Why Laser Fusion Power Is So Difficult
Is the National Ignition Facility's goal of generating practical electrical power realistic? How long would it take? Having learned the history lesson on inertial confinement fusion and its many acronyms, we know how we got here. Now we can cover the fun stuff: will we ever get it to work as a power plant?
Two distinct types of challenges face ICF "laser fusion" as a commercial power source. Engineering problems tackle transforming established scientific principles into working machines. First though, science has to know what will work. That is the question of what is possible and, if so, how to attempt it.
When we imagine an Apollo project or a Manhattan project for nuclear fusion, we frame the problem somewhat incorrectly. The basic science of using liquid propellant rocket engines and multi-stage rocket platforms had already been figured out well before Apollo began. American Robert Goddard, in parallel with German and Russian rocket scientists, had worked it out decades before. At the end of WWII the US claimed V2 rockets and their designers from Germany; we already had a working small-scale liquid fueled rocket capable of reaching the edge of space. Improving and supersizing rockets enough to blast people into space was primarily by this point an (admittedly enormous) engineering problem. A huge infusion of money and brilliant mindpower accomplished it in a matter of a decade.
So too with the Manhattan project. We believed our scientific concept that a sufficient mass of fissile material could attain a runaway chain reaction. Einstein had already supplied the matter-to-energy conversion factor. Brilliant physicist Enrico Fermi built the world's first nuclear reactor under the bleachers at the University of Chicago football field. This showed the energy output from a neutron chain reaction in fissile material. Manhattan was all about finding an engineering method to get enough uranium and plutonium to build a handful of bombs and to design a bomb mechanism capable of assembling the necessary critical mass instantly and perfectly. This was a tremendously hard issue on its own, even with most of the science previously resolved.
NIF is in a more difficult position. We don't know the science well enough to reach the engineering phase.
Scientific unknowns abound in laser fusion. First, our models of the appropriate conditions needed for useful extraction of fusion energy are not tested by experiment. Currently, we think that we need about 10 megajoules (MJ) of laser energy (i.e., about 138,000 professional tennis serves or a pickup truck hitting a wall at 223 mph) to get net power out of NIF. The facility's record-breaking laser produces only about 1.8 MJ.
Furthermore, 10 MJ from the pellet is just the energy break-even point. To see practically useful gain instead of eeking out almost no power, we think we need a 10-fold increase in power: a 100 MJ laser. Then figure that most heat is lost when being converted to power. So perhaps we need another factor of 10: a 1000 MJ (1 gigajoule) laser. Most historical predictions of the energy required for energy gain with ICF have been vast underestimates too.
The trouble is that designing entirely new lasers is more of a science question. Entirely new laser gain materials such as specialized doped glass that costs thousands of dollars per inch need to be invented. New concepts for generating bursts of photons and controlling their conversion from one frequency to another must be found. The field of optics needs time to develop and understand ideas of how to make more energetic laser pulses. Perhaps better fuel design could lessen the need for laser advances.
The moment of the pellet implosion is another deep science question. At laser impact, a chaotic ball of x-ray photons, nuclear ions, gamma rays and energy is created, which is extremely hard to analyze. An entire field, hydrodynamics, studies these messy situations. We've been working on hydronamics for centuries, but the models we use don't do a good job of predicting what happens in moments of pressure, temperature, and chaos so high as these. The poor projections of these models are part of why we are so far behind our goals with ICF so far.
Say we invent a laser 1000 times more powerful, design new pellets, and new methods to perfectly crush them. We're still not done. Engineering challenges will be an enormous obstacle. Essentially, we'll need a Manhattan Project for fusion.
Currently, NIF fires at a maximum rate of roughly once per hour. For commercial energy production, blasting a pellet roughly ten times per second is required. This means we'll need to figure out a way to run the entire experiment 36,000 times more often and also 36,000 times more quickly. Engineering and science are at play here. Powering up the laser to fire this quickly is a problem for both scientists and engineers: it needs to charge up with power 600 times faster than the current model.
More traditional engineering challenges are also manifold. A method is needed to collect the heat from the target chamber with reasonable efficiency. The fusion energy is currently not collected at all. Then we need a system capable of feeding in a pellet, triggerering the laser, entering the chamber, removing the remnants of the blasted target, replacing it with a fresh one, and resealing the chamber. This needs to happen in 100 ms or less, while also not leaking significant heat out of the chamber.
The fuel pellet too will take some engineering. Roughly 100,000 pellets will have to be blasted every day for every plant. This means we'll need to produce fuel pellets by the millions and at low enough cost not to ruin the economics of the machine. Large and stable sources of deuterium and tritium will be need to be founded.
These are just the foreseen challenges. Entirely unforeseen complications will almost certainly arise. While it is foolish to say that we will never see consistent electrical production from this method, it will be an enormous struggle. The fight is utterly impossible to win within a decade. A more realistic expectation is probably 100 years for a commercially viable powerplant.
That's a shocking number. However, it says more about the difficulty of the project than the quality of the scientists involved and their work. For fusion powr to be successful, we must plan for the long, long haul.
The author would like to gratefully acknowledge discussion with former NIF director Mike Campbell for insight into NIF and ICF energy projects.