Operators also make nuclear ships expensive to run. They need exhaustive training. Like pilots, their skills are valued by the civilian economy. Many naval reactor operators ended up running power plants ashore. Although the civilian nuclear power industry has been running down for years under the pressure of environmentalists, in the last few years there has been increasing understanding that nuclear power may be the main hope of reducing greenhouse gas emissions in the U.S. It also offers independence from foreign sources of oil – and many believe that money spent on foreign oil often ends up in terrorist hands. It seems likely that the next five years will see a revival of the American nuclear power industry. Any evaluation of naval reactor programs has to take such a possible revival into account. On one level, revival would benefit enormously from training given to naval operators. Such training might be seen as a government investment in civilian nuclear power, with the understanding that naval operators might not stay at sea nearly as long as the Navy might like.

For that matter, increasing investment in naval nuclear plants might fund expansion of nuclear reactor production in the United States and thus reduce the overhead cost associated with an expanded civilian nuclear power industry. In the 1960s, the last time the Navy seriously contemplated building nuclear surface ships (other than carriers), civilian use of nuclear power was expanding rapidly. If anything, the Navy was competing with the civilian sector for reactor production. There was no need to support the producers, but now there may be a very important need.

Kirovs have an unusual mixed plant. Their two reactors suffice for 24 knots, and oil-fired superheaters apparently provide extra steam to achieve their full 31 knots. If this description is correct, it probably reflects the fact that it is difficult to boost the power of a nuclear plant once it has been designed.

The $70 calculation probably concentrated on first cost and fuel cost. It is unlikely to have taken into account personnel cost, which many naval officers remember as the single worst factor in operating nuclear surface combatants. Another question, only rarely raised, is the effect of battle damage. A carrier operates so far offshore that the current littoral threats are probably irrelevant to her. A destroyer or a large amphibious ship might have to come a lot closer. These ships have very little protection, at least as currently designed. It would seem unwise to rule out battle damage. Even a relatively small shaped charge missile might penetrate to a reactor compartment. In that case, what sort of damage control would be possible? Would the standard response have to be scuttling the ship? In that case, working out the cost of a nuclear-powered surface ship would have to take into account her very different survivability. This sort of calculation was irrelevant to the Navy of the 1960s, which envisaged a short, sharp war against the Soviets rather than the sort of protracted, global low-level conflict in which we now find ourselves. In the short, sharp war, a ship out of action for a few weeks would have been as good as lost, because within those weeks either both sides would have come to terms or the war would have escalated into a nuclear holocaust (this vision may not have been terribly realistic, but it dominated defense thinking at the time). Now ships are periodically damaged – like USS Stark in the Gulf or USS Cole at Aden – and it is well worth our while to put them back into service, because we rightly think in terms of decades.

The foreign nuclear ships amount to Russian, British, and French nuclear submarine forces, the French carrier Charles de Gaulle, and the Russian Kirov-class nuclear cruisers. The Royal Navy uses technology similar to ours. In the 1960s, it saw nuclear power as its future, but it had to abandon that idea because nuclear ships seemed unaffordable as its own resources plummeted. Like us, the British have embraced the one-shot reactor as a way of keeping nuclear warships affordable. Their submarine reactor technology was originally developed from ours, but their current natural-circulation reactors are apparently entirely their own.

The French took an approach very different from the United States. U.S. and British reactors are fueled by highly enriched uranium, which gives them long operating lifetimes. The French decided that high rates of enrichment were too expensive, so they chose instead to design ships specifically to use limited-life fuel, which they call caramel (because the pieces of fuel resemble large candies). Reactor lifetime is about eight years, and the current plant in a Le Triomphant- class strategic submarine produces about 41,500 SHP.  When the high-level decision was taken that the next carrier would be nuclear, no money was allocated for a powerful new reactor; it may be that caramel fuel is not compatible with power outputs much above 40,000 SHP. The nuclear reactor project for the Charles de Gaulle replaced an earlier one for a much smaller nuclear-powered helicopter carrier, and possibly some of those involved did not realize how different a full-scale carrier would be. Carrier internal volume was inevitably limited, partly by the (again political) decision to build the ship at Brest naval shipyard, with a building dock of limited dimensions. Apparently there was enough space for two of the big Le Triomphant-type power trains, including the reactors, giving the carrier 83,000 SHP – about a third of what the U.S. Navy uses to power its Nimitz-class carriers, which are something more than twice her displacement. Unsurprisingly, Charles de Gaulle proved slow for a carrier, and the French Navy plans to use conventional power if it builds a second carrier. For the future, nuclear power in the French Navy will probably be confined to submarines. That does not indicate any disenchantment; the French abandoned their non-nuclear submarine force because they found nuclear power far more worthwhile.

The Russians embraced nuclear submarine construction on a larger scale than any other navy. They did not, apparently, accompany that embrace with sufficient investment in means of refueling reactors or, for that matter, decommissioning them. The Soviet system emphasized quantity production over developing infrastructure, and apparently it did not support the kind of longer-life reactor fuel developed in Britain and in the United States. The big reactor in an Akula or Sierra is rated at 50,000 SHP, which is probably somewhat more than in a U.S. Seawolf. Kirovs have an unusual mixed plant. Their two reactors suffice for 24 knots, and oil-fired superheaters apparently provide extra steam to achieve their full 31 knots. If this description is correct, it probably reflects the fact that it is difficult to boost the power of a nuclear plant once it has been designed. Soviet warship designs notoriously grew as work progressed, as more and more improvements were piled on. The choice of reactor had to be made at a very early stage, the design of the reactor proceeding more or less in parallel with that of the ship. At some point, a nuclear plant that would have been perfect for, say, a 10,000 tonner fell far short of what a much larger ship needed. At that point, the only way to regain speed would have been the oil-fired booster. Like much of the former Soviet fleet, the nuclear cruisers of the Kirov class have suffered from lack of maintenance, and only one, Pyotr Velikiy, is active. Two more are undergoing repairs or are laid up, and may or may not go to sea again.

Whether new U.S. Navy nuclear-powered surface ships other than carriers will one day put out to sea remains an open question.