possible but yet unproved method of increasing power and aiding in reactor control. Operation with heavy water will increase the conversion ratio of boiling reactor systems-that is the percentage of fuel generated to fuel burned-but heavy water is expensive and its leakage must be carefully controlled. Superheating the steam by nuclear heat may make it possible to obtain steam temperatures approaching those prevailing in modern steam plants. Experiments on these and other improvements are planned.

Work was discontinued on the Argonne Boiling Reactor (ARBOR), which had been planned as a flexible boiling water test facility. Based on on further studies by the Argonne National Laboratory, including preliminary design by United Engineers and Constructors, it was determined that the flexibile facility required to accomplish all the objectives of the project would cost considerably more than was first estimated.

Anticipated results did not appear to warrant the expenditure in view of the schedule delays and the competition of other projects for funds.

Homogeneous reactors. Following dismantling of the Homogeneous Reactor Experiment No. 1 (HRE-1) in 1954, the Homogeneous Reactor Experiment No. 2 (HRE-2) was undertaken on the same reactor site at Oak Ridge National Laboratory. The HRE-2 is a two-region forced circulation aqueous homogeneous reactor: a central region consisting of a zirconium tank containing enriched uranium fuel in the form of uranyl sulfate dissolved in heavy water, and a surrounding region containing heavy water and acting as a reflector. HRE-2 is designed for a thermal output of 50 megawatts. A small amount of heat produced is supplied to the same turbogenerator used with HRE-1 to produce 300 gross kilowatts of electrical power. The remaining heat is dumped to air-cooled condensers.

The use of fluid fuel permits continuous removal of fission products and plutonium, avoids the expense of fabricating fuel elements, and leads to desirable control characteristics. Fluid fuel reactors, now being developed, employ fuel in the form of a solution or slurry.

The HRE-2 achieved criticality December 27, 1957. After critical and low power-level experiments, the reactor was gradually brought up to its full design power of 5,000 thermal kilowatts, which was attained on April 4. Shortly after the reactor reached full power, instruments indicated that some fuel from the main core region was transferring to the heavy water reflector region. The reactor was thereupon shut down. Before the shutdown occurred, the plant had produced about 326 megawatt-hours of nuclear heat and had operated

in a stable manner at design conditions. At no time during the entire operation did any radioactivity escape from the reactor system. Examination after the shutdown determined that an internal leak in the zircaloy core vessel was responsible for the transfer of fuel from the core to the reflector region. Until the cause of the core vessel failure has been determined, and a decision made on whether or not to replace the core vessel assembly, the HRE-2 will be operated as a singleregion reactor.

The long-term objective of the program for development of aqueous homogeneous reactors continues to be the development of a thermal breeder reactor system for the thorium-uranium 233 cycle.

Oak Ridge National Laboratory also is exploring the technical and economic feasibility of using molten salts of uranium fluoride and thorium fluoride in homogeneous reactors. Molten salt systems could operate at high temperature and low pressures. Results in out-ofpile loop experiments have demonstrated the exceptionally high resistance of the container materials to corrosion and to extended mass transfer.

A smaller development program on homogeneous reactors utilizing uranyl phosphate fuel is being completed at Los Alamos Scientific Laboratory. The first small experimental reactor in this program, Los Alamos Power Reactor Experiment No. 1 (LAPRE-1), was abandoned because of excessive corrosion. The second reactor, Los Alamos Power Reactor Experiment No. 2 (LAPRE-2), will be assembled and operational tests will be conducted before the program is terminated.

Fast reactors. Development of fast breeder reactors has been pressed from the start of the civilian power reactor program because of their ability to breed more fissionable material than they consume in the uranium-plutonium cycle. Reactors of this type operate with neutrons at high velocities instead of using moderators to slow down neutrons to thermal velocities. The United Kingdom also has an extensive fast reactor program and there has been a free exchange of information in this field for the past 3 years.

One of the most important experimental facilities in the fast reactor program is the Experimental Breeder Reactor No. 1 (EBR-1) at the National Reactor Testing Station, which produced the world's first electric power from atomic energy in December 1951.

Undesirable operating characteristics were observed during 1955 after a second core was installed, and in experiments investigating these characteristics, a partial meltdown of the core occurred. 21 Studies indicated that the cause was "bowing" or distortion of the

21 See pp. 45-46, Twentieth Semiannual Report to Congress (January-June 1956).

21

core components because of temperature differences within the core. A third core was designed to control this "bowing" by using tightly clad fuel elements and tightening and supporting devices, both inside and outside the core. Preliminary operation of the EBR-1 with the third core during late 1957 and 1958 indicates that the new core design has eliminated the prompt positive temperature coefficient, that is, the tendency of a small increase in temperature to increase reactivity and thus cause a further increase in power and temperature. The third core also showed no evidence within the design power rating of the reactor of the other undesirable characteristic previously observed oscillations of power without external cause when the reactor reached a certain power level.

Major construction is under way on the Experimental Breeder Reactor No. 2 (EBR-2), a fully integrated experimental power reactor station with its own chemical processing plant. The station will have 16,500 kilowatts net electric output. The EBR-2, being built at the National Reactor Testing Station, is designed with operating characteristics similar to those of a full-scale powerplant. It will, for example, utilize components of such design that they can be scaled up to a size or capacity required for a full scale plant.

Several other major developmental facilities pertaining to fast reactor technology are in operation or are planned. Now that successful operation of the EBR-1 is demonstrating basic stability, the major area of fast reactor safety that remains to be demonstrated is the possible effects of extreme operating conditions. Two facilities are planned which will investigate these effects, namely, the Transient Reactor Test (TREAT) and the small Nevada Fast Reactor Safety Testing Station.

The TREAT reactor, scheduled for operation in early 1959 at the National Reactor Testing Station, will have a special test hole to study dynamics of fuel meltdowns. In this test hole, a sample fuel element can be heated to and beyond melting temperatures without overheating the reactor itself.

The Nevada facility, located at the Nevada Test Site, and scheduled for completion in July 1959, will provide a remote site for test of fast reactor cores under extreme conditions.

The Zero Power Reactor No. 3 (ZPR-3), a fast critical assembly at the National Reactor Testing Station, Idaho, is utilized for fast reactor physics work. The Zero Power Reactor No. 6 (ZPR-6) facility planned for construction at Argonne National Laboratory will permit an extension of these studies.

Sodium Reactor Experiment (SRE). The sodium graphite power reactor program was formally initiated in 1954 with start of construc

tion of the Sodium Reactor Experiment. The sodium-cooled, graphitemoderated SRE began operation in April 1957 at Santa Susana, Calif. Although several problems with components and fuel limited its output to about one-third design power during early tests, operation has been generally satisfactory. During this report period, the SRE reached its design power of 20 thermal megawatts for the first time. Initial experience with reactor control and maintenance has been particularly gratifying. Reactor stability was such that negligible control rod motion was required at constant power. Routine maintenance and modifications of the system have been carried out without difficulty and in a relatively short time. On one occasion, a fuel element, accidently broken during insertion, was removed during a single day with grappling tools devised for this purpose.

Early difficulties with the pumps and cold traps in the coolant system have been largely overcome by minor modifications to the system. Operation of the intermediate heat exchanger and steam generator at low capacity has revealed some difficulties which now are understood and can be eliminated in future designs. Because of potentially high thermal stresses during scrams (instantaneous shutdowns) of the reactor, initial operation was restricted to onethird design power until early 1958 after eddy current brakes had been installed in the sodium circuits to control coolant flow. Subsequent full power operation has been entirely satisfactory in this regard.

Although the fuel elements presently installed are operating satisfactorily, evidence has accumulated that the life of these elements will be much shorter than originally estimated. This is partly because the excess reactivity of the reactor, required to overcome fission poisons buildup, is low and would diminish as poisoning increased. A second factor is that the swelling of metallic uranium at high temperatures, partially a result of fission product gases, may cause failure of fuel elements. A loading of thorium-uranium alloy fuel elements is being fabricated and will be installed during 1958.

The attainment of full power operation of the SRE on May 21 is a noteworthy achievement. Even more significant, however, is the fact that a year of operation, including the difficulties encountered, has pointed the way toward major improvements in design. Planning is under way on modification to demonstrate improved sodium-graphite designs.

Invitations were issued May 8 to industry for design, development, manufacture and test operation of major nonnuclear components for use in all reactor systems which use liquid sodium as coolant. Fifteen proposals were received for development of intermediate heat exchang

ers and steam generators of advanced design and improved performance, and the Commission has selected the Griscom Russell Co., Alco Products, Inc., and Combustion Engineering to carry out preliminary design and analysis. In response to an invitation for proposals for development of flow control mechanisms, 11 proposals were received, and are being evaluated.

Liquid metal fuel reactor. Experimental work and design studies on liquid metal fuel systems were continued by the Brookhaven National Laboratory and The Babcock & Wilcox Co. A conceptual design and preliminary research and development for a large liquid metal fuel reactor were completed with sufficiently promising results to warrant continuation of the development program. The basic facility in the development program will be a 5 thermal megawatt Liquid Metal Fuel Reactor Experiment (LMFRE-1).

The LMFRE-1 will use a fluid fuel of uranium dissolved in molten bismuth in a moderating structure of unclad graphite. The reactor heat will be dissipated by an air-blast heat-exchanger with no generation of electric power. Reactor construction is scheduled to start in 1959 with completion in 1961. Babcock & Wilcox will design, construct, and operate the reactor.

Organic Moderated Reactor Experiment (OMRE). Investigation of organic moderated and cooled reactors was initiated in 1955. The noncorrosive nature of the organic coolant, the low level of radioactivity created in the coolant by neutron bombardment, and the low pressure of the organic coolant at operating temperature, are advantages of this type of plant. However, organic coolant decomposes under radiation, and there was some concern that the decomposed organic material might foul fuel element and heat transfer surfaces as well as lower the heat transfer properties of the organic. In order to determine the rate of organic decomposition and to evaluate possible fouling, construction of the Organic Moderated Reactor Experiment (OMRE) was undertaken at the National Reactor Testing Station.

The OMRE, which began operation late last year, is a relatively simple and inexpensive reactor experiment. During initial operation there has been no evidence of fouling of heat transfer surfaces. Preliminary results of decomposition studies indicate that the cost of replacement of the decomposed organic coolant would not be prohibitive in a power reactor of this type. Extensive experiments are planned with the OMRE to study long-term irradiation effects, and to try to develop improved organic materials.