04 August 2006

Would Fusion Power Be Too Cheap To Meter?

One of the pitches behind the invention of power generation from nuclear fission power plants was that electricity would be "too cheap to meter." Now, the same pitch is also being used in support of nuclear fusion. A typical expression of this view is found in a Daily Kos diary by sabianq:

The first country and government to bring on-line a Fusion reactor will become the most powerful and richest country on earth.

In reality, we should be selling energy to the world[,] and not starting wars over a non renewable scarce resource that pollutes the air and causes environmental damage.

Nuclear fusion would be a wonderful development that would make life on Earth more sustainable and could dramatically reduce pollution. But, this doesn't mean that it would make power too cheap to meter, or generate economic growth. It could, and initially probably would be, more expensive than current methods of generating electricity.

Where Are We Now?

Nobody has come up with a commercially viable nuclear fusion plant design. It is decades in the future and has been since the 1950s. The current lead fusion research project ITER, which will come on line about 2016, and it would take another decade to commercialize, even if ITER finally discovers a commercially viable way to generate power through nuclear fusion.

Nobody has even come up with sustained positive power generation from a nuclear fusion plant yet.

We Don't Know If It Can Be Done.

There are all sorts of things that are theoretically possible, which are engineering impossibilities.

It is not even well established that it is possible to do sustained nuclear fusion on a scale useful for a power plant. The Earth's core stays warm with nuclear fission, not fusion. The smallest natural sustained nuclear fusion involves masses a little greater than Jupiter. The smallest scale net energy generating fusion reaction device developed by people is the H-bomb, and trying to get commercially useful power out of H-bomb size reaction is like trying to drink from a fire hose. Lab scale fusion reactions have been momentary and taken more energy in than they put out (with perhaps a handful of cases which generated net power production for a mere moment).

There is no theoretical bar to sustained commercially useful fusion, but it is possible that there are subtle relationships that make this impossible from a practical perspective.

The Accountants Strike Back

Also, it doesn't necessarily follow that a nuclear fusion power source would be too cheap to meter. Sure, fuel would be a minor part of the cost. But, heavy water is still more expensive than fuels like sunlight, wind, waves and water flowing downhill. And, hydropower, wind generated power, and solar power are not free. Indeed, wind and solar power are just starting to become economically competitive with coal fired power generation after several decades of sustained development and many billions of dollars of R&D money.

If it costs several billion dollars to build a 1000 megawatt (i.e. typical sized) power plant, powered by nuclear fusion, it wouldn't necessarily be cheaper than the alternatives. It is almost certain that a first nuclear fusion power plant, including R&D costs, would cost that much. And, that price is hardly out of the question on a per unit basis if it is a machinery intensive tokamak design. Nuclear fission and coal fired plants use a far simpler, and hence cheaper, design. Both are glorified steam boilers. Also, coal prices are to some extent a product of the cost of the alternatives. If nuclear fusion power does turn out to be cheap, coal prices may drop making that fossil fuel economically competitive again.

Now, I'd be willing to pay quite a bit for a clean, sustainable power source like nuclear fusion, but, that doesn't necessarily mean it would be cheap. And, if it isn't cheap, then there might not be any excess energy to sell, and it might not produce massive economic growth.


Data from a recent British study indicates that a coal fired plant of 1000 megawatts costs about $1.6 billion to build (with a 30 year life expectancy), and $46 million a year to operate.

A comparable sized natural gas fired power plant costs about $630 million to build (with a 20 year life expectancy), and $65 million a year to operate.

A comparable sized nuclear fission power plant costs about $2.2 billion to build, including an allowance for decommissioning (with a 40 year life expectancy), and $78 million a year to operate.

A comparable sized onshore wind far costs about $1.4 billion to build (with a 20 year life expectency), and $46 million a year to operate.

There are all sorts of caveats to those numbers. You have to think about backup power generation for intermittent sources, value the differences in the amount of pollution produced, regularize depreciation related costs, and so on. You are welcome to read the linked study which reaches some overall conclusions after considering factors like those. And, there are countless assumptions that go into the bottom line numbers above, which is why I haven't continued the analysis here. For example, regulatory costs or emissions controls can greatly impact the cost of building a power plant. But, the study provides useful order of magnitude estimates.

But, this data is sufficiently clear to show that even with free fuel and minimal operating costs (such as staffing payrolls), a 1000 megawatt nuclear fusion power plant that costs $5 billion to build, for example, is little more than a novelty outside niche applications like space travel, with little or no commercial application in the absence of government subsidies. And, while it might make sense to subsidize nuclear fusion to some extent, at some point, going the way of wind and solar power, which also have minimal environmental impacts, makes more sense.

Thus, it isn't just a matter of inventing a sustainable nuclear fusion power plant. Even if somebody does this (and this is the hope of the ITER researchers), it might not change the world much. One must also develop a plant that can be built within some pretty severe budget contraints. And, I have yet to see anyone with even pie in the sky, 30 years from now proposals for sustained nuclear fusion power generation, that would be cheaper to build, or have lower non-fuel related operating costs, than existing nuclear fusion power plants.


Paul D. said...

No nuclear industry figure ever said fission would be 'too cheap to meter'. They knew full well that fission would not be much cheaper than coal-based power generation. That phrase came from the head of the AEC in testimony to congress when asked about what we might hope to see in the distant future from the overall advance of technology. He didn't even say it was fission that would yield that possible outcome -- it could have been fusion he was talking about!

In reality, fusion has little chance of even being competitive with fission. The first wall of a DT fusion reactor is projected (Science, March 2006) to be as expensive, per kW(e) of capacity, as an entire fission powerplant. With uranium comparatively cheap and likely to remain that way, trading a more expensive plant for a slightly cheaper fuel (deuterium + lithium) is moving in the wrong direction.

Anonymous said...

Fusion would create no radioactive waste. It wouldn't require rare or difficult to obtain fuel (uranium, coal, sun or wind which are only available in certain areas). Even if the cost to run a hypothetical fusion power plant creates electricity at the same cost of current coal or fission plants it would be a good alternative. Once world coal and uranium reserves are exausted those types of plants will not be making electricity at any price. Demand for more electricty could be met by building more fusion power plants, unlike solar or wind or water which have natural limits to expansion.

Still, fusion power will require a huge expenditure to develope if it even is possible, which it very well may not be.