Fusion, the power of wishful thinking

I got Sun in a Bottle by Seife the other day, a book on the history of nuclear fusion. Here’s my mini-review… and, of course, a comment.

Sun in a Bottle is a surprising easy read, given the topic. I polished it off over a period of two days on the train and waiting at a doctor’s office. It’s extremely well written, and at a technical level that absolutely anyone can handle.

The book is well paced, covering major developments without getting dragged off into too much historical detail. My only complaint on the history side was the overriding focus on US designs, with major projects in the UK and USSR mentioned only in passing.

I mention this because the first fusion reactors were actually in the UK, predating everyone else by several years. But the text presents the US efforts first, and seems to imply the UK designs came along later. But that’s a minor issue that doesn’t cloud the issue he’s trying to illustrate.

The point is…

The main thread weaving the book together is the wishful thinking that dominates the fusion field. In spite of a continual stream of  failures, researchers continue to believe that the next machine will solve all their problems. This is Einstein’s definition of insanity, doing the same thing over and over and expecting a different outcome.

We have a term for this, “pathological science”. It’s what happens when a researcher allows their desired outcomes to color their measured results. Seife uses this term once when referring to the infamous bubble fusion example from a few years back, but personally I think it’s safe to say the term applies equally well to the entire field.

Seife sums everything up nicely at the end, by giving an example of a scientist who’s groundbreaking research in extrasolar planets turned out to be a mathematical error. His coming out party at a major meeting instead turned into a hastily prepared talk on how they got it wrong. This is how science is supposed to work.

Seife compares and contrasts with cold fusion and bubble fusion examples. When these teams were presented with obvious problems in their data, instead of backing off and looking for more data, they pretended the problems didn’t exist and pressed on with their announcements. What followed was predictable.

I heartily recommend this book.

But then again…

Now for my complaint.

In a way, Seirf falls prey to the same problem that infests the field – the underlying assumption that fusion energy is primarily a technical issue to be solved. He even goes so far to hedge his bets at the end of the book, saying that ITER just might work.

The technical issues may, indeed, be overcome. But as I covered in a previous article, there are far greater problems in the economics side of things. He touches very briefly, on why even a working ITER might not lead to production reactors, but never really takes it anywhere.

So let me try. I’ll use the easy example so you can see how stark the problem really is.

Problem!

One of the major projects in the fusion field today is a device known as NIF (it’s covered in depth in the book). NIF uses a battery of lasers to quickly heat a hollow spherical “target” so that it literally blows itself apart. Remember that every action has an equal and opposite reaction, so while some of the target is blowing itself outwards, some of it is being sent flying inwards.

Painted on the inside of the target is a thin layer of fusion fuel. The explosion of the target drives this material inward until it compresses into a tiny dot. And we’re talking serious compression here, to as much as ten times the density of lead. Not impressive? Recall that fusion fuel is hydrogen, the lightest element. The compression is like squeezing a basketball into the size of a pea.

The idea of all of this is to compress (and thereby heat) the fuel so it gets hot enough to fuse in the very centre of the implosion. If you get everything right, the heat given off by the fusion warms up the fuel outside the center, causing it to fuse, that heats the fuel outside it, and so on. This is called ignition.

Ignition is so important, that it’s part of the name of the device; NIF stands for National Ignition Facility. Without ignition you might as well give up.

Now to create the conditions that we need to get ignition, you need to do some pretty fancy things. For instance, you need a whole lot of laser power. In techno-speak, NIF delivers 1.8 MJ of energy – or in more common measures, 0.5 kWh.

In order for that energy to efficiently heat the fuel, it needs to be  ultraviolet light. The lasers are infrared, the wrong end of the spectrum. So they use extremely expensive crystals to do “upconversion”, turning the IR into UV. Half of the laser energy disappears in the process, so the original lasers aren’t 1.8 MJ, they’re 4 MJ.

Oh, but we’re not done yet. Lasers, especially the ones used in this case, are fantastically inefficient. Only about 1.5% of the energy put into them comes out as light. So driving this process is 422 MJ of electricity from a huge power supply.

Now if everything goes absolutely perfectly, and the target ignites, you’ll get about 13 MJ of energy back out.

422 in, 13 out. See the problem?

And now for the punch line. Seirf’s book was written too early to know this, but NIF has failed to reach ignition. They don’t know why, and they don’t really know how to fix it.

It will probably reach ignition at some point. But it might not. After all, the last two machines they built to try to reach ignition also failed, so history’s against them.

The real problem

Now it’s bad enough that even a working NIF would generate less than 1/10th as much energy as you feed into it, and even worse because they’re currently stuck at around 1/30th.

But as it turns out, there’s an even bigger problem that no one likes to talk about. This one is so fundamental that it’s highly likely that it cannot be overcome. It is almost certainly the case that laser fusion can never, ever, work.

You see, in order to get to the 1/30th that they’re getting now, the laser energy has to be delivered to the target extremely evenly. Sairf covers the problem, called Rayleigh-Taylor instabilities, in the book. In order to smooth out the light, NIF uses a device known as a hohlraum. This is basically a small metal cylinder covered in gold. When the lasers heat up the metal, it re-radiates the energy as x-rays. These shine on the target more evenly than the original laser beams.

Ok it’s time to do a little math. Don’t panic.

MJ and kWh measure the same thing, just different scales, kinda like Fahrenheit and Celsius both measure temperature on different scales. That 13 MJ of energy released in an ignition event is the equivalent to just under 3.6 kWh.

That energy comes out in many forms, which eventually turn into heat. That heat would be used to drive a turbine in exactly the same way as a coal plant. That conversion is, at the absolute best, about 40% efficient, so of that original 3.6 kWh of energy, we’ll end up with 1.44 kWh of electricity.

Right now the wholesale price of electricity in Ontario is about 1.9 cents in the last OEB report. So that 1.44 kWh is worth just under 3 cents, let’s round it up to 5 to be on the safe side.

And what did we need to create that 5 cents worth of power? A hollow beryllium metal sphere suspended inside a gold-plated metal cylinder, all machined to within billionths of an inch. The cost of the materials alone is on the order of a couple of dollars. Production costs are in the hundreds, or thousands (or millions).

So forget all the technical problems, those might indeed be solved someday. But the economic problems makes those look like a joke. You are, quite literally, better off burning money.

The laser approach was first outlined in depth by a John Nuckolls back in the 1960s. Even as he was publishing his first papers on the concept, people at the labs were pointing out the economic problem to him. At the time he dismissed it by imagining a sort of eye-dropper or perfume mister that would create the fuel droplets for basically no cost.

We know now, after 40 years of effort, that such a solution cannot possibly exist. The absolute minimum price for a realistic target is many, many dollars. Even the most wild-eyed estimates of best-case scenarios suggest that the minimum cost targets would be in the range of 50 cents, still at least an order of magnitude away from where it has to be. We have no idea how to cross that gap.

So while the technical performance of the reactors has improved slightly over the years, the economic performance has actually dropped. We are not closing in on a practical energy source, we’re moving ever further away from one.

The analysis for a tokamak like ITER is much more detailed and difficult to follow, so I’ll spare you the details. But let me assure you, it’s just as damming and moving just as rapidly in the wrong direction. Even if we get these things working, they will not ever be practical.

So that’s my complaint with Seirf’s book. He does an admirable job covering the technical side of things, but never seriously considered the economics of it all.

Advertisement