The Problem with Fusion

Although there are many technical challenges that still remain in order to make fusion energy a viable source of electrical power generation, the main reason humans will never harness fusion energy is one of economics. By the very nature of a fusion reaction, pressures and temperatures only experienced within stars, fusion reactors are necessarily the most complex and technologically sophisticated machines built by man (the Large Hadron Collider (LHC) at CERN is another candidate). As such, fusion reactors are also among the most expensive. No amount of technologic breakthrough will make fusion generation a cheap enterprise. In order to recoup that expense, massive economies of scale have to be achieved in order to make fusion energy a viable source of electrical power generation. But massive economies of scale can only be achieved with massive installations which bring about a whole host of other problems like local environmental impact, vulnerability to local weather and geologic phenomenon (earthquakes, tsunami, etc.) not to mention human caused events such as accidents, sabotage and military action. Finally, massive installations have huge price tags.

First let’s look at the order of magnitude kind of cost we can expect. Nuclear fission technology is about 80 years old and by all standards a fairly mature technology, even considering the latest reactor designs like advanced boiling water reactors (ABWR). The largest nuclear power plant in the world is Japan’s Kashiwazaki–Kariwa NPP which as seven reactors for a total generating capacity of 8 GW (8,000 MW). From ground breaking to first power generation took more than four years. The last reactor did not come online until 12 years after the first one requiring huge capital outlays before revenue could be collected by selling the generated power. The average cost per kW of electricity generated is about $5,000 or a total of $40 billion to build the entire complex.

Here is a list for cost comparison to other relevantly complex machines:

• Trident nuclear missile submarine – $3 billion
• Nimitz class nuclear powered aircraft carrier – $6.2 billion
• Lifetime cost of Space Shuttle program – $200 billion ($1.5 billion / flight)

The International Thermonuclear Experimental Reactor (ITER), the only commercial fusion reactor program currently funded has an estimated price of $18 billion. (The US National Ignition Facility, the only fusion project in the US, is changing priorities after failing to meet the goal of "ignition" when the latest round of funding ended this last September.) It’s designed net power generation is 450 MW. This is only 5% of the power generated by Kashiwazaki–Kariwa NPP for almost half the cost. So even assuming ITER is a complete success, which is doubtful considering the immature state of the technology, it will not be economically viable. But it is a demonstration of the technology, so economics are not the primary objective.
However, the outlook for fusion gets worse. The economies of scale required for a large fusion reactor demand generation on the order of 5 – 10 GW. This falls out of a complex analysis of cost for power generation that ranges from $2000 / kW for state-of-the art pulverized coal plants to $10,000 / kW for the latest ABWR nuclear fission power plants. In order to be economically viable, a fusion power plant has to deliver power within this price band. For maximum economies of scale, at $10,000 / kW for a total of 10 GW power generation, this equates to a total lifetime cost (capital, operating, fuel and financing) of less than $100 billion. For a 20 year lifetime, this would approximate to $80 billion to build and $1 billion annual costs. And this would make the electricity one of the most expensive sources. To come down to the $2000 / kW of coal fired plants, the build cost would have to come down to about $16 billion. That’s already less than the projected ITER cost for only 450 MW of generation.

There are few aspects of the technology required to fuse hydrogen atoms that indicate orders of magnitude cost reduction over time. Compare the costs of the largest partical colliders as they have grown in size over the last 20 years. The closest parallel technology on the electric power generation front is the evolution of nuclear fission power plants. Even at the enormous economies of scale afforded by Kashiwazaki–Kariwa NPP its cost per kW is still in the middle of the cost band for electrical generation sources. Most of the worlds nuclear power plants fall in the upper reaches of this price band. So even under the very best of technological circumstances, fusion power will never be a viable source of electrical power generation purely for economic reasons.

Sustainable Means What?

Sustainable, carbon-neutral, self-sufficient and similar terms get thrown around and used somewhat interchangeably in the press and on the internet. For the purposes of this blog, I will use the term sustainable to refer to a process that can be carried out indefinitely. This can only happen when that process consumes resources slower than those resources are replenished (or at the same rate).

A simple example is collecting dead fall from the forest floor to make a camp fire to roast marshmellows. As long as you collect the wood that falls from the trees to make your fire and don't cut any live trees, you can do this for a very long time (generations, centuries, etc.) The trees will continue to grow and produce wood for your camp fires indefintely. But as soon as you collect all the deadwood and decide to start cutting down trees to burn, you have started a non-sustainable process because you are consuming the wood faster than it is being replenished.

All human activity can be examined through the lens of sustainability similar to this method. The complexity arises in the many different inputs required for most of our activities and analyzing whether each one of its resources are being consumed slower than they are being replenished.

There is one extreme example that I want to get out of the way for any critics who find issue with my definition of sustainable processes. Sunlight is for all intents and purposes a sustainable energy source even though we all know that the sun is slowly consuming its vast store of hydrogen and fusing it into helium in a non-sustainable process. I think any resonable person will agree that 4-5 billion years is such a vast amount of time beore this resource runs out that it is effectively infinite.

On the other hand, fossil fuels are not infinite in supply. We are close (10-50 years depending on who you ask) to consuming 50% of all of the easily accessible petroleum on the planet. That is a timeframe that most of us alive today will live to see. After that point (called peak oil) the demand to consume oil each year will exceed the supply, likely resulting in wild price flucuations (remember the price of oil in the summer of 2008?). This is a common response in complex systems to restricted supply of resources.

So why is all this important? If a process is not sustainable, at some point it will stop (because there will be no more resources to keep it going). In our campfire example, no more trees (think Easter Island), no more roasted marshmellows.

For all of human history our resources have been effectively infinite (like the sun's supply of hydrogen). At the end of the 20th century I belive human beings entered a new era where on a global scale we are starting to exhaust the resources upon which we depend for everything from growing our food, to generating our electricity to building our cities. History is littered with the ruins of civilizations that collapsed because of the over-consumption of local resources. For the first time in history we are going to have to contend with the consequences of over-consumption of our resources on a global scale.