Seawolf Tries Sodium

by Will Davis

The story of America’s first successful marine nuclear power plant as used on the submarine USS Nautilus is well known.  What is less known today is the real story behind the competing, and in the end losing, design – an “advanced” design using sodium cooling that in some quarters was first thought to be the better choice.

Seawolf at sea


During the years right after the Second World War, a number of nuclear power generating projects were conceived, planned, and ultimately canceled. One of those projects included a sodium-cooled power reactor that the Atomic Energy Commission (AEC) wanted to build perhaps ultimately as a commercial (civilian) power reactor. AEC chose Knolls Atomic Power Laboratory to develop it; Knolls had originally been set up to assist in the Manhattan Project, but now had turned its eye also to power reactor development.

According to testimony given by Admiral Rickover (head of the AEC’s Naval Reactors Branch) to the Joint Committee on Atomic Energy (JCAE) a few years later in 1957, this early sodium-cooled reactor program ran into trouble – budget trouble. In order to ensure it was continued, the project was turned into a Navy submarine power plant project using the same principles in the reactor.  So in 1950, Project Genie was born.


Development of the power plant, including novel electromagnetic sodium pumps with no moving parts (and requiring a quarter of the pumping power of the main coolant pumps of the competitive water-cooled design) moved along slowly and, as with the competitive project, it was decided to build a land-based prototype first.  This was the SIR Mark A (Submarine Intermediate Reactor) prototype built at the Kesselring Site, West Milton, N.Y. The simulated submarine hull for this prototype was moved into the huge 225-foot-diameter spherical containment in a ceremony that took place on March 20, 1954. (The containment, designed by Bechtel, is still the largest spherical containment ever built.)  There was pressure to get the plant running, as the keel for the submarine that would carry the seagoing duplicate power plant had already been laid on September 15 of the previous year.  The submarine would be USS Seawolf, SSN-575.

July 1955 proved an important month in the history of the Navy’s attempt to use a sodium-cooled reactor. In that month, the Seawolf was launched at Electric Boat, Groton; however, also in that month, worrisome leaks developed in the steam generators of the SIR Mark A plant. In fact, there had already been leaks in the superheaters at this plant.

While Knolls and Naval Reactors struggled to plug the leaks and examine a permanent fix, the nuclear plant was installed in the now-afloat USS Seawolf in December 1955. In January 1956 the leaks had been repaired at the prototype, which continued operating to accrue experience (and data) which would be needed to operate Seawolf’s plant. Seawolf itself was fueled in April 1956, and sure enough, the plant on board the submarine developed both leaks in its superheaters and its steam generators.

SSN-575 USS Seawolf fitting out, dockside at Electric Boat, Groton.

SSN-575 USS Seawolf fitting out, dockside at Electric Boat, Groton.

Admiral Rickover described the frustrating nature of the sodium-cooled plant to the Joint Committee on Atomic Energy in 1957:

“We went to full power on the Seawolf alongside the dock on August 20 of last year.  Shortly thereafter, she developed a small leak. It took us 3 months, working 24 hours a day, to locate and correct the leak. This is one of the serious difficulties in sodium plants.”

Rickover then told the JCAE that the Seawolf had had about 10 percent of its heat-exchanger capacity cut out (by plugging tubes) and that the cutting out of the superheaters reduced power another 10 percent on the Seawolf.  “With the reduced power, she makes about 90 percent speed,” he told the members.

Of course, another problem (which Rickover pointed out) was the risk of radiation exposure to the crew from the sodium, which prevented rapid repairs (as the shielded area of the reactor compartment could not be entered until some time after shutdown).  “Sodium becomes 30,000 times as radioactive as water,” he told the Committee.  “Furthermore, sodium has a half life of 14.7 hours, while water has a half-life of about 8 seconds.”  But this didn’t compare to what Rickover considered the primary problem – safety, not only because of the half-life consideration but because of sodium’s violently reactive nature with water.  “There may be advantages for sodium for shore-based atomic power plants but I cannot see it for a ship.  It is too dangerous for a ship,” he directly told the JCAE.


The decisions made by Admiral Rickover as head of Naval Reactors saw to it that the sodium cooling concept for Navy subs and ships was killed.  Rickover told the JCAE in March 1957 that the prototype reactor plant in New York was to be shut down. In fact, it was to be completely removed and the containment reused by a large water-cooled power plant for use in nuclear frigates or destroyers, which would become D1G.

The outcome for Seawolf was guaranteed to be positive simply because, as a matter of design, it had been fitted with a reactor compartment far larger than the sodium-cooled plant actually required. This had been necessary in order to maintain hull diameter and integrity, not because of extravagance or foresight.  In fact, the reactor compartment was large enough that a modified pressurized water-cooled plant could be back-fitted after the old plant was taken out.  There is evidence that Rickover may already have had this in mind when speaking to the JCAE in early 1957, as he ordered a spare SIR core (that could theoretically have been put in Seawolf) to be cut up sometime in 1957.

Seawolf traveled over 71,000 miles on her sodium-cooled plant, and interestingly not one single reactor compartment entry was made during the roughly two year time span in which the SIR Mark B, later known as S2G plant was operated.  However, the end-of-core-life arrived and the boat was put in the hands of its builders in December 1958 to begin conversion to the water-cooled plant.  Roughly 20 months later, Seawolf was ready for trials at the end of August 1960 with her new and highly reliable PWR nuclear plant installed.  SSN-575, USS Seawolf was officially recommissioned on September 30, 1960.  The Navy’s period of sodium-cooled nuclear power was over.


The Atomic Energy Deskbook, John F. Hogerton.  Reinhold Publishing, New York, 1963.

Hearings before Subcommittees of the Joint Committee on Atomic Energy, Congress of the United States, Eighty-Fifth Congress; First Session on Progress Report on Naval Reactor Program and Shippingport Project.  March 7 and April 12, 1957.  US Government Printing Office, Washington, 1957.

The Seawolf Story.  Knolls Atomic Power Laboratory, Undated.

U.S. Submarines since 1945: An Illustrated History, Norman Friedman.  Naval Institute Press, Annapolis, 1994.

Will Davis

Will Davis is a member of the Board of Directors for the N/S Savannah Association, Inc. He is a consultant to the Global America Business Institute, a contributing author for Fuel Cycle Week, and he writes his own popular blog Atomic Power Review. Davis is also a consultant and writer for the American Nuclear Society, and serves on the ANS Communications Committee and the Book Publishing Committee. He is a former U.S. Navy reactor operator and served on SSBN-641, USS Simon Bolivar.  His popular Twitter account is @atomicnews.


5 thoughts on “Seawolf Tries Sodium

  1. Richard Soderholm

    A few additional comments about Seawolf and the sodium plant.
    SIR Reactor was later designated S1G and S2G. (Submarine 1st and 2nd by GE.)
    Nautilus was STR (Submarine Thermal Reactor, later S1W [prototype] and S2W [Sub]).
    I was at SubSchool when Seawolf was experiencing the super heater delays during construction, and I later qualified on S3G.
    The large containment vessel at West Milton site was required to contain a total sodium-water reaction. The submarine hull section was sufficient to contain an accident in a water cooled reactor.
    I had a number of shipmates who served on S1G or S2G and we often discussed it. Later, I was SubRon 10 Engineer for Seawolf when it had the S2Wa plant, but my office still had copies of the original plant manuals when made interesting reading on duty nights. I occasionally rode the boat.
    The sodium plant was known to have high radiation levels after shutdown, so was designed to minimize personnel entry. Various plant thermocouples had multiple redundancies, so upon failure, they just changed connections outside the reactor compartment.
    The super heater leaks were weld cracks. The metallurgy of the day could not handle the high temperature sodium.
    Some emergency Heat Exchangers were sodium cooled by sea water. These were complicated designs, with concentric tubes and double tube sheets, to eliminate the possibility of a sodium-water reaction at a tube joint leak.
    The space between the inner and outer tubes and tube sheets (third fluid) was filled with Mercury for S1G (prototype), and filled with a Sodium-Potassium liquid alloy (NaK) on the submarine. Mercury is not allowed a submarine, as the vapors are toxic. (Even the Medics had to use dial thermometers.)
    The sodium plant required considerable shore power for electric heaters to keep the sodium liquid when shut down.
    There were a number of tanks of fresh sodium near the reactor compartment, and a process to exchange the coolant with fresh, not radioactive sodium, if there had to be repairs done.
    I believe the coolant pressure was less than 50 psi, and there was a pressurizer filled with some inert gas.

    The sodium plant was replaced by a water cooled plant after the first fuel cycle. The plant used the spare reactor vessel and steam generators ordered for Nautilus. Both ships had the same pressure hull diameter, and everything fit. Seawolf used two speed coolant pumps instead of the variable speed pumps on Nautilus.
    The EM pumps on S1G changed flow by changing voltage on a nominal 480 v, 60 cycle generator. The same coolant generators were used for the new two speed pumps at 480 v.

    The water plant produced lower temperature steam that the steam plant was designed for, so the reactor plant controlled for a constant Th instead of Tave.
    The S3G/S4G operated at higher temperatures than prior submarine plants. They required a more sophisticated steam plant to avoid corrosion. They had a De-aerating Feed Tank (DFT) to remove oxygen from the feed water, and had several steam turbine driven auxiliaries to provide low pressure exhaust steam to the DFT. Very similar to a WW II Destroyer plant.

    Richard Soderholm

  2. Ed Pheil

    The reason for the sodium plant was not just another option, but the potential for a more power dense, smaller power plant for more “business” end space on the sub, more weapons ir deeper, OR the ability to go faster, more quietly, in the same size sub. We learned very abruptly of this advantage of the liquid metal plant later when the Soviet lead cooled Alpha sub came out, and our subs couldn’t even keep up long enough to get a decent torpedo shot off that could catch the Alphas, and the Alphas were also extremely quiet when not shredding the water at high speed. The US has never come close to matching that capability of speed, and no longer even try, but do stealth only, hoping the Russians dont im0rove on the Alphas and fix the leaking SG’s, the same problem as the Seawolf, except a fleet of Alphas were at sea for a while. Meanwhile Russia is working on the 300MWe BREST lead cooled commercial plant, still working on improving small On cooled reactors. One wonders what Russia is doing, NOT in public view? Liquid Pb IS very good at damping noise and can reach VERY high core power densities, 10x water plants.

  3. Will Davis


    An astute observation! So far as I recall the details from Friedman and others, Knolls was indeed ready to walk after the sodium cooled episode but Rickover, fearing that there would be only one design laboratory left, decided to refocus Knolls’ submarine efforts on very high power (that is to say, high power density) cores and plants with the result being S3G and the S4G for the Triton, as well as the D1G / D2G. So what he did was dump the losing technology and convert the losing lab over to the winning technology, to put it exceedingly simply.

  4. Charles Bergeron

    Great article Will!

    A couple of notes on Seawolf and Sodium Reactors for subs.
    Seawolf had originally been designed to have a tertiary loop of mercury, which was being used in land based (non-nuclear) power plants to help solve “leak” problems at the time. As originally designed Seawolf might have been more successful – but – the mercury might have made the crew Mad as a Hatter.
    Light water reactor designs need a material to contain the fuel that had a very low barns value. Only the invention of Zirconium alloy at Bettis Atomic Power Labs (similar to Knolls) ensured the success of LWRs. Zirconium is used worldwide in all LWR designs today.
    The EM pump drives were added to a Sodium Reactor design concept to solve noise issues. It also used properties of seawater in a final drive design that did not use propellers. That design concept was considered problematic by the inventor and studied only as a possible future concept if needed. It became the model for the “Red October” of Clancy fiction fame.
    It was interesting to be in the program in the early years and know the inside stories to go with the “public committee reports”.

  5. Rod Adams


    As usual, you have provided an excellent history lesson.

    I’ve often wondered if Rickover didn’t exaggerate the start-up difficulties associated with sodium cooling just a little bit.

    After all, he was a pragmatic engineer with a keen understanding of the importance of controlling cost, schedule and narrative. By taking direct action to focus efforts on the water cooled reactor that was working well enough, he maneuvered his congressional supporters into allowing him to kill off a design option that could only be carried by duplicating a number of different parts of the supply chain.

    He was smart enough and savvy enough to realize that there was strong support for keeping an alternative system viable because it meant more jobs, more research, more prototypes, and, quite frankly, more toys for scientists & engineers so they could remain entertained and challenged.

    All of those “mores” while benefitting many, would have added major costs to Rickover’s program without getting additional capabilities. He was truly devoted to building a large nuclear submarine fleet as quickly as possible to deter the evil communist dictators ruling over his ancestral home.

    He had no time or money to support a jobs program.

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>