Rickover was interested in the potential of nuclear power throughout the navy. His whole career had been built in naval propulsion machinery, and he had witnessed several major U.S. advances, leading up to the remarkably efficient and reliable high-pressure high-temperature power plants of World War II. Their technology had given the wartime navy unprecedented mobility. One lesson was that new power plant technology had to be spread across the fleet if it was to offer its full potential. For example, a nuclear fleet would gain high-speed mobility, which would protect it from submarine attack (in a pre-nuclear submarine era). It would not be tied to tankers, which themselves might be attacked by an enemy. Once he felt he understood nuclear engineering, he proposed design of a range of larger and smaller plants. The smaller ones might be used to build less expensive submarines. The largest were clearly intended for carriers and cruisers. Chief of Naval Operations Adm. Robert Carney approved Rickover’s program in 1954, before the prototype Nautilus went to sea. The high end of the series of reactors offered 30,000 horsepower, twice what the Nautilus plant put out. The new Forrestal-class carriers required 280,000, so eight of the high-end reactors could power a carrier, particularly if their power could be boosted slightly.

Preliminary design work on a nuclear carrier began in 1955; USS Enterprise was included in the FY 58 program, for the year beginning 1 July 1957.  She was a spectacular achievement, but she was also spectacularly expensive to build and to maintain. Each of her eight reactors required its own operators, for example. The hull large enough to accommodate this power plant was far more massive than that of a pre-nuclear carrier. Rickover argued vigorously that all future carriers should be nuclear, but the sheer cost of the new ship was a deterrent. After one more (non-nuclear) carrier, construction of new carriers paused for a few years (it had been running one per year) when money was diverted to the crash program to build Polaris strategic submarines – another type of warship which Rickover’s new kind of propulsion had made practicable.

Compared to a steam plant, a nuclear plant requires a larger cadre of more skilled operators. Rickover was acutely aware that any nuclear accident would kill nuclear power for the U.S. Navy, so he insisted on high (some would say extravagantly severe) standards for those operating the plants and commanding the ships they powered.

Meanwhile Rickover’s Naval Nuclear Reactor organization strove to simplify carrier power plants. It realized that the key was cutting the number of reactors. It proved possible almost to double reactor power, so that a carrier could be built with four rather than eight, albeit with less power than Enterprise. This ship was not built. Secretary of Defense Robert S. McNamara argued that it would still be so much more expensive than a conventional carrier as not to be worthwhile. Rickover and other nuclear supporters argued that was an illusion. The nuclear carrier would be far less vulnerable, thanks to her sustained speed, she would need far less tanker support (she would still need fuel for her aircraft), and she would be easier to maintain. McNamara’s decision was embodied in USS John F. Kennedy, the last U.S. non-nuclear carrier. Echoes of McNamara’s arguments could still be heard in the 1980s, in attempts to eliminate nuclear power so as to cut carrier cost. The issue was generally the purchase cost of the carrier as compared with the cost of operating her over her lifetime. At the time it was probably not imagined that the U.S. Navy would typically operate carriers for as long as fifty years, far beyond the operating lifetimes of earlier kinds of warships. That was possible partly because the sheer size of these ships limits the stress imposed by the sea.

Compared to a steam plant, a nuclear plant requires a larger cadre of more skilled operators. Rickover was acutely aware that any nuclear accident would kill nuclear power for the U.S. Navy, so he insisted on high (some would say extravagantly severe) standards for those operating the plants and commanding the ships they powered. Experience suggested that for a small ship, a cruiser or a large destroyer, nuclear power entailed too high a cost in personnel. That cost was well worth paying in a submarine. A carrier and her air wing require so many highly skilled personnel that the additional cost of a nuclear power plant was bearable. If, as advocates of energy independence and conservation suggest, nuclear power will have a larger role in the future, the naval nuclear program will provide most of the new reactor operators needed. The Navy will have to compete with a livelier civilian sector, and the cost of nuclear personnel will undoubtedly rise. So, perhaps, will the cost of reactors, if the companies making them have a larger civilian role. Even in the 1950s operators were seen as a major nuclear expense, because they required so much specialized training.

Enterprise was difficult to maintain because her eight reactors were closely coupled together. Like any other nuclear ship, she had to be opened up periodically so that the reactors could be refueled. In a carrier the power plant is buried deep in the ship, beneath the flight and hangar decks. These decks have to be cut open to give access to the reactors; there is no way to get at the vertical fuel rods from the side.

Thus Naval Reactors continued to develop larger reactors that would need fewer operators. By the 1960s the U.S. Navy had not only a nuclear cruiser (Long Beach) but even a large nuclear destroyer (Bainbridge, later redesignated a cruiser). Each had two reactors. The U.S. Navy was then planning a class of Typhon missile destroyers with huge radars, which, it seemed, would need nuclear power to drive them. Naval Reactors developed a single reactor that could replace the usual pair of destroyer reactors. It never entered service, but the lessons learned made it possible for Naval Reactors to double power again (and then some) into a reactor two of which could power a carrier. This new reactor was available when Secretary McNamara left office (1967) and the design of another new carrier began – USS Nimitz.

This was remarkable progress. It was little more than a decade since Rickover had received authority to develop his range of reactors. Now his organization was offering one about four times as powerful – not to mention much more fuel-efficient.

Enterprise was difficult to maintain because her eight reactors were closely coupled together. Like any other nuclear ship, she had to be opened up periodically so that the reactors could be refueled. In a carrier the power plant is buried deep in the ship, beneath the flight and hangar decks. These decks have to be cut open to give access to the reactors; there is no way to get at the vertical fuel rods from the side. That is why other modifications to a carrier are generally held back to refueling time. Alternatively, it might be said that much of the cost of operating a nuclear ship is spent when she is refueled. Eight closely coupled reactors required a huge refueling hole and an enormous amount of special piping.