The University of Michigan Industry Program of the College of Engineering NUCLEAR POWER FOR COMMERCIAL VESSELS K. Maddocks IP-144 December, 1955

ACKNOWLEDGEMENT We would like to express our appreciation to the author for permission to distribute this preprint under the Industry Program of the College of Engineering.

Advance copy, subject to revision. The Press are asked not to publish this paper, either wholly or in abstract, before 21st December 1955. Contributions to the discussion either at the meeting or by correspondence are invited. THE INSTITUTE OF MARINE ENGINEERS Nuclear Power for Commercial Vessels K. MADDOCKS, B.Sc.Tech. (Associate Member)* To be read on Tuesday, 20th December 1955, at 5.30p.m., at 85 Minories, London, E.C.3, and re-presented on Thursday, 29th December 1955, at 7.30p.m., in the small hall of the Institution of Engineers and Shipbuilders, 39, Elmbank Crescent, Glasgow, C.2. The paper presents a survey of British and American unclassified material relative to the use of nuclear power for marine propulsion. Following a brief discussion on the principles of fission and reactor operation, five types of reactor suitable for marine use and one type suitable for fuel production are described and illustrated. The gas cooled reactor is selected as most suitable for marine propulsion and a proposed closed cycle gas turbine plant is analysed in some detail. Various proposals for the use of nuclear power in specific ships are reviewed, and an economic analysis is made to compare a 30,000 ton d.w. tanker when operating with an oil fired steam turbine plant and when operating with a nuclear powered closed cycle helium turbine. INTRODUCTION unfavourable to the nuclear powered plant as has been suggested. Since the presentation of Sir John Cockcroft's paper (1) on The U.S. Atomic Energy Commission has recently prethe subject in 1953, much information has been released on the pared estimates of the economically recoverable reserves of both subject of power production using nuclear fuels and it seems conventional and nuclear fuels, an abstract of which is given in pertinent that a survey should be made to define how this Table I. evolution in technology may affect the professional marine engineer. TABLE I.-WORLD RESERVES OF FUEL With the world-wide increase in demand for power, which must accompany the present rise in the standard of living and Fuel World reserves Energy in B.t.u. the increase in population, some authorities have estimated that the limit of economical production of fossil fuels will be Coal 3,482x 109, tons 722 x 1018 reached in about 100 years' time. The alternative sources of Oil 186x 10, tons 7.6x 1018 power being developed currently are nuclear and solar energy., c. f While discussion in this paper will be confined to the former, Totalconventional 80-4x1018 it should be borne in mind that the present stage of development of the solar battery has produced an efficiency of the order of Uranium 25 x 106, tons 1,700x 1018 10 per cent. This may well be bettered and applied to trans- Thorium 1X 106, tons 71 x 1018 portation within a decade, but under the present rationing system for sunshine, it seem highly unlikely that this will be Total nuclear 1,771 x 1018 available in the United Kingdom. This paper will discuss the engineering aspects of the design, construction and operation of nuclear powered machinery. The future of any leading maritime nation may well To be of interest to the marine industry, this paper must consider eventually depend on the availability of reactor technology and the economics of the nuclear plant. Authorities contend that production potential. It will also depend on the location of nuclear power production ashore, providing existing develop- sources of fissionable material of which uranium is the most ment schedules are maintained, can be competitive with fossil promising. The military and political significance of this fuelled power production in about ten years' time. It seems question is outside the scope of this paper, but the world-wide unlikely that a nuclear powered marine plant will show any interest in the use of nuclear energy can be deduce from a economical advantage before that time since, in the author's study of the map in Fig. 1. opinion, one of the main factors will be the source of supply The details released in the recent White Paper covering of fissionable fuel at a reasonable price and this must probably Britain's ten-year plan for nuclear power development convey await the actual operation of a land power station using a breeder a note of optimism despite the capital cost involved. The reactor. Probably one, and possibly two, nuclear powered White Paper states that the fuel supply prospects are now better merchant ships will be in operation within the next five years. than previously anticipated Considerable deposits of medium These will not be competitive in either first cost or operational and low grade uranium ores are known and thorium has distinct cost with vessels propelled by orthodox machinery, primarily possibilities for conversion to a nuclear fuel. The Government because of the inevitable expense attached to the development is confident that the necessary supplies will be available when of any new type of machinery. However, outside of this required. In dealing with an installed capacity of eight power factor, it is hoped to show that the balance will not be as stations in excess of 1,000 megawatts, this White Paper con__ __ _ __ _ ___ c__ ludes, "This formidable task must be tackled with vigour and * Assistant Professor of Marine Engineering at the University of imagination. The stakes are high, but the final reward will be Michigan, Ann Arbor, U.S.A. immeasurable ". I

Nuclear Power for Commercial Vessels xCANADA xN R xCANA BRITAIN E BRI T \ ~a DENMARK \X NETHERLANDS e ^ v _,BELGIUMA EASTGERMANY, RUSSIA FRANE *CZECHBSLOVAKIA0.,., SWITZERLAND It / \ /*OZAMBIQ AUSTRALIA CHILEO'V J oARGNTINA SOUTH AFRICA \ By courtesy of Standard Oil (New Jersey) FIG. 1-World map showing location of uranium and reactors Uranium countries * Now producing or believed capable of producing at current prices o Not fully explored but possibly capable of producing at current prices Reactor countries X One or more built A Active research or announced plans to build The author suggests that the statement could well be involved. Radiation decay and corrosion and the properties applied to power production in the marine industry, of the many new materials now in use also present problems. The main object in writing this paper is to foster interest The heat produced by fission is, of course, on the credit in this new source of power, by replacing the " imagineering ", side of the ledger, while the three "by-products ", alpha, which has so far accompanied the magic words "atomic beta and gamma radiation (briefly this order denotes increasing energy ", by a more rational survey of the facts. Those path length and decreasing ionization) are on the debit side, colleagues whose daily menu includes isotope hors d'oeuvre, as they require special shielding to prevent a health hazard. potage uranyl sulphate, plutonium pie with beryllium dressing, All reactors to date have been designed to use either etc., will find little sustenance in this article. They are never- Uranium 235, Uranium 233 or Plutonium 239 as fuel. Thorium theless very welcome to partake in the feast over the author's 232 is a fertile material, which can be converted to a fissile bones, which it is hoped will follow the presentation. Security material in a breeder type reactor. (The numbers indicate precautions will no doubt limit the range of the discussion, the atomic weight or the sum of the neutrons and protons in but the author feels that the present state of published knowledge the nucleus of the atom of that material). It is significant to offers ample scope for debate. remember that we are discussing the use of a metal as fuel. All the subject matter referred to in the presentation of Uranium has a density 1-5 times that of lead, its melting point this paper has been taken from unclassified sources and thus is about 2,000 deg. F., it is malleable and ductile and can be there may be an incomplete discussion on certain items. readily machined or cast. The fact that when in powdered form uranium is highly inflammable has no connexion with DERIVATION OF NUCLEAR POWER its use as nuclear fuel. In the homogeneous reactor, it is used Several excellent texts (2,3,4) have been published on the in the form of a salt, uranyl sulphate, and dissolved to form a principles and applications of the new technology and to include liquid fuel. a similar complete treatment is outside the scope of this paper. Natural uranium, U.238, contains only 0-7 per cent U.235 However, in the interests of continuity, the following points by weight; the other two fuels, U233 and Pu.239, must be proshould be borne in mind. duced artificially. The degree of enrichment of the fuel is the Fissioning is the splitting apart of the atomic nuclei of the proportion of fissile U.235 to non-fissile U.238. The higher material used as fuel. The addition of another neutron to the enrichment, the more efficient " burn-up " of the fissionable the nucleus of a fissile material is sufficient to cause an agitated U.235 can be expected. Of course, the production cost of the state and subsequent split up of the nucleus. The kinetic fuel is in proportion to the degree of purity required. In the energy of the fission fragments is dissipated in the form of heat present state of the art, any of the fissile fuels discussed above and other radiation. are costly although the prices are currently in control of governTheoretically, one pound of nuclear fuel, which has a ment agencies, e.g. Atomic Energy Authority in Britain or volume slightly in excess of one cubic inch, if completely Atomic Energy Commission in America. The price of fissile fissioned, releases energy equivalent to 43 x 109 B.t.u. or fuel should be considerably reduced, when in the not too approximately 1,000 tons of fuel oil. It is not possible to distant future, the breeder type of reactor is put into service arrange complete fission of the fuel. Only a fraction can be for power generation ashore. As described in a later section, used before chemical treatment is required, due mainly to the these stations are expected to produce adequate quantities of inevitable simultaneous production of reactor "poisons ". fissile fuel as a by-product. The transfer of such highly concentrated heat to a usable form of working fluid requires a complex system for coolant circula- TYPES OF NUCLEAR REACTORS tion. Critical control mechanisms, remote fuel handling, Reactors are classed as either " thermal ", " intermediate " shielding and heat exchangers are also required. Many of these or "fast", depending on the energy level of the neutrons. require entirely new concepts in design, due to the increased In a "thermal" reactor, the neutrons are slowed down conheat transfer rates and the variation in static and fluid mechanics siderably by a moderator before continuing the fission chain 2

Nuclear Power for Commercial Vessels reaction. When the moderating action allows a higher neutron B energy level to operate, the reactor is said to be of the " intermediate" type. If no moderator is provided to slow down the neutrons, then the reactor is "fast" and in some cases a fuel diluent may be necessary to spread the nuclei and so decrease the thermal flux density....:: The form of the fuel, coolant and moderator create further classifications. In a homogeneous reactor, the fuel, coolant and moderator (if used) are mixed, often in liquid form. In a heterogeneous reactor, these are separated usually in solid form, which allows a definite geometric arrangement, e.g. round rods of fuel can be slipped into the moderator block in much the same way as marking pegs are placed into a cribbage scoreboard. I t A reactor is said to be regenerative if it replaces all or part of the fissioned (burned up) fuel by creating new fuel from nonfissionable fertile material. If this replacement is equal to the amount of fuel consumed plus some excess, then the reactor is known as a breeder. Other distinguishing characteristics are the enrichment and type of fuel, the coolant used, and, for the thermal reactor, the moderator used. A complete discussion on the recommended materials to fulfil each of these functions is given in reference 5. li' " OPERATION OF A NUCLEAR REACTOR Unlike the oil fired boiler furnace where the fuel is pumped:. ^ in, atomized and burnt in a continuous process, the nuclear 1: "~ fuel supply for a complete run between refueling ports would By courtesy of the Westinghouse Electric Corporation be carried in the core of a reactor (this excludes the homogeneous reactor). The quantity of fuel so carried would be FIG. 3 mote controls nd pescope sighting device for determined not only by an allowance for " burn-up " and an handling radioactive materials excess to overcome the effects of reactor " poisons" which are simultaneously produced when the fuel fissions, but also from neutrons, some means of adjusting the " k " value is required the concept of providing an accumulation of fuel sufficient to to prevent a spontaneous build up to criticality. This is proproduce and maintain criticality. vided by control devices which have a high capacity to absorb The reactor is said to be critical if at least one of the neutrons neutrons. In the heterogeneous reactor, these devices are (two or three are produced with each fission) is available to usually in the form of rods and shims of either cadmium, split up another nucleus and thereby maintain the chain reaction. cobalt, hafnium or boron steel alloys. The movement of the This neutron multiplication factor (usually denoted by "k ") rods gives a coarse control and the shims a fine control of the therefore determines whether the reactor is critical or not. If the neutrons. In the " cold " position, both rods and shims would value of " k " is unity or above then the reactor is critical, but if be full in. The power level of the reactor is selected by the the value falls below unity, the reactor becomes subcritical and withdrawal of the rods a predetermined amount. Then the the chain reaction ceases. gradual withdrawal of the shims further increases the " k" As a fissile fuel has the property of continuously emitting value of the reactor until it becomes critical and the system gets under way. The control shims are also used to compensate for fuel "burn-up" during reactor operation. An emergency control device can be fitted so that if a shut down is required both the control rods and shims are rapidly driven into the core and the reactor immediately becomes sub-critical. This arrangement is aptly known as the " scram" control. Since operating personnel would normally be located in a control room (6) without access to the "hot" reactor, the various operating factors such as control rod position, strength of neutron flux and coolant flow, etc., must be measured and transmitted to the control room by various instruments. These individual signals must then be used to operate automatically the reactor in a stable condition at the power level required. Possible radiation hazards in the form of gamma rays or escaping neutrons must be detected by an elaborate system of sensing instruments located both inside and outside the reactor. The marine installation would also require these instruments on the ship's hull and on the ventilating system and sanitary system. The removal of the spent fuel and its replacement by new fuel requires the use of remote controlled handling gear such as the mechanical tongs shown in Fig. 2 and the operating station shown in Fig. 3. These are photographs of the equipment used in the development of the full scale pilot plant, which was operated at the National Reactor Testing Station, Idaho, before a second reactor plant was installed in the U.S.S. Nautilus. Shipboard equipment would be essentially of the same design. 13 cour.tesy of the W/estinghou.se Electric Corporation The major portion of the so-called " spent " fuel is fissile after chemical processing; therefore, it has a high salvage value. It FIG. 2-Remote controlled handling devices for radioactive would usually be removed from the ship and dumped in a material tank of water to permit the fission product heat to decay to a

Nuclear Power for Commercial Vessels tolerable level. Then, encased in a " coffin " of lead, it could investment of $200 million will be made in developing five be transported to the re-processing plant. The disposal of the separate types of power reactors. It is anticipated that this actual waste product must ensure that it does not become a plan will bring within sight the objective of harnessing nuclear health hazard. This waste will remain radioactive for a power on a basis economically competitive with coal and oil. considerable period and two methods have been used for its The first four reactors are versions of those included in the disposal. One is to bury it in a concrete vault in as remote a United States A.E.C. plan. The fifth will use a closed cycle location as possible. The other is to encase it in concrete and gas turbine as a power-producing unit, operating with helium dump this out at sea. This is certainly not a commodity that as a working fluid. can be kicked around until it is lost. The characteristics of each reactor design will now be considered. The illustrations should be read liberally. Detail SELECTION OF REACTOR TYPE FOR MARINE USE design of a particular reactor depends on the type of vessel in It will be appreciated that by various combinations of the which it is to be used. For example, the control rods and the characteristics outlined above, the number of "possibilities" fuel rods of the heterogeneous reactors may well be more is very great. conveniently fitted on mutually perpendicular axes. Also, the Fortunately, this range can be narrowed considerably by use of concrete for shielding is shown on all the reactors. Steel, space and weight considerations and also from the fact that a lead or a composite structure could be more suitable. The plant capable of producing replacement fuel is precluded. probability is that lead will, in general, be found to be the most Such a breeder reactor would require a far too extensive suitable shielding material for marine use. ancillary chemical plant and shielded material handling equip- Pressurized Water Reactor ment to be accommodated on shipboard. The design of a The pressurized water reactor, as shown in Fig. 4, which is suitable vessel must include provision for loading and dis- essentially as fitted in the U.S.S. Nautilus, has become the charging packaged fuel and waste material. These arrange- " pioneer " marine plant(7). ments will be dealt with in a later section. At this stage, also, The fuel rods of the reactor could be of enriched uranium some approximation to the power output of the proposed plant or plutonium, probably clad for strength with a metal which has must be made. To exploit fully the main advantage in reduction low neutron absorbing capacity, such as aluminium or zirconium. of fuel weight and to offset as far as possible the expected high The control rods would be machined from a cadmium steel first cost and fuel cost, a minimum of 15,000 s.h.p. is indicated, alloy or a boron steel alloy. The present monopoly of the steam turbine in this range Highly purified water under pressure forms both reactor of power, with the consequent accumulation of design and coolant and moderator to make the reactor operate at "thermal" operating technique, has no doubt influenced the choice of steam energy level. For best heat transfer conditions in the reactor, as the working medium. It has, in fact, been stated frequently the primary water velocity is increased by restricted passage that the only difference between a nuclear fuelled steam plant cross-sections. By maintaining a high rate of flow through the and a fossil fuelled steam plant is the type of boiler used. primary circuit, the temperature rise of the pressurized water That this is an over-simplification will be seen from the dis- is kept to a minimum. cussion of some of the possible reactor designs which follow. Heat is transferred from the primary circuit to the secondary Also, the operational control of the fluid conditioning unit circuit in a tubular exchanger and the steam so formed is separ(reactor) and the turbine must be more closely integrated than ated in the steam drum of the steam generator. Presumably is the case even with advanced steam plants using automatic this could be arranged for either natural convection or forced combustion controls. convection, depending on the steaming rate required. A list of feasible types of reactor. with no implication of One obvious disadvantage of this system is the impracticathe order of precedence, then becomes: bility of producing superheated steam. As an example, 1. Pressurized water reactor. assume the primary circuit water is pressurized to 1,000 lb. 2. Boiling water reactor. per sq. in. abs., the maximum temperature in the primary circuit 3. Homogeneous reactor. must be below the equivalent saturation temperature 4. Sodium loop reactor. (544 deg. F.). Allowing, say, 10 deg F. loss during transmission 5. Gas-cooled reactor. to the heat exchanger and a mean temperature difference This list was computed from study of the U.S. Atomic between the pressurized primary water and the evaporating Energy Commission's published five-year plan in which an secondary water in the heat exchanger of, say 60 deg. F., then Access to fuel and control rods o. *: ~:.'.'.','l ~ I...'.......:.'.'4'...'.'..,::'.'.'.'~''.' ~ ".':... /, * ~ beJ __' 4','_. _.' -.'.'''''' 4'. >e ~ o / Pr'imary shield ^....:g y —-~ _____G.' 4 Steam ~: - -:i- l=:\ Turbine:O.' core~_~ ___ _...' Condenser 4 FIG. 4 —Pressurized water reactor

Nuclear Power for Commercial Vessels.-"' ~ -''-. ~.: i ^' a''.'9.'' ^. ~ utilization of saturated steam. Possible modifications to the;' Rod - 4 modern steam turbine designed for superheat would be interRod' stage extraction and centrifuging of the steam, the fitting of.', a: ^ _ _ ~.. semoisture throwing rings as blading shrouds, and it would nil _. Secondary shield appear necessary to use blading at the l.p. end of the turbine,:'~.'' I~~ aI'.~ which has a high resistance to erosion. With present-day techniques in design and material production, designers are now:'~,l l,:. *in a far better position to tackle this problem than their pre~1 ~ =~ DP Steam decessors, who faced exactly the same problem before the..^I Xo^.~ D~ \~ sdevelopment of an effective superheater. Nevertheless, the ~",.. ~ v'. ^: J jprinciple of the acceptance of this inherent deficiency is, in the *I%* s n..! 8 u. i^edA author's opinion, open to criticism. 1:.'O~ |11,-'' dm -^ _^ e11Boiling Water Reactor ^~~.;~ iiF~,~~~ ='aI~.'I~ iFig. 5 shows the boiling water reactor in which steam is; ~a "-1; "'- 6 Turbine generated by direct contact during water circulation through b'. I; _ _:the core. The arrangement is thus a simplification of the 4:.'.' ~'i' i t ^. pressurized water reactor in that one of the loops is eliminated.''~,, "D 1.Condenser This, however, has two additional disadvantages. First, during -':. I CoreED * c,} + ) the boiling process of the water, which acts as moderator as Water:.':' Hil. 1 \ R11. ~~i~ well as working fluid, the variation in density allows a variation make-up' Water in leakage of neutrons, thus causing a fluctuation in power level..~ _ _ -= ~'_.'. * ~ ~ j > of the reactor. Secondly, the steam passing off to the turbine.;. C ( -TT-\ ~' ) will be radioactive, thus producing an additional shielding'.'.'.1 -'' problem. It seems likely, therefore, that the boiling water |.*~. ~c' TP. ~''' d *'.|reactor would be operated at a lower power level than the.*... -. pressurized water reactor. However, it is reported that the Ob'.~ I" ^''U.S. General Electric Company has expressed a preference \Primary sh'ied for this design as a long term possibility, and indeed experiments conducted at the National Reactor Testing Station in Idaho FIG. 5-Boiling water reactor and at the Oak Ridge National Laboratory have confirmed that these reactors can give stable operation. saturated steam at 400 lb. per sq. in. pressure will be produced. The materials for the fuel and control rods, which form the The United States A.E.C. reactor of this type will generate reactor core, could be the same as those used for the pressurized approximately 300,000 kW of heat which will be transferred to water reactor. Another feature that the two types of reactor the heat exchanger by circulating water at 2,000 lb. per sq. in. have in common is that they can only produce saturated steam. and 525 deg. F. Saturated steam at 600 lb. per sq. in. pressure This can be seen at a glance from the diagram in Fig. 5. will be generated in the heat exchanger and passed to a steam turbine generator which will have an output of some 60,000 kW. Homogeneous Reactor It is expected that this installation will be in operation late in The homogeneous reactor shown in Fig. 6 was designed 1957. primarily to overcome the essential limitations of the heteroAn inherent advantage of this system is that the expansion geneous reactor of which the two previous reactors are examples. of water with temperature rise allows an increased leakage of These limitations are: neutrons and, hence, a system which to some degree is self- (a) The separate core components of fuel, control rods, stabilized. coolant and/or moderator in the limited area of intense No claims are made that this type of reactor will produce heat, create a real heat transfer problem. economical power, but it is the type on which most experience (b) The core structure is subject to radiation damage. is available, and, therefore, it could be claimed to be the most (c) The accumulation of fission products caused by the reliable. An improvement in neutron economy could be made absorption of neutrons necessitates the periodic removal by using heavy water instead of light water, and the thermal of fuel for reprocessing. This, as previously discussed, efficiency might be raised slightly by increasing the circulating is a complex operation. water pressure. Pressure tightness of the system becomes of As the name implies, the homogeneous reactor operates increasingly greater importance with either of these changes. on an intimate mixture of fuel and coolant/moderator in the The principal bogey remains in the form of the efficient form of a solution of uranium salt in ordinary water. The ~:'' Q. ^^.">*.<':'~'.:.; Fission.. Steam products ~'..'t ea extraction Waterea' -='/Heavy water —--. P... ".''a —~~~ i l i~ \ exchonger Condenser solution D' 2l o,0 storqage mote-up'.. ~* 4',, v~'''4,..*' ~4','-.' " *,'',' * * *' 4','' ~'' *. ~'. ~'..'.''.A'.''.'''.'o.'.''"*co'' FIG. 6 —Homogeneous reactor 5

Nuclear Power for Commercial Vessels primary circuit, through which this solution is circulated under one of the main application problems by allowing the production high pressure, consists basically of the spherical container of moderately superheated steam (say, 600 lb. per sq. in. and which forms the core and a restricted passage to the heat 800 deg. F.). The higher level of reactor power output exchanger. increases the degree of radioactivity in the sodium and this The form of the core is such that during passage through provides a more difficult shielding problem than is encountered this sphere a sufficient accumulation of the solution occurs to in a thermal reactor. Also, the preparation of a sodiumreach criticality and, hence, the fuel will fission. Whereas, cooled reactor for operation must include arrangements for during passage through the constricted circuit to the heat external melting of the material prior to circulation in the exchanger, the spreading out of the solution will decrease the system. In fact, an auxiliary oil fired " boiler " will be required mass below the critical and so quench the chain reaction. Steam for this purpose. can be produced in the heat exchanger at reasonably high Another major potential danger is the possibility of leakage pressure but again it is saturated. In this case, however, the if the highly radioactive and strongly alkaline sodium were in steam is not radioactive. close proximity to the steam system. Any such leakage would The reactor core is surrounded by a neutron reflector of produce a violent reaction with water. Thus, the coolant heavy water and arrangements are made for addition of fuel and system is separated into two stages to provide a partial solution. removal of fission products while the reactor is in operation. In the primary heat exchanger, the radioactive sodium from One of the most striking features about this design is the the core gives up heat to non-radioactive sodium which is then absence of control rods. These are not required as the system used to heat the secondary heat exchanger or steam generator. has been proved to be self-regulating; e.g. with the main Another method of safeguarding against the contact of circulating pump stopped temporarily and the heat exchanger liquid sodium and water is the use of double-wall concentric cooling down, the uranium salt solution becomes more dense. tubing in the secondary heat exchanger. The annulus of this On recommencing circulation, the power output of the reactor tubing could be filled, say, with lead, giving a good heat transfer shoots up until design level is restored. Then, by the time the bond and a leak detecting medium. solution reaches design temperature, the solution has expanded A further line of thought has been developed (9) in the to offset the reactivity and the power output levels off. suggestion of benzine as a working fluid to replace steam. To summarize the characteristics of this homogeneous Benzine is chemically inert with sodium and, therefore, the reactor, it could, therefore, be said that the nuclear stability danger of violent reaction resulting from any leakage is or safety is purchased at the expense of providing a completely eliminated. It seems probable, however, that the use of leakproof system for a highly radioactive and corrosive solution benzine would not find favour in the marine field on account subject to a pressure of at least 1,000 lb. per sq. in. The f the fire hazard introduced. description and results of an experimental model of this type roa the aor en have now been published(8). It is interesting to note that Probably the major engineering problem to be faced with a lthough several leakiIs w ere experienceduring thoe start-p this type of reactor is the pumping of a liquid metal at a temperaalthough several leaks were experienced during the start-up ture of about 1,000 deg. F. Any leakage would produce both phase, the plant finally operated for twelve months without te ioacti e and a fie hazard Any leakage would produce bot ~any leakage being detected. a radioactive and a fire hazard. A typical specification for leakage tolerance is one cubic centimetre in ten years and to Sodium Loop Reactor meet this demanding service two types of pumps have been Fig. 7 shows the variant in heterogeneous reactor design developed. One is an electro-magnetic pump which eliminates using liquid sodium as a coolant. A version of this type of the usual rotors and, consequently, the shaft glands. This type reactor is to be used on the second nuclear powered submarine is reliable but its efficiency is low. The other type is a centriU.S.S. Sea-Wolf. Sodium being a weak moderating material, fugal pump using fluid bearings. This has a much higher a separate moderator will be required and this could be of capacity, but is subject to the usual mechanical failures. A graphite block construction similar to the original piles at pump suitable for slightly less arduous duty is described in Harwell and elsewhere. For shipboard use, the quantity of detail in reference 10. moderating medium required can be considerably reduced by Standby pumps in the system must be provided in triplicate designing the reactor for operating at an energy level above the or probably quadruplicate, as any maintenance required on a " thermal ". Sodium at atmospheric pressure has a boiling pump can only be undertaken after a waiting period of maybe point of 1,600 deg. F.; therefore, the upper reactor temperature days before the radioactivity has "cooled ". There is also is not controlled by system pressure and large temperature the problem of " freezing" of some remaining coolant which variations in coolant can be arranged. While this leads to could easily provide additional work and the scrapping of design problems incurred in thermal stressing, it also overcomes components. fuel and control rods Steom to turbine'.'j. l..... ~~ ~''.~.~"..'''..'. --.'Secondory heat:'.''..'~.'- -.~''.' Primary heat exchanger.. exchanger l r b ter from condenser.'.':'> A. -.....",'-.:'~~;'.' <;'.:.'...,i, FIG. 7 —Sodium loop reactor 6

Nuclear Power for Commercial Vessels Fuel and control rods /Fuel ad control rod metal container, thereby reducing fuel reprocessing costs,.;.. n r i ~I..- -." although this could introduce a danger of failure of the fuel 4. H'.' * j 1""''4 < ~~ " "4 A^. elements during operation. The moderator is provided in the'o.'. ~'4l^ i^ l'a h'" -'' 4' C 1;form of block graphite or beryllium oxide and the control rods ~~4"~.. 1,' __ _ -'.!.^ could again be of boron steel or cadmium. Control of this.''.4 ~<:^ type of reactor would be both easy and safe, consisting merely of Hotgoasoutlet moving the control rods in and out. -.'.__ Preliminary designs and outlines of equipment have been 7''"~ 2 C 7.''\~.4 Aprepared for a submarine installation (13) to compare the use, a /'' of water, sodium, and helium as coolants in the nuclear power plant. The characteristics are summarized in Table II. TABLE II.-COMPARISON OF WATER, SODIUM AND HELIUM AS COOLANTS Reactor coolant Water Sodium Helium 1:[.^1 E ^4 ^.''' I~Shaft power output 0.90 1 1 Overall plant weight 0-97 1 0-64 4-:- ^odMOa t /t - Specific weight, lb. per s.h.p. 1-08 1 0-64 -:-~~~' d / 4'. -; $Blower Space occupied, cu. ft. per s.h.p. 1.10 1 066: / X:' Shield weight 077 1 0-51 l.'~ Pressure I'.~' vessel l X - / Cool gas inlet While these figures appear to favour the use of helium,;.......their validity awaits the result of a good deal of development'"-'.:'~'' *'.'::' --'.'4 *'~''* ~'"''*'~.'' ~ work to confirm a number of assumed factors. Of the five reactor types already discussed, the marine Shield application of the gas-cooled reactor, as outlined in a later FIG. 8_Gas-cooled reactor section, offers what the author considers to be the most favourable balance between first and operating cost and simDetails of suitable materials for a reactor of this type are plicity and safety in operation. given in reference 11. THE BREEDER REACTOR Gas-cooled Reactor Although not likely to be used as a shipboard power The gas-cooled reactor, as indicated in the recent White producer, some mention of this type of reactor is justified in Paper, is Britain's choice for development as a power reactor that it seems likely that it will provide a definite link in the ashore. The arrangement of these land plants is presumably as application of nuclear power to marine propulsion. This link envisaged in the recent Institute Section paper(l2). This pro- could well be the production of fissile fuels, available to the vides for the reactor coolant gas to circulate a steam generator in marine industry and others, at a price lower than that at which much the same manner as existing types of waste heat boilers, uranium ore could be mined, processed and marketed. except, of course, that for the nuclear plant the gas would be Two reactions have proved of interest in the manufacture in a closed circuit. The steam produced, again probably of fissile fuel. The first is that when the natural uranium saturated, with its attendant complications, would be utilized U.238 is subjected to a bombardment of neutrons, as occurs in a turbo-generator set. in the core of a reactor, the nucleus picks up an additional The gas-cooled reactor shown in Fig. 8 is of the hetero- neutron and thus becomes a new element or isotope, U.239. geneous thermal type. The fuel could be slightly enriched This has a nucleus which is not stable and, therefore, it decays uranium, the rods of which are clad with zirconium, aluminium to plutonium 239. The second is a similar reaction commencing or stainless steel to minimize neutron absorption. There is with thorium 232 which captures an additional neutron to also the possibility of using fuel in powdered form sealed in a become Th.233 and then decays to U.233. - Sodium r-Condensate Cylindrical building potossium storage tanks cycle Nak expansion / X team tank generator Sodium pump Feedwoter Rmotor \ heaters -, Fuel transfer \__.Uunit F —F eedw oter,e'......'7'~::'' ~ ~ i" I[~Condenser,,.i~::l}::it.f". f'-'f:f'..l' ~' /-$~ umpotasL-Bo'lerf:eed \""............. o~Sodium cycle By coy courtesy of the Detroit Edison Company FIG. 9-Cross section of reactor power station

Nuclear Power for Commercial Vessels By courtesy of the Detroit Edison Company FIG. 10-Model of proposed power plant Both Pu.239 and U.233 require chemical processing to separate them from their respective parent materials. There are, of course, many ways of modifying the various types of reactors already discussed to take advantage of this nuclear phenomenon. The homogeneous reactor, for instance, could be made into a breeder by replacing the heavy water neutron reflector by a thorium solution "blanket ". In the heterogeneous reactor either natural or depleted U.238 could be arranged to surround the core and then be removed for processing after the transmutation occurred. l In order to achieve the highest breeding gain, the neutrons must be kept as energetic as possible. To do this, the modera- e ting material must be eliminated to allow the energy level to Irise so that the reactor can operate on the " fast " energy level. The first experimental reactor of this type was referred to in reference 1. It was built and operated at the Argonne National Laboratory and is to be followed by another of similar design, but of much larger scale, to give a heat output of 62,500 kw and an electrical output of 15,000 kW. Encouraged by the results from Argonne and after a survey of the many possibilities, the Atomic Power Development Associates, one of several groups of American power companies and machinery manufacturers, are now working on the design and development of the power plant shown in Fig. 9. A photograph of the model of this plant is shown in Fig. 10. This represents what is probably one of the most advanced designs proposed and the reasons for its choice, despite the " pioneering" work required, merit quotation (14). " Our reasons for selecting the breeder type of reactor were (a) our belief that a reactor which will produce both heat and fuel holds the greatest possibility of commercial success, and (b) our belief that large scale use of atomic energy for power generation can be achieved only by utilizing a large part of the total heat potential of uranium, rather than the 3 per cent to 7 per cent which seems to be the limit of most thermal reactors which use U.235 or plutonium as fuel. A breeder reactor theoretically offers a possibility of using all of the heat potential of uranium, but from a practical standpoint it likely would.1 _ li succeed in utilizing only about 50 per cent. At the same... time it would produce more atomic fuel than it consumes." The plant consists basically of the same design as shown in'" " Fig. 7, except that for the larger plant both primary and. secondary loops can be either duplicated or triplicated and the medium to be used in the secondary loop is a sodium-potassium By courtesy of the Detroit Edison Coi pani alloy in place of the straight sodium. The design as envisaged at present will produce steam at 600 lb. per sq. in. and 730 deg. FIG. 1 —Cross section of breeder reactor

Nuclear Power for Commercial Vessels 100 F00 12-Helium-air //he / H.P. receiver-c Transfer L.P. accumulator 8o "o OF / /,', /~ i 7-y_ / ~ ~ Make-up supply i Iort. /haim-ti nesa to mak sc___er Control c t eroo psilF. whi ch, while conservative by present munads sstanbes is 1-lst-vill high vuieplt /f / / B/ crFter o turbine orin importance in that it is necessary to make it bypas covampact as Film coefficient B.Th.U. per hourper sqcft per deg. F. FIG.reahed between the extremes of providing a good heat transfer characteristics bond and the effects of decay due to the high level radiation. in Switzerland for sixteen years and all the machines so far put F. which, whilustration of the rod control and fuel handling into service have by present standards, the maximum output of any pmenough in Fig. to operate set being 12,500 kW. Nitrogen, which acceptable efficiency. Safety measures have controlled the design of the structure, cent of air, has very similar characteristics. Carbon dioxide as choican be seen fro m these lower sheat transfer properties, but allows a ree become radiocasing, which completely encloses the reactor plant, is airtight active when heated in a nuclear reactor. Helium, however, to prevent the spread o f radioactive contami nation in the un- if kept free of contamination uring circulation, has a ikely event of a failurmportance in the system nucleus which is necessary stable under neutron flux, thus compact does ~~~~p~~~not become radioactivbe. It follows wthain this core that a compromise must be FIG. 13-Closed-cy gas turbine plant NUCLEAR-POWERED CLOSED-CYCLE GAS TURBINE PLANT only the reactor xtrequires of providing allowing a far more flexible bondThe hea t en ergy in the gas of decay due to the high level ra n be directly machinerland for sixteent an a considerabll the saving in weight. An illustration of the rod contrlosed-cycle gas turbinel and fuel andHeli umng into service haveeat transfer charated on ristic the maximum output of any equipment is given work ing fluidg. 1. could be eitherng 12,500 kW. Nitrogen, carbon- gen, which compis enhanced with increase in pressure as shown in Safety measures have controlled the design of the structure, cent of air, has very similar characteristics. Carbon dioxide as can dioxid e or heliu m, the latter is preferr ed. The s team generating Fig. 12. Thus, the heat transfer s urface required wibut all three become radioand condensing, which completely enclost ares thus eliminatedactor plant, s are also r educ ed compared with an air r ni trog en system, but the highver, to prevent the spread of radioactive contaminationing and the ma ny other specif ic heat free of hlight contamination design of turbo machinery, has a ancillary event of a failure the systems connectednucleus with a moder stable under neutron flux, th es it does not become radioactive. It follows that in this type of plant, NUCLEAR-POWERED CLOSED-CYCLE GAS TURBINE PLANT only the reactor requires shielding, allowing a far more flexible The closed-cycle gas turbine has been under development temperatur e is roughly arrangement and aoportional to ble saving in weight. y cortesy o the Aerical work n a closed-cycle gas turbine and Helium has a better heat transfer horpora cteristi c than trowhile the working fluid could be either air, nitrogen, carbon- gen, which is enhanced with increase in pressure as shown in dioxide or helium, the latter is preferred. The steam generating Fig. 12. Thus, the heat transfer surface required will be and condensing equipment are thus eliminated, as are also reduced compared with an air or nitrogen system, but the high make-up feed, boiler water conditioning and the many other specific heat of helium makes the design of turbo machinery ancillary problems connected with a modern steam plant. more difficult. The number of stages required for the same The closed-cycle gas turbine has been under development temperature rise is roughly proportional to the specific heat FIG. 14Typical helium turbine 9

Nuclear Power for Commercial Vessels':.. ~-.-:~, discuss the merits of the closed-cycle air turbine and a large amount of the subject matter applies to the helium turbine. Fig. 13 is a diagrammatic representation of the suggested plant. Expansion is in two stages to isolate the power turbine from the compressor drive. Reversing can be accomplished either by a reversible pitched propeller (19), but the upper limit of power which could be absorbed by the propeller may not be compatible with the minimum power required for an effective nuclear plant; or the power turbine can be built as a reversible inwards flow radial machine; or a turbo-electric drive could be l.:. adopted. A typical axial flow turbine suitable for an output I' ~ of about 20,000 s.h.p. is shown in Fig. 14. One of the most pressing engineering problems in producing a gastight system is the design of a suitable turbine gland. The heat transfer x/'_.; I >\ surface in the recuperator and precooler could be of the plate fin type as shown in Fig. 15 and the intercooler of shell and U-tube type. A typical section and general arrangement of a 60-MW turbo plant of the same type as that proposed is shown — I _ I in Figs. 16 and 17, from which it can be seen that the elimination \ J1 of gas ducts between components ensures a minimum pressure ~k —?; loss and potential source of gas leakage. Control System am~-~ X g. N^The power output of this plant varies with the system -^~~ ^ pressure. This pressure level control is effected by addition or withdrawal of working fluid from the circuit, and emergency FIG. 15- Griscom Russell plate fin heat transfer surface speed control of the power turbine is effected by bypassing. Helium that is not being circulated in the plant is stored Fin material......... SA-204 carbon 1/2 molybdenum in accumulators for subsequent use, making this a no-loss Fin thickness, inch...... 00145 system. The accumulator system consists of two (or two groups Effective fin height, inch... 1238 of) storage bottles, one being the receiver and the other the Centre line to centre line flat accumulator interconnected by a transfer pump. In this plate, inch......... 0'2655 plate, inch........... 0022 system, the total amount of helium in the power plant and tanks Plateh thickness oflow, inch...... 002is constant at all times. Any leakage loss is made good by Fin pitch paranllel to flow, inch 5100 addition of helium to the receiver from time to time as required. Fin surface/total surface... 0-7974 A simplified diagram of this sytem is shown in Fig. 13. Equivalent hydraulic diameter, Manual control of valves between the sytem and the receiver inch......... 0 125 and accumulator tanks allows the selection of any desired pressure level. (1 25 B.t.u. per lb. for helium compared with 0.24 B.t.u. per lb. An overspeed governor is provided on both the high for air). However, the cycle analysis which follows will show pressure compressor/turbine set and the power turbine. The that the compressor temperature ratio required for maximum governor on the compressor/turbine set is a top speed governor cycle efficiency decreases with increasing recuperator effective- only, tripping a compressor bypass valve when this set exceeds ness and this fact is made use of in the design of closed-cycle a predetermined speed limit. The governor on the power helium plant by trading static heat transfer surface for stages turbine is designed to come into play only in the event of of turbo machinery. The inert helium also removes the prob- emergency, the presence of which makes it necessary to shut lem of chemical attack on the power plant components. down the plant. In action, the power turbine governor opens In common with most desirable commodities, the use of the power turbine bypass valve, immediately reducing the helium has one major snag-its cost. To minimize this expense, helium flow through the power turbine. Since this reduces a leak-proof system will be required. the back pressure on the compressor drive turbine, it tends to Several excellent articles and papers (15, 16, 17 and 18) overspeed, thus actuating the compressor bypass valve. Further, L.P.turbine H.P. turbine L.P.compressor H.Pcompressor H-...-. i:'. By courtesy of the American Turbine Corporation FIG. 16-Turbine section 10

Nuclear Power for Commercial Vessels -b~Ja~K -..UP' ~~1~~ —-------— 4 I ~"~ FIG. 17-General arrangement of turbine -~ 2'5-0> Starting Compressors Turbines motor ~~11~1 111 0 1 n To generator Accessory /' -t-~ I. - gear /I.1itil~ IPrecooler/ l, LRecuperator To reactor!ntercooler From reactor By courtesy of the American Turbine Corporation the power turbine overspeed governor trips the system pressure ing this would be to withdraw a small stream of helium from regulator, resulting in the discharge of the contents of the the cold end of the compressor intercooler and pass it through system to the receiver. Simultaneously, the control rods are a heat exchanger in which it would be cooled to whatever dropped into the reactor, reducing the heat input to the system. temperature would be necessary to reduce the xenon content to When the power plant load is eliminated and the reactor a permissible level. Since the xenon is present in such small activity level reduced, a means must be provided to cool the amounts, even its complete removal would leave the helium reactor for a period after shutdown. During both the normal essentially undiminished in quantity. This cold helium stream procedure of shutting off the plant and the emergency condition would pass through a turbo expander wherein its pressure previously discussed, the compressor/turbine set will circulate would be dropped to essentially the suction pressure of the helium through the reactor until the minimum self-running speed is reached. At that time, a secondary, motor-driven circulating compressor with an auxiliary cooling loop is ener- Helium free of volatile gized, circulating helium through the reactor until activity is fistioe Products so first reduced to a point resulting in a safe temperature level. Xenon Removal The helium used in this plant is available commercially at a purity of 99-99 per cent. Impurities consist of argon, - - carbon dioxide and nitrogen, none of which is in sufficient ^^ ~ quantity to be of concern. There is, however, the possibility.enoerato _ of contamination of the system by gaseous fission products Xenon sts escaping from the reactor fuel elements. The principal volatile Xen ntus Helium contatning volatile radioactive impurity of the fission process is xenon and it is Expansion fission products from desirable that this be removed to prevent even a small build-up turbine compressor intercooler of radioactivity of the working fluid. By courtesy of the American Turbine Corporation The xenon can be effectively removed to any degree desired by solidification in a cold trap. One procedure for accomplish- FIG. -Xenon rap 11

Nuclear Power for Commercial Vessels compressor. In passing through the expander, the helium The design conditions selected are as follows:would be cooled sufficiently so that it could act as the refrigerant Total compression ratio 2-4:1 for cooling the xenon cold trap exchanger. A typical arrange- Pressure losses, per cent:-Intercooler 0-75 ment of this type of trap is shown in Fig. 18. H.P. recuperator 1-50 Since the gas flow required to hold down the xenon Reactor 1-50 concentration in the working fluid of the power cycle is only of L.P. recuperator 2-25 the order of 1 per cent of the mass flow, the dimensions of the Precooler 1.00 cold trap heat exchanger would be small, as would the turbo expander required to provide the cold end drop in temperature. Total, per cent 7-00 The passages of the cold trap would gradually become plugged Expansion ratio: with xenon and its decay products until eventually it would 2-4 (1-007) 2-233:1 require replacement by a new unit. The size of this trap Compressor inlet temperature:would be such that it could be cleaned or disposed of, depending based on 75 deg. F. sea temperature, deg. F. 90 upon which would seem to be desirable in the final design. Turbine inlet temperature, deg. F. 1,400 Choice of the Cycle Details Recuperator effectiveness, per cent 92.3 In order for any closed cycle nuclear power plant to be Mechanical and other losses, per cent 5-0 attractive economically, it must be a high temperature machine, Physical Constants for Helium i.e. it must operate at cycle temperatures in excess of 1,200 Specific heat at constant pressure, deg. F. All experience to date with closed-cycle power plants B.t.u. per lb. Cp 1-25 has been at a cycle temperature of 1,250 to 1,300 deg. F., as Ratio of specific heats 8 = 1-658 dictated by the limiting tube wall temperature in a fired air Gas constant R = 386-2 heater. In a nuclear plant this restriction is removed and the (8-1)/8 = 0-398 turbine inlet temperature is only limited, within reason, by Analysis of Cycle (Fig 21) reactor outlet temperature. However, a plant of conservative Compesso (e l e i ec ge design would limit such temperature to 1,500 deg. F. In Absolute temperature at inlet sa 550~R. establishing a cycle for the helium plant, the values of 1,300, Compressor ratio per stage =t2 4 =RT 1 55R:1 1,350 and 1,400 deg. F. cycle temperature were assessed against R(8 -1)/8 = 1191 a 1,300 deg. F. cycle temperature with 100 and 200 deg. F. 1 —Rc(8-/8) = 0191 ~~~~~~reheat. ~Adiabatic temperature rise ATad = 105~F. Adiabatic efficiency qcom. = 0-88 44 _ _ _ Actual temperature rise AT = 120~F. 2-Stages Absolute temperature at outlet T2 = 670~R. intercooling T3 = 550R. 42o30degj 2 4 6700 egRHR. 4,_-, _- O^ei - /% -0L -Total temperature rise ___, ^ 1400deg. _ (compression work) =2 x 120= ATcom. = 240~F. I-Stage I Zl 1140 I intercooling H.P. Turbine /7-0 -7X3sode Absolute temperature at inlet T6 = 1,860~R. i-^.,~~ ATcom 240~F. T7 1,620~R. -38 t I 0 ~eg --- | ~ ~ ~ Adiabatic efficiency exp. 0-888 /1300 deg+100__degRH Adiabatic temperature drop ATad = 270~F. /6't ll3O~04j> T71 1,5900R. 36- ~ ~ ~ ~ ~ T~~ ~ ~6 _ 117 T7i - 54 t ~ I 8 | - i = 2 -5 1 2.0 2.5 Rc 30 2.0 25 Rc 3 0 Expansion ratio Re HP = (1.17)2.51 = 1.48:1 Reheat and non reheat One and two stage intercooling L.P. Turbine By courtesy of the American Turbine Corporation T e nin ri e Total expansion ratio Re TOT= 2-233 FIG. 19 (left)-Comparison of reheat and non-reheat cycles 2-233/1-48 = Re LP = 1-51:1 FIG. 20 (right)-Comparison of single and two-stage inter- (ReLP)-398 = 1178 cooling T7 = 1,620~R. T7/1-178= T81 = 1,374~R. Adiabatic temperature drop ATad = 246~F. A comparison of cycle efficiencies on the basis of pressure Adiabatic efficiency nexp = 0-888 ratio is shown in Fig. 19. These were computed using Actual temperature drop reasonable polytropic (stage) efficiencies for the turbo machinery (output work) ATw = 218~F. and taking pressure losses for the two types of system into account. There is little difference in efficiency between a T8 = 1,402~R 1,400 deg. F. non-reheat plant and a 1,300/200 deg. F. reheat Recuperator plant. The 1,400 deg F. non-reheat cycle was chosen since it T = 1,402~R. shows an optimum efficiency at a lower pressure ratio than the T4 = 670~R. reheat cycle and also does not involve the use of a reheat Available temperature range t = 732~F. exchanger. Effectiveness, per cent r = 92-3 Appreciable gains in efficiency in a closed-cycle power Increase in temperature in plant are effected by moderate intercooling. The value of recuperator AT = 675~F. single versus two-stage intercooling was assessed and plotted Outlet temperature 670+676= T5 - 1,346~F. in Fig. 20. The dual stage intercooling provides only an in- Reactor T6 = 1,860~R. crease in efficiency from 42 per cent to 42.5 per cent, while T5 = 1,346~R. requiring an addition in pressure ratio to achieve this optimum Increase in temperature from 2-4:1 to 2 8:1. Thus, the single stage intercooling was (heat supplied) ATR = 514~F. selected. 12

Nuclear Power for Commercial Vessels Precooler The assumption of a sea temperature of 75 deg. F. T4 = 670~R. will certainly be on the high side for the majority of steaming Loss in recuperator = 732-676 = 56~F. time and, when this is so, an improvement in cycle efficiency T9 = 726~R. can be expected. T1 = 550~R. The effect produced on work rate and cycle efficiency with Temperature drop variation of overall pressure loss and sea water temperature is (to circulating water) At = 176~F. plotted in Fig. 22. Cycle Efficiency Output Tw = 218~F. Input TR = 514~F. ECONOMIC ASPECTS OF THE USE OF NUCLEAR FUEL -cycle = 42-4% A power of 15,000 s.h.p. has been mentioned earlier as a Work Rate - minimum to fully exploit the advantages of the use of nuclear 2,544 fuel. Crever and Trocki (13) make two significant comments on Tw x 1.25 W = 9-35 lb. this consideration: (a) " As the amount of shielding is practically per h.p. hr. independent of power output, a nuclear power plant of low power To assume a value of 7 per cent for the overall pressure will be penalized excessively with respect to its power output ". loss in the cycle may be regarded as optimistic. It is believed (b) "Power plants for propulsion of larger ocean going vessels that this value can be attained with careful design and without (of the order of 10,000 h.p. and above) are of sufficiently large power output to fall within the favourable range for a nuclear'soo w power plant of current design ". /800 To illustrate the finance involved in powering a vessel today, estimates of propulsion machinery derived from the costs k/,I for five ships are given in Fig. 23. The vessels represent 1600 / 7 1000 /' a x -/ 0/2 -.600- 1/ V __3____* *1i________________________ FIG. 23-Graph of cost estimates for steam turbine Entropy machinery FIG. 21-Proposed cycle for helium plant current design of cargo ship, tanker and passenger liner, but excessively large heat transfer surface. Referred to an ar this variation does not impair the comparison of the machinery cycle plant, this is the equivalent of a total overall pressure loss involved as they are all fitted with geared steam turbines. Costs of 11 per cent assigning 5. 5 per cent total pressure loss to the include boilers, turbines, shaft and propeller, but do not allow reactor, which would be the equivalent pressure loss in a fired for cargo handling machinery, steering gear, etc. In the power heater. range pertinent to this discussion, the cost of the steam gener2 __ ators and auxiliary equipment is of the order of 20 per cent of ~ ^^ the machinery costs indicated. -c ^^.? High-powered Cargo Ship ______ An economic analysis has been made of the application A[ / 0 /-^ I _"^ of nuclear power to the " Mariner" class ships (20). These _ g,t^ ^_ ^ ~ ~- ~cargo vessels have a displacement of 21,000 tons and develop ~._-"" 17,500 s.h.p. to give a cruising speed of 20 knots. The maxie 8 - ~ mum output is 19,250 s.h.p. and is, therefore, within the range considered feasible for nuclear powering. The design and 7-, l operation of these ships is described in references 21 and 22 -C- _______ _respectively. I's, ~^^^^The design study made in considering the use of nuclear t47- ~ - power in a vessel of this class has not been published. It is j' ^^^ ^^^^- ^^^^ presumed that a sodium loop reactor would be used to provide F-42- ~: steam of sufficiently high superheat. Fig. 24 shows the model 840 i^^^^ ______ ______of this projected ship, the Atomic Mariner.:~98' The conclusion reached from the economic analysis of a t3__________ ____ 8 - _nuclear-powered Mariner was that it could not compete with the w^ ~ ~ ~ ~'" conventional power plants at present, but that the advance in W __ ___ __ __ 8~~~~~.. reactor technology would improve the competitive position of Sea tertemperature, 0To- deg. F. this new power source. One factor which will adversely ~5,Q~ 15~ ~ affect the suitability of this class of vessel for nuclear propulsion System pressure loss, -percent is its relatively low" load factor ", determined from the number FIG. 22-Correction for off-design conditions of days at sea and the power developed. The Harvard Report 13

Nuclear Power for Commercial Vessels By courtesy of the Newport News Shipbuilding and Drydock Company FIG. 24-Model of proposed nuclear powered Mariner (23) assumed 170 days at sea at 10,000 s.h.p. as typical for these of the ship in ballast would be 4 per cent higher than the loaded ships. This gives a load factor of:- service speed. 10,0002 x 17 266 per cent A comparison will be made between this ship and an equal 17,500 365 sized vessel powered by a helium cooled reactor and a closedThe deficiency applies in some degree to all types of general cycle gas turbine. cargo carriers. Bulk Cargo Carriers I A survey of the most desirable conditions under which to I operate a nuclear-powered vessel gives a good indication of the 8 type of vessel most likely to benefit from its adoption. As a first requirement, a high powered installation running on a long / haul fully exploits the saving in oil fuel. Intermittent operation of a nuclear reactor is a wasteful procedure, as, at reduced loads, /6 it is probable that arrangements must be made to " dump " the temporarily unused heat. Even on shutdowns, the reactor output can only be gradually reduced to prevent overheating. Thus, in both cases, a waste of valuable fissile fuel can occur. / _ To minimize this loss, berthing and cargo handling time must ^ b/ be reduced. Another consideration is the special terminal i facilities necessary to handle radioactive material. A shuttle a / service with fixed terminal ports would thus be desirable. Bulk ~ /2 I cargoes such as ore, grain, or oil are therefore indicated and the' latter appears to be preferred, particularly as the offshore loading and discharge of oil cargoes is now an accomplished __ fact. This is an additional advantage both in reducing i / / manoeuvring time and in providing a safety measure by isolation. - / The choice of an oil tanker is not a paradox as it is incon- ceivable that the use of nuclear energy will reduce the demand 8- / for oil within a period of time equivalent to the combined lives of several ships. / An excellent review of modern tanker and ore carrier design practice is given in references 24 and 25. The specific vessel 6 / selected as suitable for analysis in this paper is described in references 26 and 27. To meet the limitations of available dry docks and permit passage through the Suez Canal, the ___// principal dimensions are: — 7'4-_ ~ ~ Length overall, ft. 660 /Breadth, ft. 85 Loaded draught, ft. 34 2 / _ The deadweight capacity loaded is 30,000 tons and the 8 10 12 14 16 18 20 model test speed-power curve is reproduced in Fig. 25. The speed (knots) ballast condition curve was estimated from the model test data FIG. 25-Speed/power curves, loaded and ballast given, using the assumption made in reference 24 that the speed condition 14

Nuclear Power for Commercial Vessels For the purpose of this analysis, a typical voyage from a Overall Propulsion Plant Conditions and Weight Flow North European port to either the Burma or Borneo oilfields Net output (at shaft), s.h.p. 16,500 will be considered, a steaming distance of, say, 10,000 nautical Mechanical and other losses, per cent 5 miles each way. Gross output, h.p. 17,350 Reduced power operation during the voyage will be assumed Helium flow, lb. per hr. 17,350 x 9 35 = 162,000 as follows: Reactor load, B.t.u. per hr. (a) Suez Canal passage plus berthing at both ends, equivalent 162,000 x514 x 1-25 = 104 106 to 24 hours at 50 per cent service power. Overall propulsion plant efficiency, per cent (b) Loading and discharging equivalent to 24 hours at 15 per 16,500 x 2,544 404 cent service power. The three cargo pumps fitted are 104 x 106 actually each powered by 500 h.p. motors and are capable A 10 per cent addition to the turbine power output will of discharging the rated cargo capacity in 12 hours. be used to cover the engine room auxiliaries. The latest published performance figures for this tanker(27) Estimate of U.235 "Burn Up" during Voyage averaged over the outward and homeward passages of eight Grams voyages, show a speed of 18-2 knots with a fuel consumption At full speed 16,500 x 1 x 46 x 24 1927 of 93-6 tons per 24-hr. day. This is higher than the predicted 0-404 x 25,750 speed at the rated power of 16,500. At reduced speed 8,250 x 1 1 x 24 _ 0.404 x 25,750 Estimate of Fuel Oil Consumption for Round Voyage 2,480 x 24 Tons In port 0404x256750 0.404 x 25,750 At full speed 20 x 93...... 4,280 81 50x 2 x 24 Total 1,954 At reduced speed 8250 x 06 x 24 T 2,240... This is the weight of fuel which is actually destroyed in In port 2,480 x0 6 x24 16 producing the power for the voyage, but it only represents a 2,240..... fraction of the total fuel with which the reactor must be charged. A reasonable "burn up" percentage for the fuel for TOTAL.................. 4,350 the heterogeneous gas cooled reactor, would be 25 per cent. Following this burn-up the fuel elements would require Time for Round Voyage chemical processing, as described previously. The capital Days outlay for the plant, therefore, must include the cost of some At full s d 20,000 46 5,862 grams of U.235 carried as dormant fuel per voyage. A132 x 24. a * a 46 A detailed estimate of the first cost of the gas-cooled reactor Reduced speed and in port......... 2 is outside the scope of this paper, but, using the limited information available, a figure of ~1 million, or ten times the equivalent TOTAL................ 48 steam plant, agrees with majority opinion. -___ If the price of U.235 is ~X per gram and using the same The comparison may be simplified and still remain within fixed charges on investment the "steaming cost" for the voyage the limits of accuracy allowed by other necessary assumptions, then becomes: if the cost of the turbines and transmission is considered to Fixed charges = ~ be the same in each case. The comparison then becomes one 48 x 1 1,000,000+5,862X) =14,480+ 85X between the "steaming cost" of the orthodox ship and its 365x 100 / equivalent with nuclear powering. Fixed charges on the Fuel at ~X per gram = 1,954X invested capital will be included. From Fig. 23, the cost estimate for the machinery of a 16,500 s.h.p. installation is 14,480 +2,039X ~475,000 of which, say, ~100,000 represents the cost of steam generators and ancillary equipment. To "break even" with the equivalent orthodox steam MacMillan and Ireland(28) use the following make-up for plant, the cost of U.235 must then be: the fixed charges on this type of investment: 1,900-14,480 er r X = 31,90014,48 say, ~8 10s. 0d. per gram. Per cent 2,039 per annum At the international conference on the peaceful uses of Interest 2-6 atomic energy held in Geneva in August 1955, the price of Depreciation 4-9 uranium was quoted at $25.00 per gram of U.235. It was not Insurance 2-0 stated whether or not this figure included an allowance for Maintenance 1-5 fuel element fabrication. Even allowing for some error in this figure, with the present rate of exchange at $2.80=~1, then 11.0 it would appear that a balance in "steaming costs" for the two plants could very nearly be made. Using these figures, the "steaming cost" for a round Additional Considerations in the Comparison voyage would be: The bunker fuel capacity of this 30,000-ton d.w. tanker Fixed charges = 48 x 1- x ~100,000 = ~1,450 is 4,500 tons and would be filled at the port of loading. Using 365 100 a nuclear-powered plant would increase the machinery weight Oil at 140/- per ton ~30,450 by some 500 tons, giving a net increase in cargo capacity of 4,000 tons. Total ~31,900 The crew wages of all classes of vessels are now a major item in operational expense(22). On completion of the initial In calculating the fissile fuel consumption of the nuclear voyage, it appears unlikely that any additional operating plant, the heating value of 1 gram of U.235 is taken as personnel will be required. For instance, the advances in 65.5 x 106 B.t.u. or equivalent to 25,750 horsepower hours. reactor control technology have proved that such reactor The cycle developed in the foregoing section will be used controls are complex in design, but simple in operation. and the cycle efficiency of 42 4 per cent should not vary appreci- No account is taken of the additional investment to cover ably over the whole range of powers, this being a characteristic the charge of helium, but it is not expected that this would of the closed cycle plant. unduly affect the comparison. 15

Nuclear Power for Commercial Vessels..............::.:.....:........................~........... By courtesy of the U.S. Maritime Administration FIG. 26-Concept of reactor installed in a tanker The cost of the turbo machinery has been considered helium machinery since more stages will be required. Also, equivalent to that of the orthodox steam plant and a more special arrangements will be required to prevent leakages not detailed analysis would show a definite increase in cost for the only at glands and other openings in the casings, but also due Water level l 7 Main eck / Port 0 / Starboard wing tank wig tank, C.V. K. t j I By courtesy of the U.S. Maritime Administration FIG. 27-Dockside handling arrangement for radioactive material (1) Manipulators; (2) Reactor cover and control rods; (3) Dockside crane, (4) Catwalk; (5) Shield hatch; (6) Cask; (7) Watertight deck; (8) Shield; (9) Core; (10) Pressure vessel; (11) Reactor supports; (12) Reactor foundation 16

Nuclear Power for Commercial Vessels to porous castings. This latter problem is far more acute when Proposed Machinery Arrangement for the 30,000 Ton D. W. dealing with helium under pressure than when dealing with Tanker Powered by a Closed Cycle Helium Turbine with a steam. Helium Cooled Reactor The advantage of the nuclear power plant increases with increasing power level, therefore the decision to limit the The first consideration is the selection of a suitable drive power of the vessel selected to 16,500 s.h.p. may appear to be and the choice is limited by the desire to keep the turbine dequestionable. This was made to introduce an element of sign as simple as possible which calls for unidirectional rotaconservatism into an otherwise "pioneer" plant. The pro- tion. The transmission then must include either a reversing pulsive advantage of a single screw installation is thereby gear a controllable pitch propeller or an electric drive. It maintained in a well tried power range. should be remembered that one of the reasons for selection of Alternative Proposals for Tanker Propulsion this type of vessel as most suitable for nuclear powering was To compare propulsion plants, Shoupp and Witzke(29) the minimum of maneuvering required under usual service selected a tanker of 20,000 normal s.h.p. with a service speed conditions. of 18 knots. This ship would have a cargo capacity of 35,000 A reversing gear to transmit 16,500 S.H.P. appears to be tons and make eight voyages of 17,000 miles per year, giving too far ahead of current development to warrant serious conan annual total for cargo handled of 280,000 tons. Using a sideration. Unfortunately the same remark appears to be true specific fuel consumption of 0-5231b. per s.h.p. hr. for all purposes this gives an annual fuel consumption of 232,000 bbls for the controllable-pitch propeller. The author considers that poses, this gives an annual fuel consumption of 232,000 bbls.,. t which at $2.00 per barrel would cost $1 65 per ton of cargo ths is the most promising line of development for the future carried. Additional ship operating costs, including capital Correspondence with a leading manufacturer has ascertained charges, overhead, port dues, maintenance and supplies, wages that 7,000 S.H.P. is the present maximum in satisfactory and subsistence, bring the total ship operating cost to $7-15 service. Baker in his discussion on Ref. 19 recalls an Ameriper ton of cargo. can vessel on which an attempt to transmit 14,000 S.H.P. Details of the type of nuclear reactor proposed for this failed. McMullen (31) records that the open cycle gas turbine ship were not available at the time of writing, but mention is plant of 6,000 S.H.P., at present being installed in the Liberty made in the paper of the use of steam as the thermodynamic ship "John Sergeant," will use a controllable pitched propelfluid. Also the cost of the boiler and the boiler auxiliary equip- ler. Mention is made in this paper that if successful, gas turment was estimated at $570,000 from the "Mariner" class bines will be adopted for selected applications in the 7,500 to estimates. It was assumed that the cost of the turbines, con- 15,000 S.H.P. range and it would be interesting to hear whether denser, shaft, propeller, etc., was not changed when the con- the same type of drive would be proposed. ventional oil fired boiler was replaced by a nuclear reactor. The safe selection therefore must accept the additional Analogy with land plant data was necessary due to security weight, space and expense inherent in the electric drive. restrictions on much of the information that would be more The auxiliary machinery will all be electric motor driven. directly applicable to this field and, on the above basis, it was Power demand for port use should be of the order of 2,000 concluded that the cost of natural uranium was not likely to H.P. and the convenient possibility then presents itself of using compete with conventional fuel on a straight economic basis. the main propulsion turbine to produce this power. This To compare the two types of power source in more detail: computing the economic comparison with scheme was used in computing the economic comparison with For equal costs: orthodox steam power plant. Allowing for the inefficient run(a) With zero nuclear fuel cost, the maximum permissible nin of i i i o o price for the reactor plant would be $4,800,000 nn g of the plant at about 12% of designed rating this would not (b) Assuming zero investment in the plant, the nuclear be a significant loss since the port time/sea time ratio is so fuel price cannot exceed $27.80 per gram. small. However the scheme demands continuous operation of This paper agrees that a saving will be made in the com- the main plant and hence the placing of too many eggs in the bined weight of plant and fuel and suggests that every 1,000 single basket. tons of additional cargo capacity can pay for an additional plant Until a sufficient degree of reliability has been proven, it investment of $500,000. is recommended that a separate diesel powered plant be inPerhaps a more significant comment for immediate interest stalled say of four high speed units giving a total output of is that these figures, and the conclusion drawn from them, 2,000 H.P. This would provide in port power only, the sea load apply to an American-operated vessel. The equivalent being tapped from the propulsion power. Detailed discussion European ship would have a considerably reduced ship operating of the possible electrical systems to be used is outside the cost, many items of which would be at least halved. This scope of this paper and can be found in Ref. (32). The diesel means that the cost of the machinery and fuel forms a larger plant would be a sound investment during the early years of part of the European ship operating cost. Hence the balance the ship's operation as the carriage of some 100 tons of diesel would be more in favour of the nuclear power plant if compared fuel could ensure a means of making port up to a distance of with the oil fired plant when using figures applicable to a 1500 miles in the event of a complete failure of the nuclear European ship. powered plant. Another proposal is made by the Engineering Research p plant. Another proposal is made by the Engineering Research Ihi laying out the scheme shown in Figs. 28, 29 and 30, the Institute of the University of Michigan. This is included in a for guidance on the space recently completed feasibility and preliminary design study for abe fr te ainer. e diese anti a nuclear power plant suitable for a large ocean-going tanker(30) avalable for the machinery. The diesel plantis accommodated An artist's impression of this installation is shown in Fig. 26. n the ornal boler room, the reactor is housed in what was The conclusion drawn from stuy is tt se the aftermost centre cargo tank and the main pump room lies The conclusion drawn from this study is that safe operation d bl can be expected, but that there will be no saving in operational drectly below the reactor. Direct access from deck to pump cost compared with an oil fired installation. This latter factor room is afforded both on the port and starboard sides by the was rather to be expected since the original specification for trunk/cofferdam immediately forward of the engine room this project called for a tried type of reactor. This, of course, bulkhead. By eliminating the original fuel settling tanks the considerably reduced the field of choice. A pressurized water total machinery space is increased only slightly. reactor was' selected and one of its inherent advantages is The main turbine/compressor has been shown as a single illustrated in Fig. 27, which shows the convenient arrangement in-line unit. If conservation of engine room length is considof providing a transparent shield by flooding the access hatch ered to be of sufficient importance, then either or both of two above the reactor. The loading of fissile fuel and discharging modifications could be effected: of fission products is thus simplified by the direct observation (a) The main turbo alternator could be replaced by two or afforded. even three smaller units. 17

Nuclear Power for Commercial Vessels ~I \~ ~ ENGINE ROOM ACCESS DIESEL / \ SERVICE TANKS n n REACTOR ACCESS. \ ACCOMODATION OR STORES STEERING GEAR i I REACTOR 11 AFT RECESS I —-l --- -: —I - ROOM I low ==r —-' F4'. ~ I — I \CARGO PUMPS 0 10 20 30 40 P I I I I I 4 FIG 28. PROFILE OF PROPOSED TANKER SCALE IN FEET PUMP ROOM ACCESS STARTING AIR - AND WING A -G STORE CARGOE io~ ~SCALEI MAINN EET FIG29. -PLAN AT OPERATING LEVEL PURIFIERS 0 ALTERNR REACTI / S P A R Ez TURBINE CONTROLS I^^^ ^^ O CARGO 0 ff i TANKe HELIUM PUMPS AND STORAGE PUMP ROOM ACCESS 0I~O~ 100 20 30 40 3C~~~~~~~~~SCALE IN FEET FIGFIG.29. PLAN AT OPERATING LEVEL PRECOOLER & INTERCOOLERS 7 CIRC'G PUMPS ALTERNR SEATING WING " —SPARE TAILSHAFT \ CARGO;JIN THRUST 1 V ~pu F~1-1 M- ~0 \CARGO 0E' HPUREMPS \O ~ ^ "^90 ~0 0TURBINE SEATING 0 — C FIRCN. PLAN AT OPERA LEVEL

Nuclear Power for Commercial Vessels (b) The in-line unit could be replaced by a co-axial ar- REFERENCES rangement in which the alternator and low pressure 1. COCKCROFT, J. 1953. "Atomic Propulsion with Special turbine is on one shaft and the two compressors and Reference to Marine Propulsion". Trans. I.Mar.E., the high pressure turbine on the other. Vol. 65, p. 105. For future vessels in which it may be considered the 2. FEARNSIDE, JONES and SHAW. 1951. "Applied Atomic diesel-alternators can be dispensed with, the space so vacated Energy", Temple Press, London. could house the reactor thus increasing cargo space. 3. MASSEY. 1953. "Atoms and Energy", Elek Books, London. From provisional estimates of stability all the above ar- 4. STEPHENSON. 1954. "Introduction to Nuclear Engineerrangements would be acceptable to the naval architect. ing", McGraw Hill. A funnel is shown in chain dots to indicate that its inclu- 5. LEESER. 1955. "Evaluating Basic Materials for Nuclear sion depends upon factors outside direct machinery require- Power Plants", A.S.M.E., Paper 55-S14. ments. Dispersion of the diesel engine exhaust and the display 6. STONE. 1955. "Shielding Concepts for Nuclear Reactors", of owners insignia are two of these which could be found alter- A.S.M.E., Paper 55-S20. native locations. The author suggests however that a ship 7. RODDIS and SIMPSON. 1954. "The Nuclear Propulsion without a funnel would resemble a Manx cat. What better mark Plant of the U.S.S. Nautilus, S.S.N.-571", Trans. of esteem could be given to the individual who will sail in S.N.A.M.E., Vol. 64. charge of this unique power plant, then to use the funnel as 8. GALL. 1955. "The Homogeneous Reactor Experiment", housing for a luxurious suite of rooms for "The Chief"? A.S.M.E., Paper No. 55-S15. 9. MCCHESNEY and SHANNON. 1955. "The Use of Benzine as a Thermodynamic Working Fluid for a Nuclear Power Plant", A.S.M.E., Paper 55-S29. CONCLUSION 10 CLARK. 1953. "Mechanical Pumps for High Temperature Following a study of the information now available on the " Mechanicl various types of nuclear reactors and power plants, the author Lqud Metal", A.S.M.E., Mechanical Engineering", August 1953. selected the helium-cooled reactor with a closed-cycle gas 11. W LSEY. 1955. "Selection of Materials for a Sodiumturbine power plant as the most attractive for commercial Gap Ra Sstem A Paper marine use. The economic comparison shows that this plant.. "es d Oper n of a N er can compete with the oil fired steam turbine plant for high Power St, Tra..ar.. ol p. 1 Power Station", Trans. I.Mar.E., Vol 67, p. 179. powered ships once the design values used in the paper have 13. CREER and TROCKI. 1954. "Nuclear Power Plants for been verified in practice and the first cost of the plant proven Conferee Pap, I.E.. General to be of the order suggested. The latter will only be so when M i ng, 7h Jn,... the components of the plant are in normal production by the "Prect of the Atomic Power Developequipment suppliers. "Project Assofiates, Inc.", A tomerican Power DevelopHere, then, on both counts, the time factor governs the Associates, Inc. American Power onference, date on which nuclear power will be adopted for ship pro- 30th March, 1955. "r pulsion. Whatever type of nuclear plant is selected, a vast "Trial s of an Aerynamic Turbine amount of research and development work must yet be accomp- K EE n S 1 "le le ANovemr 1.T lished. This will require the combined efforts of an integrated Installations for Marine Propulsion", "The Air Turbine 16. KEInstallations for MarN. 194." los ed Cyce ir Tine and team of engineers, physicists, chemists, metallurgists, mathemati- as rine eeer an anar. cians and probably the representatives of other professions. Turbe", DecemR W 195 3 and January 19G 4. These men are available and the British shipbuilding industry "Gas and Oil Power", February 1955. and their equipment suppliers undoubtedly have the potential to Gas and Oil Pwer", February 1 tackle the many problems peculiar to this new source of power. 18 MCMULLEN 195. "A Gas urbine lant with Refer If this paper is instrumental, to whatever small degree, in to a Cargo Vessel", Trans. S.N.A.M.E., Vol. 59, p. 502 fostering the necessary interest in this concept of marine 19. BURRILL. 149. Latest Developments in Reversble power, then the author will be well satisfied. Propellers", Trans. I.Mar.E., Vol. 61, p.l. 20. WITZKE and HAVERSTICK. 1954. "Nuclear Power Plants for Ship Propulsion Application", Electrical Engineering A.I.E.E., February 1955. ACKNOWLEDGEMENTS 21. RuSSo and SULLIVAN. 1953. "Design of the Mariner The author gratefully acknowledges the assistance, in Type Ship", Trans. S.N.A.M.E., Vol. 61, p. 98. supplying information and illustrations, given by the following 22. ALLEN and SULLIVAN. 1954. "Operation in Service of the organizations: Mariner-Type Ship", Trans. S.N.A.M.E., Vol. 62, p.522. "Scientific American", New York. 23. Student Committee on Atomic Energy. 1951. "Swords Standard Oil Company of New Jersey. into Plowshares-a Study of Current and Future DevelopNewport News Shipbuilding and Dry Dock Company. ments of Commercial Atomic Energy", Harvard University Westinghouse Electric Corporation. Graduate School of Business Administration, Cambridge, Detroit Edison Power Company. Mass. J. Stone and Company (Charlton), Ltd. 24. ROBINSON, RoESKEand THAELER. 1948. "Modern Tankers", The Caledon Shipbuilding and Engineering Co., Ltd., Dundee. Trans. S.N.A.M.E., Vol. 56, p. 422. Escher Wyss, Ltd., Zurich. 25. HENRY. 1955. "Modern Ore Carriers", S.N.A.M.E., Engineering Research Institute of University of Michigan. Spring Meeting 1955. United States Department of Commerce, Maritime Administra- 26. GOLDSMITH. 1950. "S.S. Atlantic Seaman", A.S.M.E., tion, Office of Ship Construction and Repair. "Mechanical Engineering", Vol 73, 1951. In particular, the American Turbine Corporation for the 27. BRITTON and TEN BROECK. 1955. "Comparison of Design supply of their recent design study of a 60-MW closed-cycle and Service Performance of Some Modern Tankers", gas-turbine nuclear power plant and other data on which the A.S.M.E., Paper 55-S37. discussion on the closed-cycle gas-turbine plant is based. 28. MACMILLAN and IRELAND. 1948. "Economic Selection of A personal word of thanks is offered to Dr. Tom Sawyer Steam Conditions for Merchant Ships", Trans. S.N.A.M.E., and Messrs. Harold Ohlgren and Marx Weech for their helpful Vol. 56, p. 149. criticism and to the author's colleagues in the Naval Architecture 29. SHOUPP and WITZKE. 1955. "Nuclear Power Plants for and Marine Engineering Department of the University of Marine Service", Atomic Industrial Forum, San Francisco, Michigan for their encouragement. 4th April, 1955. 19

Nuclear Power for Commercial Vessels 30. FOLSOM, OHLGREN, LEWIS and WEECH. 1955. "Nuclear Propulsion of Merchant Ships-an Engineering Summary", Engineering Research Institute, Ann Arbor, June 1955. 31. McMullen, 1955, "The Gas Turbine Installation in Liberty Ship "John Sergeant." Paper No. 4 SNAME Annual Meeting, New York, November, 1955. 32. Linsell, 1948, "Electric Power for Ships." Trans. I. Mar. E., Vol. 60, pp. 167-170. 20