record being 49% 8 efficient, making the savings worth it for coal. After all, coal is a significantly more expensive fuel; EIA statistics find that the fuel cost of coal energy per kWh is consistently 4 times higher than fuel costs for nuclear energy 9.

Assuming we hit that magical number of 49%, the savings from supercritical steam would be around 35% of fuel costs. That’s worth adding extra 600$ or so to the price of a normal PWR. It doesn’t seem like we’d be able to build them so cheap. I mean, the cost difference between traditional and supercritical steam coal plants is in the area of 20-30% 10. I don’t think we’d be able to do the upgrade cheap enough for it to be worth it when the nuclear fuel is so cheap anyway, even assuming that a brand new nuclear reactor matches the world record for efficiency on its first try.

This one is tougher to evaluate. What kind of efficiency improvement would we achieve? Well, one GE turbine reached an efficiency of 50% 11.

CO2 piping would be more expensive, more complicated, and more dangerous than using water as a coolant instead. These factors would quite likely add significantly to the cost of the coolant loop. I’ve already mentioned how using supercritical steam in coal plants is just barely worth it, thanks to the cost savings from their superior efficiency. The fact that the coal industry has yet to upgrade to supercritical CO2 is telling. They should know better than anyone just what a supercritical CO2 turbine is capable of, and still don’t build them. They’d make far less sense with the far lower fuel costs of a nuclear reactor.

Sodium is probably the best metal of choice. I don’t really think there’s a metal or salt more suited for a reactor, so if any liquid-metal reactors would replace today’s PWRs, it would probably be a liquid sodium reactor. The corrosivenes of sodium is a big issue, however. Sodium reactors have faced serious engineering challenges, and concomitant increases in costs, as a result.

And considering the magnitude of the energy efficiency gains, 40% thermal efficiency seems to be a common number, I’m not sure it can justify the complicated and expensive solutions needed for the corrosion issues. If using a sodium-cooled reactor added a small amount to the cost of a plant, say 10% or less, it’d probably make sense. But I find it hard to believe that the corrosion issues and other complexities of a sodium-cooled reactor could let it be built so cheap.

I want end all this negativity on a more positive note. My quick run-through of alternate reactor technologies seems to be all doom and gloom. So I want to talk about how a reactor might actually be able to replace the leading PWRs that we use today. I’ll give you a brief table, showing situations where these promising new reactor technologies would be superior economic options than a PWR.

The bottom line is that the fixed costs for nuclear energy are disproportionately huge, so that any attempt to use a newer, more efficient technology, cannot possibly justify itself through fuel savings. None of the most promising new nuclear technologies promise lower capital costs, at the cost of efficiency staying the same. Yet the fuel cost makes up such a small percentage of the total costs that even huge reductions in fuel costs can’t justify even small increases in fixed costs. And as the discount rates go up, trying to save money through more expensive, if efficient plants, simply becomes ever more dubious an idea.

I think that the crudest, most simple, and most affordable reactor to build would be a BWR. It uses super-cheap purified water as its coolant. It only has 1 loop. It avoids paying the price of heavy water in its core, or the vagaries and complexities highly-enriched uranium. I don’t think we’re getting any bleeding-edge technology that manages to be cheaper and simpler for pure capacity than that.

In the end, I don’t see how an advanced, bleeding-edge nuclear reactor would be able to cost less to build than the currently dominant PWRs. And I don’t see how the more complicated alternatives will be worth their extra costs, either.

References: