Fusion, and other fusion-level bets: Part II

The three categories qualify on the same test as nuclear: each could in principle supply cheap, abundant, zero-carbon primary energy at a global scale, though he expects none of them to be as flexible as nuclear is. Geologic hydrogen earns its place for an unusual reason, which is not that hydrogen is a good fuel. The post reprints, at length, Lubershane’s own verdict from about two years earlier, that hydrogen is a swiss army knife, versatile and unwieldy, with poor energy density per liter and a talent for embrittling steel and escaping through gaps. That quoted passage is his earlier writing, and the present prose keeps its judgment while carving out what survives: hydrogen the molecule is irreplaceable as feedstock for ammonia, methanol and refining, and is probably the leading route to decarbonizing iron and steel. The problem is that no low-carbon way of making it, green or blue or any other color, is cheaper anywhere in the world than splitting methane and venting the carbon, and he does not see one coming within striking distance. The exception would be hydrogen you pump rather than make. That is the entire bet, and since published estimates of the resource span four orders of magnitude, the only way to narrow them is to drill. His path is one word: dig.

For renewables the prize is a specific benchmark rather than cheapness in general. Solar and wind are already attractive where sun or wind, open land and transmission coincide, but even in the best American locations they are not quite competitive with the marginal cost of burning fuel without federal tax credits, and their cost in the United States has been rising for five years. The cheapest solar in the world is Saudi, built on conditions almost nowhere else has, so reaching Saudi prices in Arizona or Portugal would take at least a 50% cut, and he adopts that as the benchmark. The bigger prize past displacing power plants is competing with the thermal energy embodied in coal and gas, because industry is more than a third of global energy use and most of it wants heat rather than electricity. Then the reversal. The lever everyone associates with cheap solar, module manufacturing, is spent: prices have kept falling only because Chinese manufacturers sell at a loss, and he argues module prices should rise, for economic sustainability and geopolitical comity both. Panels are now under a third of a utility-scale project anyway. So the cut has to come from panel efficiency, which physics caps at about a third for the single-junction silicon we know how to make cheaply, hence the interest in layering perovskite films on top; and from land, labor and materials, now nearly half of project cost, hence construction robots. He is explicit that neither gets there alone.

He also explains why wind is not in the category, and the reasoning is empirical rather than principled: he has simply not encountered an idea with a realistic hope of halving the cost of wind energy. The three-bladed turbine is mature, the industry’s main cost strategy of building bigger machines is running into the fact that the largest blades no longer fit on roads, and the main alternative paradigm, generating power from devices flown at high altitude, has a twenty-year record whose most prominent entrant spent thirteen years iterating before shutting down. A few small companies are still trying and he puts the odds of a breakthrough low.

Geothermal is the one he treats as fusion-level on its own terms, and better than fusion: limitless, power-dense, available around the clock, with no fuel supply chain and no radioactivity. It may also be harder. Three things have to coincide, high-grade heat, permeable rock and a working fluid, and the binding one is heat, which means depth. That we are already extremely good at drilling is what makes the challenge daunting rather than encouraging, because cost climbs exponentially with depth and the numbers are known. Then the turn that matters most in the piece: he does not expect the fusion-level version to be what happens. He expects Enhanced Geothermal Systems, a category he says is effectively one company, to work first. If its full-scale Utah project succeeds, oil and gas capital and skills follow, costs come down a curve, and better drilling and more precise fracking gradually widen the map eastward. That path is not geothermal anywhere and is not a fusion-level bet, but he can imagine it yielding hundreds of times more geothermal energy than today several decades out.