On my previous post, I received an excellent comment from Alex Brown, that I would like to try to answer in a full post. Alex said:

While the enthusiasm is a good thing there are several reason why this is more or less impossible for TVA to achieve. First off is the fact that TVA is a power producer, they do not design nuclear reactors nor are they active in getting new designs approved. TVA really only has the options to choose between the already approved designs, and since so far none of those is a thorium powered fluoride reactor TVA cannot build them. Furthermore, if TVA were to attempt to push this technology they would be amazingly lucky to even be able to liscensce and construct a single plant before 2020. Even if they were able to start building in 2010, they would need 14 1000MW reactors to replace their coal fleet, the manpower does not exist to build that many new reactors simultaneously, the manufacturing plants do not exist to produce so many heavy forgings on such notice, and more importantly, the financing does not exist to pay for them. At 2 Billion a reactor (a VERY conservative estimate for a new technology) they would still be incurring 28 Billion dollars over 12 years, or about 2 Billion a year. Since TVA currently has a gross revenue of 8 Billion a year they would have to spend 25% of gross revenues on construction just to avoid breaking the 30 Billion dollar debt ceiling imposed by congress. Also, this does not even bring into account replacements for the 8000 or so MW of gas turbines that also burn fossil fuel, only the coal plants. OF all the utilities in the country right now I cannot think of one that is doing more for nuclear energy. With Browns Ferry 1, Watts Bar 2 and the 2 reactors at Bellefonte TVA has chosen to go 100% nuclear for new base load generation, there really isn’t much more they can do. As it stands now mine-mouth coal plants in Kentucky are already trying to cherry pick TVA customers, if they are forced to raise rates to replace their own coal units they will no longer be able to stay competitive.

Alex, first of all, thanks for taking the time to write this…

How could TVA replace their coal plants? First of all, let’s look at the magnitude of the problem. TVA operates 11 coal plants across Tennessee, Alabama, and Kentucky. These coal plants produce about 60% of the total electricity generated by TVA. Each is located on the Tennessee River or on a navigable offshoot of the Tennessee, and thus has river access and abundant coolant water available.

To replace these coal plants will require TVA to come up with about 15 gigawatts of electrical generation capability, nearly as much as the Three Gorges Dam in China. These coal plants are:

• Allen (753 MW)
• Bull Run (870 MW)
• Colbert (1198 MW)
• Cumberland (2530 MW)
• Gallatin (988 MW)
• John Sevier (712 MW)
• Johnsonville (1254 MW)
• Kingston (1456 MW)
• Paradise (2273 MW)
• Shawnee (1369 MW)
• Widows Creek (1629 MW)

Assuming that a liquid-fluoride thorium reactor (LFTR) was built with a capacity of 1 GWe, we would need 15 of them to replace these coal plants. Now given TVA’s history building Browns Ferry, Sequoyah, or Watts Bar, is that feasible?

Certainly not…if we were assuming that the fluoride reactor would encounter similar development issues to the light-water reactor. Building LWRs in the United States has been a business fraught with developmental risk. The first one comes when you announce a site, since it immediately becomes a focal point for anti-nuclear opposition, and at best, distrust from the local population. Evacuation zones have to be established with the NRC, detailed seismic surveys must be conducted, and a detailed environmental impact must be made of the site. If you find the rare red-speckled, golden-bellied, spotted toad things could be all over.

Then there is the question of actually building the facility. You’ve got to pour an immense amount of concrete, hundreds of tons of rebar, and then there’s the reactor vessel and turbine hall themselves and their incident equipment. Specialized casting of the primary vessel–and such a capability is no longer available in the US–also puts you in a long line. You’ve got to nail down your fuel supply, which at today’s elevated uranium prices could land you in a world of hurt if you happened to lock in at the peak of a speculative market. Trouble, trouble, trouble in every direction.

So why consider something as radical as thorium and the fluoride reactor? You correctly observe that it’s not something you’re going to go buy from some vendor. You’ll have to develop it, and if you’re TVA, why should you?

Simple answer–because once you do and replace your coal plants, everyone will be coming to you wanting the same thing, and you’ll make a lot of money.

Now let’s be honest here. No one, with the possible exception of PBMR, is developing new reactors today. The AP-1000. ABWR, and EPR are simply the latest repackaging of 50-year old light-water reactor technology. I don’t have anything against 50-year-old technology, if it does what it needs to. Fluoride reactor technology is 50-year-old technology. But today’s newest incarnation of light-water reactors use very similar enrichments, the same sort of power conversion, similar burnup levels, and the same basic non-approach to fuel management.

One might say, well, it works, so why change it? Well, I would contend that it doesn’t work. Today’s LWRs make money because they are paid off. The process of getting them built and paid off bankrupted several utilities, and is not a risk that Wall Street is anxious to repeat. But now the issues of CO2 emission and the risk of a carbon tax is making them wonder which is the greater risk–continue with coal or build nuclear (LWR)?

So a fluoride reactor would certainly be a new development. And for TVA to implement this ambitious goal they would probably have to be the developing agent. But why consider it? Because the fluoride reactor has the promise of vast reductions in capital costs and fuel cycle costs. Fluoride reactors can be built compact and inherently safe. Not probabilistically safe. And that’s a big difference.

You don’t need the heavy specialized castings of a light-water reactor in a fluoride reactor. It operates at ambient pressure. You don’t need the complex triply-redundant emergency core cooling systems and backup power and core pressurization systems. The salt melts a drain valve and safes itself in any loss-of-coolant situation. Inherently safe, not probabilistically safe.

Qualifying fuel–today’s LWRs have qualified two fuel enrichments over the last 40 years, 2.3% and 3.1%, if memory serves. Changing fuel forms or enrichments is prohibitively expensive. For a fluoride reactor, qualifying fuel is like falling off a log. Stick a sample in a high-flux reactor like HFIR, barbecue the heck out of it, and give a lifetime dose in a few months. Many would argue that the fuel is already qualified.

The long poles in the tent for the reactor are going to be the power conversion system and the fuel reprocessing system, in my opinion. The gas turbines of the PCS use helium and a cycle totally different than today’s open-cycle combustion turbines. There will be a fair bit of work there, but we have the CFD and turbine design codes today to really shave development time off that task. Fuel reprocessing will requ
ire four steps: protactinium isolation (using bismuth extraction), fluorination, reconstitution through hydrofluorination, and distillation. Each of these four steps can be developed in parallel to the reactor development. Four steps to a closed-fuel cycle. Imagine that and compare it to any other reactor out there. Four steps from a thorium feed to a fission product waste stream. That’s incredibly impressive.

Finally, how to build these units….do we build them on a site? No–this will be one of the most radical breaks of all, and possibly one of the most lucrative. We build the reactors in a factory in a submersible configuration. LWRs can’t really be built in a factory because 10 meter pressure vessels nine-inches thick aren’t very mobile. But fluoride reactors with a high-core-power-density operating at low pressure could be built in that fashion. And just as Henry Ford showed when he started building cars on an assembly line instead of one at a time, you can realize incredible economies-of-scale and production rates when you do it that way. We don’t do it that way with LWRs because we can’t. With fluoride reactors we can and we should.

There’s already an excellent site to do this work: the Delta 4 manufacturing plant in Decatur, Alabama, right on the Tennessee River. I’ve visited this plant on several occasions, and it was built for the production of about 60 Delta-4 rockets per year. Right now it’s only doing a few per year and is incredibly underutilized. Here’s a place with milling, bending, welding, forming, and shaping equipment, along with rail and river access.

Let’s say that it took until 2015 to get reactor production started. A few years to convince TVA, 5 years of design and development, and production beginning in 2015. At a rate of one reactor built every four months in the plant, by 2020 you’ve produced 15 reactors and can deploy them. They float in the river up to where the coal plant is today and plug into the grid, assuming the electrical demands currently met by the coal plant. A year or two later and you could replace all the natural gas plants too.

The key to the plan would be to build a mobile, power-rich energy source. Fluoride reactor technology enables the compact reactor. The thorium fuel cycle enables the reactor to avoid having to “return-to-base” every 18 months for refueling like an LWR, and it also lets the reactor avoid further speculation in the uranium market, giving the TVA customer long-term rate stability.

If TVA were to develop this technology every utility in the country would be screaming for their own thorium reactors. Because by 2020 they’ll be paying a lot of money to burn coal, gas, and possibly even uranium.