CHAPTER 7 DESIGN AND CONSTRUCTION OF EXPERIMENTAL HOMOGENEOUS REACTORS* 7—1. INTRODUCTION 7-1.1 Need for reactor construction experience. The power reactor de- velopment program in the United States is characterized by the construe- tion of a series of experimental reactors which, it is hoped, will lead for each reactor type to an economical full-scale power plant. Outstanding examples of this approach are afforded by the pressurized water reactor and boiling water reactor systems. The development of pressurized water reactors started with the Materials Testing Reactor, followed in turn by the Sub- marine Thermal Reactor (Mark 1), the Nautilug Reactor (Mark II), and the Army Package Power Reactor. KExperience obtained from the construce- tion of these reactors was applied to the full-scale plants built by the West- inghouse Electric Company (Shippingport and Yankee Atomic Electrie Plants) and Babeock & Wilcox Company (Consolidated Iidison Plant). Although many have argued that the shortest route to economic power will be achieved by eliminating the intermediate-seale plants, most experts believe that eliminating these plants would be more costly i the long run. To quote from a speech by Dr. A. M. Weinberg [1], while discussing large- seale reactor projects: “The reactor experiment—a relatively small-scale reactor embodying some, but not all, the essential features of a full-scale reactor—has become an accepted developmental device for reactor tech- nology.”’ An alternative to the actual construction of experimental nuclear reactors has been proposed which consists of the development of reactor systems and components in nonnuclear engineering test facilities, zero-power critical experiments, and the testing of fuel elements and coolants in in-pile loops. This approach, although used successfully in the development of various solid-fuel coolant systems, is not completely applicable to circulating-fuel reactors because of the difficulty of simulating actual reactor operating conditions in such experiments. In in-pile loops, for example, the ratio of the volume of the piping system to the volume of the reacting zone is never quite the same as in a reactor, making it impossible to duplicate simul- taneously the conditions of fuel concentration, enrichment, and power density. In cases where these variables are important, the m-pile loops *Prepared by J. A. Lane, with contributions from 8. E. Beall, 5. I. Kaplan, Oak Ridge National Laboratory, and D. B. Hall, Los Alamos Scientific Laboratory. 340 7-2] WATER BOILERS 341 an at best provide information of an exploratory nature which must be verified In an operating fluid-fuel reactor. A second aspect of circulating-fuel reactors, which precludes relying solely on engineering tests and in-pile loops, is the close interrelation of the nuclear behavior and the operational characteristics of the fuel circulation system, which can be determined only through construetion and operation of a reactor. Other aspects of reactor design that can be best determined nm an operating homogeneous reactor are continuous removal of fission products produced in the nuclear reaction and remote decontamination and maintenance of reactor equipment and piping. 7-1.2 Sequence of experimental reactors. It is obvious from the fore- going that the construction of a sequence of experimental reactors has been an important factor in the development of homogeneous reactors. In this sequence, which started with nonpower rescarch reactors, seven such reactors have been built (not mecluding duplicates of the water boilers). These are the Low Power Water Boiler (LOPQO), the High Power Water Boller (HYPO), the Super Power Water Boller (SUPQ), the Homogeneous teactor Iixperiment (ITRIE-1), the Homogencous Reactor Test (HRE-2), and the Los Alamos Power Reactor Experiments (LAPRE-1 and -2). [ the sections of this chapter which follow, these reactors are deseribed m detail, and their design, construction, and operating characteristics are compared. Their construction covers the regime of homogencous reactor technology involving the feasibility of relatively small reactors fueled with aqueotls =olutions of urantum. Since their construction and operation does not include svstems fueled with aqueous suspensions of thorium oxide and or uranium oxides necessary for the development of full-scale homo- geneous breeders or converters, additional experimental reactors will un- douhtedly he built. 7—2. Warter BoiLers* 7-2.1 Description of the LOPO, HYPO, and SUPO [2-4]. Interest in homogencous reactors fueled with a solution of an enriched-uranium salt was initiated at the Los Alamos Scientific Laboratory in 1943 through an attempt to find a chain-reacting system using a minimum of enriched fuel. The first of a sequence of such reactors, known as LOPO (for low power), went critical at Los Alamos in May 1954 with 565 grams of U2?5 as uranyl 431003 | 722, ~ aalpA dwng A—. < SEERN Joppindag juswuinyuy y abBing i "5 , . =223 0 [auiquioaay & SR * T ~18zZ1Inssaly m . 2 10)DIaudn) | < \~yn . - P 234G g, 2BURYIXT tDOH mmc.._flw ' EN::@&E§~M g l&* umog-e] ¥4 2josuspuo’) m B g T ., - : v u .‘«s ] 1asuapuon) i : 7 o \¥ ~ - siaqiospy npoud uoissiy < umoq-§91 pag uoudiosqy 1 aBung e BUIPO| PUD JaUIqUWOI3Y W = aiosuspued F7777) _ T T~ e uonjog yerjupig pue 2103 I waysAg \vf\\\\\ A wnnapa T osuigun) o wosyg soay- 1D FETNE pinbi puo sno P4 DHO i¥up|g o0 [ yapig o) ¢ 364 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 necessary to accommodate the greater volume of blanket fluid—the entire blanket system is not shown. The fuel system is described as follows, In the high-pressure system the fuel solution is pumped into the bottom of the reactor core at 256°C and is heated to 300°C as it proceeds upward to the outlet pipe. At the top of the outlet pipe is attached the pressurizer in which condensate is electrically heated to a maximum temperature of 335°C to produce 2000-psi steam. The fuel flows past the pressurizer to the gas separator, where directional vanes cause the fluid to rotate suffi- ciently to separate the radiolytic gas (D2 and O2), the excess O2, and fission- product gas. The separated gas forms a vortex along the axis of the pipe and is bled to the low-pressure system. The reactor solution continues from the gas separator to the U-tube primary heat exchanger, where it is cooled from 300°C to 256°C by transferring heat to the boiler feedwater surrounding the tube bundle. The 244°C, 520-psi steam produced on the shell side of the heat exchanger is bled partially to the small (345-kva output} turbine generator remaining from the HRE-1 and partially to an air-cooled steam condenser. The uranyl sulfate solution flows next to the intake of the 400-gpm canned-motor circulating pump and thence is pumped to the core for reheating. The blanket fluid follows an identical cycle at a flow rate of 230 gpm. This lower flow rate was based on a pump of the same horsepower but designed to circulate a thorium oxide suspension, rather than pure Ds0. The gases and some entrained liquid removed by the gas separator are transferred to the low-pressure system through a “'letdown heat exchanger,” a jacketed pipe which cools the gas-liquid mixture to 90°C. A valve downstream of the heat exchanger throttles the gas-liquid stream to atmospheric pressure. The mixture then discharges into the “dump” tanks, which have sufficient capacity to hold all the reactor liquid. An evaporator built into the dump tanks provides continuous mixing and, more important, steam for dilution of the deuterium and oxygen below the explosive limits. The gas-and-steam mixture flows upward through the iodine bed to the catalytic recombiner, in which the deuterium and oxygen react on a bed of platinized alumina pellets to form water vapor. The heavy water is condensed by the shell-and-tube condenser following the recombiner and normally flows back to the dump tanks. However, the water may be diverted to weighed storage tanks in case it is desired to change the concentration of the fuel solution. Water which is returned to the dump tanks is mixed with the excess fuel solution stored there (ap- proximately 25 gal) and then fed to the intake of a sealed-diaphragm injection pump, which returns the liquid to the high-pressure circulating system at a rate of about 1 gpm, thus constantly replacing the liquid removed via the gas separator. The small volume of intensely radioactive fission gas plus the excess THE HOMOGENEOUS REACTOR TEST (HRE-2) 365 7-4] “Jourejuod JodsA pus ppaIys Z-JHH '8-L ‘PIg - S Lo s (70 “ut ) siog Budicjuiay oIg "ul-t Buippo) 9245 ss3jutig A Pl soo|4 Juawdinby Aipyjixny S ey ™ [eyems 19IDM PUD pung sajliog N o - Al N T ) fi B ] 7y A . A - - .J/ Iz\nt.fi..“ Y i “ DDA L i “, > .cED_oU adiy "ui-¢ | poayyng 4 ” DR~ X 4 0’618 uUoHDAR|] O} 0.—0._UCOU >._UC_mu._o R . U0 M ST {a1910u07) sa4dung) 49Apq Bn|g do) {s18A0q Bn|q usamjag) Joayg |peg 3606 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 oxygen remaining after condensation of the re-formed heavy water is dried in cold traps at —23°C and sent to the beds of activated carbon for a period of decay. Gas leaving the bed is diluted with 1400 ¢fm of air and discharged to the atmosphere from a 100-ft stack. Samplers are provided to sccure small quantities (5 ml) of the fuel and blanket liquids from the high- and low-pressure systems for chemical analysis. These units are located in bypass lines; material circulated through them is trapped by closing the sampler inlet and exit, after which the contents are discharged into a portable container through a drain valve. All the primary reactor equipment is located in an underground, box- like, steel tank called the “shicld pit,” shown in Fig. 7-8. The design of the shield pit was influenced by several factors, including a requirement for accessibility and flexibility because of the experimental nature of the in- stallation, provision for complete containment of the contents of the re- actor should a leak develop or should the pressure vessel or heat exchangers rupture, efficient utilization of the space within an existing structure, and capability of flooding with water for maintenance or replacement operations. The reactor shicld pit occupies the center high-bay area of the building and is constructed of 3/4-in. welded steel plate reinforced in such & manner that an internal pressure of 30 psi will be contained. This pressure corre- sponds to the instantaneous adiabatic release of the entire contents of the reactor svstem. The chemical processing cells, each 12 ft wide by 25 {t long, are designed similarly. The upper surface of the blocks forming the roof of the shield pit 1s at ground level. This roof is made of high-density concrete 5 ft in total thick- ness and consists of two layers of removable slabs with a completely welded steel sheet sandwiched between the layers and extending across the top of the pit to form a gastight lid. The roof blocks are anchored to the girders and supporting columns by means of a slot-and-key arrangement, shown in Fig. 7-8. The vertical columns are embedded in a concrete pad which 1s 3 ft thick and is heavily reinforced with steel. The wall hetween the reactor pit and the control area is a hollow box 5% ft wide, constructed of 1/2-in. steel plate welded to the north side of the reactor shield tank. It is filled with high-density barytes, sand, and water. The use of the fluid shield between the reactor and control-room areas allows flexibility in the locations of service piping and instrument or electrical conduits. All lines leaving the reactor tank are welded mnto the shield wall; conduits are connected iuto junction boxes inside the pit with gastight seals on the individual wires. The design is such that nowhere outside the shield will the radiation dosage exceed 10 mrep/hir when the reactor is at 10 Mw. l'or the purpose of decreasing the neutron activation of equipment inside the pit, the re- THE HOMOGENEOUS REACTOR TEST (HRE-2) 367 7-4] lasuspuor) wnaig 0 G 3 o o 3 w ® 2 JaUiquIocoay jajun|g ‘matA ur[d ‘QuowedurLie juauodwiod - "6-2 PLI yue] dwng 43N0 J9up|g (1oxuoig) 19Bupy>x3 josy 7o ML Auo dwngq jaxjuo|g ° sdosj p|od 2un|g 101003y Juoy 8ipsUapuUO jaxyuD|g SSAIDA sdwing pP334 i93ub|g —— J9suspuor xn|45y 43Z1Inssaly SBAIDA sdwny pead |20y juo] dwng . JanQ |any sdoa) pjoy jeng {]ony) JeBubyox g joay lasuspuc) Jaulquodey |any 368 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 F1a. 7-10. HRE-2 container looking southeast {at 509, completion). F1a. 7-11. Reactor tank with shielding plugs in position during hydrostatic test. 7-4] THE HOMOGENEOUS REACTOR TEST (HRE-2) 369 actor vessel 15 surrounded by a thermal neutron shield consisting of a steel tank with a 2-ft-wide annulus filled with boron ore and water. The arrangement of reactor components within the main reactor pit is shown in Fig. 7-9. The reactor is near the center of the pit, enclosed by the 2-ft-thick thermal shield. To the left of the reactor is all the fuel- system equipment, and to the right is the blanket equipment. Most of the high-pressure components are located close to the pressure vessel. The pressurizers are positioned directly above the reactor; the gas sepuarators are in the S-shaped outlet lines to the steam generators on the south side of the pit. The areas to the left and right of the reactor in the center bay are reserved for future equipment modifications or additions. To the far left is grouped all of the fuel low-pressure equipment, including dump tanks, recombiner, condensate tank, and cold traps. These compo- nents are assembled on a rigid structural-steel frame; the blanket low- pressure equipment is to the far right. Insofar as possible, valves are situated close to the control-room wall so that air lines and leak-detector lines can be kept short. The arrange- ment of the valves is such that all flanges can be easily disconnected from above. 7-4.3 Schedule of construction [19,20]. Construction of the HRE-2 was started in July 1954, immediately after the dismantling of HRE-1. The initial step was the excavation of a large hole beneath the building (which had previously housed the HRIZ-1) for the large rectangular steel tank (60 ft long, 304 ft wide, and 25 ft deep) which contains the reactor and its associated equipment. [igure 7-10 shows this tank at approximately 209 completion, The north wall of the reactor tank, seen to the left in Fig. 7-10, is com- mon to both the reactor tank and the control-room area, and it is through this wall that the many service, instrument, and electrical lines which must interconnect these two areas pass. The next step in the construction was to install the approximately 600 lines which penetrate this 5i-ft-thick wall. Figure 7-11 shows the tank after the complete roof structure had been assembled and welded closed as it was prepared for a hydrostatic test. In this test the tank was filled with water and then pressurized to give the equivalent of a 30-1b internal pressure at all points within the tank. Strain gauges were attached at many points so that the complicated stress pattern could be studied in some detail for assurance that the tank was safely within design limits. While the reactor tank and control-room areas were being constructed, reactor equipment was being procured and constructed at several places. Much of the equipment required in the low-pressure system had been inspected and tested when the container was completed. In November [cHAP. 7 . a2 £ < = g mm ....,,. = A s W e W 1 g o g of HRE TAL REACTOR DESIGN AND CONSTRUCTION MEN EXPERI 370 -12. View Fic. 7 Fia. 7-13. Artist’s coneept of homogeneous reactor test, HRE-2. 7-4] THE HOMOGENEOUS REACTOR TEST (HRE-2) 371 DO | Supply To Oxygen System {3000 psi) Vent to Air Control Area Reactor Cell 8 Pressu.e Drop Alarm ] 2 Seal Weld - —s i< s SIFQOF:'LJC:;?j y femmamel < . Secl Weld Indi‘vfiual Test . { e e, Point Line - Level Cnnfro\\\ ] Pressure Goge —+ 9 B Y ! North Shield wall Flanged T ot Joint I < ot i e~ e e Legend: e et ——{>fme Valve Normolly Open d -t Valve Normally Closed - Sectionalizing Heoders Fic. 7-14. HRE-2 leak detection system. Pressurizer Vessel 1955, these parts and the thermal shield surrounding the reactor pressure vessel were installed. Figure 7-12 shows the reactor core and pressure- vessel assembly just before installation in January 1936. The heat exchangers were subsequently installed and followed by the main circulating pumps so that the high-pressure piping which connected the pumps, heat exchangers, and pressure vessel could be attached. This work occupied most of the months of February and March. Construection of the reactor was completed in May 1956. ['igure 7-13 is an artist’s concept of the completed reactor. 7-4.4 Nonnuclear testing and operation. Pretesting, operation of the reactor as o nonnuclear facility, and a lengthy flange-replacement job occupied the period from completion of construction in May 1956 to De- cember 1957. A chronological summary of the events associated with the nonnuclear operation of the reactor during this period is shown in Table 7-6. From Table 7-6 1t can be noted that preoperational testing of HRE-2 was interrupted by stress-corrosion cracking difficulties, which were caused by chloride ion contamination in the stainless steel tubes that are used to detect and prevent leakage of radioactive solution from flanged joints [20]. Figure 7-14 shows how an individual leak-detector line is attached to the groove of the ring-joint flange. The tubes from all the flanges terminate at a valve header station in the control room. Normally this system is kept pressurized with water to a pressure of 300 to 500 psi above the system pressure. A leak m any flange results in leakage of water from the header and o loss in pressure, which actuates an alarm at a fixed level above the fuel or blanket pressure. This is normally o sensitive and satisfactory means 372 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. TABLE 7-6 SumMaryY oF HRE-2 NonNNUCLEAR OPERATION Period Test or event May 1956 June 1956— July 1956 July 1956— August 1956 August 1956 October 1956 October 1956- November 1956 December 1956-- January 1957 January 1957- March 1957 April 1957- August 1957 August 1957- September 1957 September 1957 October 1957 November 1957— December 1957 3000-psig hydrostatic test of high-pressure system and 750-psig test of low-pressure system Cleaning of piping systems with 3%, trisodium phos- phate followed by 59 nitric acid and initial operation of pumps. Tests for dump-tank entrainment and efficiency of catalytic recombiner Initial tests of equipment removal and underwater maintenance Flushing of flange leak-detector tubes to remove chlo- ride contamination Initial nonnuclear operation of reactor at 280°C and 2000 psig. Thermal cycling of flanged joints Removal of typical flanges for metallographic examina- tion to detect possible stress-corrosion cracks. Further tests of remote-maintenance tools Further operation of the reactor with water and with depleted uranium at various conditions of tempera- ture and pressure Removal, inspection, and replacement of flanges and leak-detection tubing Recleaning of system with trisodium phosphate and nitrie acid solutions and hydrostatic testing Final operation at design conditions with water and with depleted uranium. Final leak test of reactor piping with radioactive tracers Final insulation of reactor piping; installation of new refrigeration system and new iodine absorption bed, followed by nonnuclear operations with heavy water. Criticality achieved December 27, 1957 7-4] THE HOMOGENEOUS REACTOR TEST (HRE-2) 373 of preventing the leakage of radioactive liquid from the reactor (the leak- detector fluid leaks out instead) and detecting the leaks when they do occur (by measuring the pressure or volume loss of the leak-detector fluid). Volume changes in the header can be read to 41 cc from o graduated scale next to each header sight-glass. DBy observing level changes at regular intervals, noting which lines are isolated from the header, leaks of less than 2 ce per day can be detected. Since this system is a secondary portion of the reactor, the leak-detector tubing unfortunately did not receive the same attention from the stand- point of specification and materials control as the stainless steel used in the primary piping. The 1/4-in. type-304 stainless steel had been purchased to standard ASTM tubing specifications, but in 30- to 40-ft lengths instead of the usual 20-ft lengths. The tubing was received and installed without difficulty. It was given a hydrostatic test after installation and put in service with distilled water. After approximately three months it was observed that some of the water drained from the leak-detector system was badly discolored. An analysis revealed the liquid to contain approxi- mately 1000 ppm of chloride. Since conditions of operation up to this point had been relatively mild, it was thought that the chloride might be removed simply by flushing, and approximately six weeks were devoted to disassembling the reactor and washing out all detectable indications of that contaminant. After discussions with the tubing manufacturer it was concluded that the chloride had originated from a die-drawing compound which had not been removed from the inside of the tubes prior to annealing. The presence of the chloride-containing hydrocarbon caused carbide precipitation at the grain boundaries during annealing and created tiny caves into which the chloride penetrated. The pickling and cleaning treatment which followed did not remove this material; in fact, it was learned that the manufacturer’s pickling tanks did not accommodate the full length of the tubing, making 1t necessary to pickle by dipping approximately half the tubing at a time. The net result was that large quantitics of chloride remained inside the tubing to be leached out later when filled with water. At the time the stress-cracking damage was discovered late in 1956 the reactor had been made ready for a series of engineering tests, and for this reason it was decided to make a brief inspection of the damage resulting from the chloride contamination before proceeding with the planned ex- perimentation. This preliminary inspection provided the basis for a decision to prepare for the replacement of the 259 flanges and the 15,000 ft of leak-detector tubing in the system. It was further decided that engineer- ing tests which had been interrupted could proceed for the period of approximately three months which would be required to procure new flanges and leak-detector tubing. 374 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION Lcaap. 7 Dismantling of the system was begun in April, and a very careful in- spection of the 259 flanges was made to determine how many should be replaced. The Super-Zyglo dye-penetrant method of flaw detection was chosen as the most sensitive test which could be used practically in the field. Of the 259 flanges inspected, 167 were found to be acceptable (i.e., as good as new), 67 were rejected because of cracks, pits, or other possible flaws, and 25 were judged questionable. Nearly all of the rejected flanges were in high-temperature portions of the reactor. While the inspection method was selected as the best available, it was not judged to be infallible; e.g., differentiation between mechanieal scoring and corrosion pitting was not always clear-cut, and any cracks covered by smeared metal resulting from excessive gasket pressure could not be detected. Hence it was de- cided to replace all the high-temperature flanges with new flanges or, where this was not possible, to remove 0.02 in. of metal from the flange surfaces. A total of 132 flanges were replaced; 15 were remachined. In addition, the 1/4-in. stainless steel tubing to all the high-pressure flanges (approxi- mately 10,000 ft) was replaced in the leak-detector system. This repair work was completed in August 1957. To remove any organic material introduced during repairs and to pre- treat the fresh metal surfaces incorporated into the system, the reactor piping was subsequently flushed with hot 39, trisodium phosphate solution, followed by water rinses and a 5% nitric acid wash. After hydrostatic testing the reactor was test operated with condensate for 150 hr at 280°C, then charged with depleted uranyl sulfate solution. Test operation with depleted uranium included: (1) a series of concentration and dilution experiments to study the transient and equilibrium behavior of ions in the system, (2) checking of the inventory-control methods by comparing the fuel analyses and indicated system controls with the quantity originally charged, and (3) observation of the corrosion behavior by analysis of fuel samples during a 159-hr run at temperatures above 250°C. At the conclu- sion of the run the charge was recovered and found to agree well with the computed inventory, although chemical analyses of high-pressure-system samples during operation had indicated a uranium concentration 5 to 10% lower than the amount added would predict. Nickel analyses of the fuel solution pointed to a system corrosion rate of slightly less than 1/2 mpy. Before the reactor was charged with enriched fuel, the piping and shield were subjected to careful leak tests. To obviate the presence of helium in the piping in case further testing with helium became necessary, the reactor was first pressurized to 500 psig with nitrogen, to which was added 40 curies of Kr®? as a tracer gas. The shield was sealed and the reactor allowed to stand pressurized for five days, after which air samples were drawn from the shield and heat-exchanger shells for beta-activity scanning to detect 7-4] THE HOMOGENEOUS REACTOR TEST {HRE-2) 375 the presence of any leaking krypton. This test was inconclusive at the time, however, because of difficulties encountered in purifying the samples for counting. Large samples of the air being tested were stored in gas eylinders; then the piping was vented and repressurized with helium. After an addi- tional waiting period the sealed volumes were again checked using helium leak detectors sensitive to 1 ppm of helium in air. This test demonstrated that the piping leakage was less than 0.2 ce/day of helium, with a pressure ditferential of 15 psi across the leak. These results were subsequently confirmed by krypton data after analytical difficulties were resolved. With the integrity of the reactor piping established, the seal pans were welded in place on the shield roof, and the roof plugs were locked in place. By pressurizing the shield and flooding the roof with water, leaks in the seal pan welds were located for subsequent repair. The major shield leak was found to be through a 16-in. valve in the ventilating duet; when this was repaired, the total leakage fell from 25 ¢fm to approximately 1,/2 cfm. The shield was judged sufficiently tight at this point to proceed with the eritical experiments, after which repairs were continued. By pains- taking individual checking of all shield penetrations and the use of ther- mosetting resin to seal the metal lips to which the roof seal pans are welded, the leakage was finally reduced to 4 to 4.5 liters/min at 15 psig. 7-4.5 Nuclear operation [21]. Fuel charging began on December 24, and eriticality was achieved on December 27, 1957, with the core and blanket near room temperature and at a pressure of about 800 psig. Nuclear instrumentation for the test consisted of three fission chambers, viz.. two permanent chambers in the instrument tube outside the reactor pressure vessel and a temporary chamber inside the blanket vessel. An antimonv-bervllium neutron source was suspended in the thimble in the center of the core (zee I'ig. 7-18). The reactor was brought gradually to the eritical condition by the m- jection of enriched uranium into the fuel solution added to the dump tanks in batches of 100 to 400 g. Tuel feed pumps and purge pumps were operated continuously to provide mixing between the dump tanks and the high- pressure system. Following each addition the solution concentration in the high-pressure system was allowed to reach steady state, as indicated by a leveling-off of fission count rates. After 2060 g of U35 had been added and the temperature of the solution lowered to 29°C, the neutron source was withdrawn. At this point, the fission count rates continued to rise, indicating that eriticality had been achieved. Raising the temperature of the reactor slightly by pumping warm water through the heat exchangers stopped the nuclear chain reaction. By further varyig the temperature and concentration of the fuel solution, it was demonstrated that the nu- clear reaction could be easily and safely controlled in this manner. 376 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 After the initial eritical experiment the neutron source was moved from the core to the blanket thimble, and the reactor was brought to eriticality seven times at suceessively higher teraperatures ranging up to 281°C. In each experiment the reactor temperature was raised above the desired point by supplying steam to the heat exchangers, and a batch of fuel so- lution was injected into the dump tanks., After steady count rates showed complete mixing of the new fuel, the temperature was slowly lowered until the critical temperature was reached. It was held at this point for about 1/2 hr before procecding to the next experiment. IFigure 7-15 compares the experimental measurements of the HRIS-2 critical concentration as a function of temperature with concentrations calculated by various methods. It can be seen that the two-group caleulations predicted values about 209 below those observed. The harmonics caleulation, which used a convolu- tion of an age and a Yukawa kernel to represent slowing down in D20, gave quite satisfactory results. The multigroup ecalculation also gave results in agreement with the experimental data. The first operation of the reactor at significant power levels took place in February 1958. In April 1958 the power level was raised in steps of I Mw to the design power level of 5 Mw. Operation was exceptionally smooth, and no mechanical difficulties were encountered in the first 500 hr after charging the reactor with U??5 Unfortunately, shortly after reaching full-power operation a crack developed in the tapered portion of the zirconium core tank, permitting fuel solution to leak into the blanket After a series of tests to determine the magnitude of the leakage and cal- culations to determine the behavior of the reactor with fuel in the blanket, it was decided to operate the reactor as a one-region machine (i.e., identical fuel solution in core and blanket). Operation of the reactor under these conditions was resumed in May 1958. 7-4.6 Operational techniques and special procedures. Reacfor starfup. As the size of homogeneous reactors increases, the use of control-rod neu- tron absorption to perform a startup becomes progressively less attractive; e.g., to maintain criticality in HRE-2 while heating from 20 to 280°C re- quires a reactivity increase of more than 259, Ak because of the large negative temperature coefficient. It is mueh more convenient to provide a means of varying the amount of fuel in the core as required to overcome temperature and power coefficients. During the initial experimental stage of HRE-2 operation, startup was begun by evaporating heavy water from the dump-tank solutions, con- densing, and pumping to the core and blanket circulating loops. The filled loops were then pressurized, circulation was initiated by starting the canned-motor pumps, and the circulating stream was preheated to operat- ing temperature with an auxiliary heat source. The concentrated fuel was then pumped from the dump tanks to achieve criticality. 7-4] THE HOMOGENEOUS REACTOR TEST (HRE-2) 377 12 l 1 | | i | | 10 [— O Gravimetric ] % O Apparent (From Multigrqup o Inventory Data) Calculatien o < 8 | wy ™ N:) Harmonics o Calculation c 6 — ] 2 s £ With Thimbte e In Core & 4 | o U o o T \ Two-Group Calculation S 2 = B i | | l | | | 0 40 80 120 160 200 240 280 320 Temperature, °C Fi1c. 7-15. Critical concentration of HRE-2 as a function of temperature. An alternate procedure involves starting the reactor without preheating hy varving the fuel concentration. When sufficient fuel has been added to raize the reactor temperature to its normal operating level, power with- drawal 15 begun. Temperature adjustments may be made by removing pure =olvent to temporary storage tanks or adding pure solvent to the cireulating fuel solution. A vuriation of this procedure is to fill the reactor slowly with fuel of the final coneentration. In this case the reactor will become chain-reacting at a low temperature with the core tank only partially filled with fuel solution. As the quantity of liquid in the core is slowly increased, the temperature will rise until the desired temperature is attained with the core completely full. The time required for startup is determined by the rate at which heating can be permitted. The same limitations apply to homogencous reactors as to other reactors in this respect. Generally, the heating rate of 100°F/hr is considered reasonable and unlikely to produce excessive stresses. Prob- ably more important, but more diflicult to determine, 18 the temperature difference which exists across heavy walls. Keeping these temperature dif- ferences to less than 100°F prevents excessive stresses. Once the tempera- ture limitations are established, the startup rate of fuel addition can be set to mateh them. Thus, although aqucous homogeneous systems have been demonstrated to be inherently stable, restrictions are nevertheless placed upon the oper- 378 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cuar. 7 ator to prevent excessive power surges and the resultant excessive pres- sures or heating rates. These restrictions are generally in the form of elec- trical interlocks in the control circuit. IFor example, a typical interlock might prevent the operator from concentrating fuel if instruments indi- cated that an excessive concentration was being reached, or an interlock might stop the addition of fuel solution if the temperature-measuring de- vices ndicated too high a heating rate. Although practically none of these mistakes Is serious enough to cause a reactor aceident, they are evidenee of poor operating technique, and if permitted might result in more serious mistuakes which could cause damage in spite of the inherent stability of an aqueous homogeneous system. The reactor 15 considered “'started up” when the desired operating tem- perature hus been achieved and the reactor is ready for power extraction. Operation. LExcept for several preliminary preparations, such as warming-up the steam turbine, power extraction is the simplest part of the reactor startup routine. This involves merely turning steam to the turbine, bringing the turbine up to speed, and making the necessary elec- trical switching changes to distribute the clectrical output. Once the gen- erator is synchronized and feeding into a larger power network, there is little for the operator to do except to see that the equipment is checked routinely for proper performance. For HRE-2 this normally takes a crew which consists of a supervisor, an assistant engineer, and nontechnical helpers. Checks of all continuously operating equipment are made at 1- to 2-hr intervals to verify that the equipment is performing properly. From time to time samples of fuel solution are removed from the high-pressure circulating system to analyze for nickel and other unwanted ions. Samples must also be removed from the steam system to show that boiler feedwater treatment is adequate and that oxygen production is not excessive in the steam generators. Radiation levels must be observed to determine whether there are leaks of fuel solution or weak places in the shield structure. In addition to these routine service functions, the operating crew must also start up and maintain the chemical processing plant associated with the reactor. One engineer and one technician, in addition to those already mentioned for the reactor proper, are required to operate the HRIE-2 chemical plant. Shutdown. The shutdown of the reactor is normally accomplished in two steps. The first step in the procedure is cessation of power removal, which in- volves nothing more than closing the steam throttling valve. This might be accomplished by running the turbine governor down to zero load. Although this action leaves the reactor critical at 280°C and does not com- pletely stop heat generation, the power is limited to the normal heat losses from the system (a few hundred kilowatts for HRE-2). Minor repairs or 7-4] THE HOMOGENEOUS REACTOR TEST (HRE-2) 379 adjustments to equipment which do not require further cooling can now be made. The second step is to make the reactor suberitical by diluting the fuel. This type of shutdown normally requires 3 to 5 hr, since it is desirable to have the reactor subcritical at the storage temperature of approximately 25°C when the reactor is emptied. The rate at which the temperature can be reduced is again determined by the permissible cooling rate of the system components. Scram. A more rapid shutdown, equivalent to an emergency scram in a solid-fuel system, is the “‘dump.” In this situation the reactor is kept circu- lating for 2 min to permit recombination of the radiolytic gas in solution during which time the steam valves are closed to reduce power output. Then the pressure in excess of the vapor pressure of the core and blanket is vented to make pressure balancing between core and blanket easier, and the dump valves are opened, permitting the rapid emptying of the core and blanket vessels so that the reactor is shut down within minutes. This type of shutdown is only resorted to in case of emergencies such as excessive pressures or evidence of a leak of radioactive solution from the reactor. Decontamination of equipment. Conventional methods of decontaminat- ing fuel processing plants have proved to be inadequate for a stainless steel homogeneous reactor system which has been exposed to uranyl sulfate- sulfuric acid-fission product solutions at 250 to 300°C. This was demon- strated with HRIE~1, which was decontaminated (without descaling) prior to disassembly (Article 7-3.13). Since the HRE decontamination was in- complete, laboratory studies were carried out to explore the nature of the contamination and develop methods of decontaminating the stainless steel. Tt was found that chromous sulfate, a strong reducing agent, would modify the oxide film and permit dissolution in dilute acids. A 0.4m CrS04+—0.5 m Ho80,4 solution has given excellent removal of the film by modifying and dissolving the oxide corrosion film. Decontamination fac- tors of 5 X 10° were achieved on specimens from in-pile corrosion loops, where the activity was reduced to the induced activity of the structural material, by contacting for 4 hr with the chromous sulfate solution at 85°C. The solution was also tested satisfactorily on four 22-liter uranyl sulfate corrosion loops which had run for 22,000 hr at 200 to 300°C. The loops had a very heavy oxide coating such that in thermal cycles large flakes broke off the wall and plugged small lines. After a 4-hr contact with the chromous sulfate-sulfuric acid solution at 85°C, the walls of the loop were completely free of all clinging oxide. The total time involved in the preparation of the chromous sulfate solution, in descaling the reactor system, and in disposing of the extremely radioactive scale-waste would probably be at least 48 to 72 hr. If one is 380 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 considering decontamination as an aid to maintenance, time required for decontamination should be weighed against the reduction in repair time which would result from the lower levels of activity in the working area. 7-4.7 The HRE-2 Mockup [22].* During the design and construction of HRE-2 some of its major pieces of equipment were assembled and operated at design conditions on unenriched uranyl sulfate solution. The purposes of this engineering mockup were (1) to study the behavior and removal of gases in the high-pressure system, (2) to study fuel-solution stability in a circulating system similar to the HRE-2, (3) to establish a reference corrosion rate for the reactor circulating system, (4) to study the behavior and removal of corrosion- and fission-product solids in the system, and (5) to establish the reliability of components, pointing out weaknesses and possible improvements. The prototype reactor components tested in the mockup include: (1) a Westinghouse canned-rotor centrifugal pump, (2) an electrically heated steam pressurizer, (3) a centrifugal gas separator, (4) a 1/8-scale heat exchanger similar to the HRE-2 steam generator, (5) a letdown heat ex- changer to cool the fluid and gas bled from the high-pressure system, (6) a bellows-sealed letdown valve to throttle the fluid and gas from the high-pressure system, (7) a liquid-level controller which adjusts the letdown-valve position, (8) the dump-tank and condensate system for excess fuel solution and storage of condensate for use as purge to the pressurizer and circulating pump, (9) diaphragm feed pumps to return fuel to the high-pressure system from the dump tanks, (10) an oxygen-feed system to maintain oxidizing conditions in the eirculating fuel, and (11) an air-injection system for tests of the effectiveness of the gas-separator unit and letdown system. In May 1956 a corrosion- and fission-product solids-removal system was mstalled m the mockup, consisting of a 5-gpm canned-rotor ORNL pump, an assembly of a hydroclone and underflow pot, and two 1/2-in. I'ulton Sylphon bellows-sealed air-operated valves separating the system from the main circulating stream. A commercial pulsafeeder was used to inject rare earths containing radioactive tracers and corrosion products. In ad- dition, a through-flow bomb was filled with the required solids and con- nected into the system. Gamma-ray counting equipment was installed to detect any buildup of radioactive tracers on the heat-exchanger surfaces, in the horizontal connecting pipe to the pressurizer, in the gas separator, and in the underflow collection pot. A multichannel gamma counter was used for determining activity levels at the various counting stations. The equipment was removed in November 1956 after demonstrating satis- factory removal of solids from the system. *Article 7-4.7 is based on a paper by L. Spiewak and H. L. Falkenberry [22]. 7-4] THE HOMOGLNEOUS REACTOR TEST (HR-2) 381 Operation of the mockup for a total of more than 13,000 hr in the three- year period from Iebruary 1955 to February 1958 provided valuable in- formation pertinent to the construction and operation of HRE-2. Items of major importance include: (1) a satistactory demonstration of the opera- tion of the equipment for removing gases from the high-pressure system, (2) the out-of-pile chemical stability of the uranyl sulfate solution over a long period of time at operating temperatures and pressures, (3) determina- tion of the oxygen injection and excess sulfuric acid requirements to pre- vent uranium precipitation in the high-pressure loop, (4) detection of an unsatisfactory pressurizer design in which excessive corrosion and precipi- tation of uranium occurred, and tests of a revised model which proved suitable, (5) & demonstration of the suceessful removal of injected fiscion- product solids and insoluble corrosion products by means of a hydroclone, and (6) long-term operability of the circulating pump and other pieces of equipment, 7-4.8 The HRE-2 instrument and control system [23].* The control system for & homogeneous reactor such as the HRIS-2 differs drastieally from that for solid-fuel reactors hecause control rods and fast electronie circuitry are not necessary for systems with such large temperature co- efficients (nearly 0.9, Al/°C at 280°C for the HRIS-2). I'unctions similar to those performed by control rods in heterogeneous reactors, but without such exacting speed-of-response requirements, are performed by valves which control the concentration of the fuel, which ary the steam-removal rate from the heat exchangers, or which allow the fuel to be discharged to noneritical low-pressure storage tanks, Ifor these rensons the nuclear control eircuits for a homogeneous reactor are designed to limit very rapid changes in fuel concentration and steam-removal rates. Other circuits control the pressures und temperature limits of the cireulating liquid fuel, principally to prevent equipment damage. In addition to these general considerations, the instrument and control system for the HRE-2 includes the following special features: (2) All instrument lines through the shield wall are blocked by valves on a signal of high shield pressure to prevent the escape of radioactivity. Electrical leads are sealed by glass-to-metal seals. (b) Thermocouple, electrical, and air lines may be disconnected by remotely operated tools for equipment repair or removal. (¢) All eritical core- and blanket-system transmitters, except electric level transmitters and thermocouples, are located in two shielded nstru- ment cubicles (5 ft diameter X 15 ft high) located adjacent to, but out- side, the main reactor tank. This arrangement wus selected to avoid opening the main tank to replace instruments, to provide a location for *Article 7-4.8 is based on a paper by D. 8. Toomb [23]. 382 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cmAP. 7 Pressure Switches Pressure - Switches Eiectric Controller }-{ Recorder Pneumatic Converter Control Circuit Dump Tank Pulsafeeder Pump Fia. 7-16. Key control loops utilizing both pneumatic and electric transmission. instrument components which could not be easily protected from the water-flooding of tne main tank during remote-maintenance operations, and to minimize the radiation exposure of instruments. (d) The cell air monitors, which provide an alarm in case of a leak of radioactive vapor from the reactor system, are installed in one of these mnstrument cubicles. Cell air is circulated through a 2-in. pipe from the reactor tank, past the enclosed monitors, and then back to the cell. The blower is sized so that only 5 sec is required for cell air to reach the radia- tion monitors. I'igure 7-16 indicates several key control loops: (a) The pressure of the core system 1s controlled from sensed pressure by the proportioning of power to the pressurizer electric heaters; the blanket pressure is similarly controlled by a core-to-blanket differential-pressure signal. (b) The liquid levels in the pressurizers are controlled from sensed levels by pneu- matic control of the letdown valves. Pneumatic control actions are derived from transducers which receive signals from electric transmitters. Electric interlock control of the pneumatic signals to final control elements is achieved by the use of solenoid-actuated pilot valves. Other control loops not illustrated are: (¢) The reactor power is con- trolled by a manual or turbine-governor signal to a valve throttling steam from the core heat exchanger. Because of the large negative temperature coefficient and low heat capacity, this reactor system cannot produce more heat than is being removed, except for very short times. Heat is removed by opening the steam throttling valve, which causes a decrease in the temperature of the fuel leaving the heat exchanger. Because of the nega- 7-4] THE HOMOGENEOUS REACTOR TEST (HRE-2) 383 Overflow Pipe To Wcsie/ Reactor Thermal Shield Boral Sleeves 12#0in, ———= L | Fia. 7-17. HRE-2 nuclear instrument thimble. tive temperature coefficient of the system, the cooler fuel entering the core causes an increase in reactivity and the fuel is reheated until it over- comes the excess reactivity. (d) The average temperature of the core system is controlled by varying the concentration of the fuel solution. The inlet and outlet temperatures will vary with the power extraction, while the average temperature is a function only of fuel concentration. (e) The blanket temperature is controlled by a signal, derived from the difference in average core and blanket temperature, which operates the steam withdrawal valve of the blanket heat exchanger. Nuclear instrumentation. Neutron level transmitters are two Westing- house fission chambers and two gamma-compensated ilonization chambers designed by the Oak Ridge National Laboratory. Neither type of instru- ment requires a gas purge, and both are amenable to operation in the water- filled thimbles (I'ig. 7-17), which allow the chambers to be positioned or replaced during reactor operation. Varying the distance of the chambers from the reactor core affords a means of sensitivity adjustment, which is needed to accommodate different operating powers. The gamma-compensated ionization chambers are required to be in a 384 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHaPp. 7 neutron flux of approximately 10'¢ at 5-Mw power to utilize their measur- ing range of 10°. The lower limit of their range of operation is comparable to the maximum desirable flux for proper operation of the fission chambers, Therefore 1t was possible to install the two types of sensing elements in the same area in the reactor cell. At high fluxes the fission chambers which are required to follow the neutron flux for five decades during start-up are withdrawn into a protective boral shield to limit fission-product buildup in the chamber lining. Proper operation of the compensated ion chambers is ensured by the lead-shot-and-water fill around the thimbles, which re- duces the gamma-ray background after power operation from 250,000 r/hr to 250 r/hr. The fission-chamber signals are fed to conventional preamplifiers, A-1 linear amplifiers, logarithmic count-rate meters and a dual-pen re- corder. For initial reactor startup before gamma-ncutron reactions pro- vided a sizable neutron source, the fission-chamber output was used to drive a low-range pulse counter. Control panel. Figure 7-18 shows the main control board and console. Here are located only those instruments necessary for the safe operation of the reactor. These are arranged in a “visual aid” form to reduce opera- tional errors and to facilitate the training of operators. The graphic section is essentially a simplified schematic representation of the chemical process flowsheet with instruments, control switches, and valve-position indicators located In positions corresponding to their location or function in the actual system. Each annunciator is placed in the control board directly over the instru- ment or portion of the system on the graphic board with which its signal is associated. Key measurements are displayed on “full-scale” recorders in the center section of the panel and include the fuel temperature, a multipoint tem- perature recorder, the multiarea radiation monitoring recorder, the re- actor power, the logarithmic neutron-level and count-rate meter signals. The patch panel on the extreme left 15 a “jumper board,” which is a schematic representation of the electrical control circuits. Provision is made for jumping certain individual contacts in the control circuit with a plug. Lights indicate the position, open or closed, of the contact in the system. The board is valuable for making temporary control-circuit altera- tions necessary for experiments and makes 1t unnecessary to use jumper wires which are normally placed behind the panel and if overlooked and not removed might introduce a hazardous condition. The board is also an aid in familiarizing operators with the electrical control circuitry and, since the lights indicate contact position, as an operations aid during startup. Switches and controls on the console are restricted to those necessary for nuclear startup, steady-state power operation, and emergency. THE HOMOGENEOUS REACTOR TEST (HRE-2) 385 7-4] "[Pued JUIWNIISUI PUB WOOI [013U0d G- H 'S1-L D1 3806 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHaP. 7 Data-collection instruments and the transducers that drive the miniature pneumatic slave recorders on the graphic panel are located in an auxiliary instrument gallery beneath the main control room, as are a 548-point thermocouple patch panel, a relay panel, the nuclear-instrument amplifiers, and the nuclear-instrument power supplies. Other panels located near their respective equipment elsewhere in the building include the steam control station, the turbine control panel, two sampler control panels, and a control station for the refrigeration system supplying the cold traps. Standard 2-ft-wide modular cabinets and panels are used throughout to facilitate design changes. Protective interlocking. Extensive protective interlocking of the controls circuits is provided to prevent unsafe operating conditions. Examples of interlocked systems include the following: (a) The pumping-up of fuel to the reactor core instead of condensate is prevented by several interlocks which keep the fuel-addition valve closed until the core is full of condensate and has been heated to 200°C. This prevents power surges and consequent pressure increases. (b) For the same reason the fuel circulating pump is started in reverse to provide a low flow rate as protection against pumping cold fuel too rapidly from the heat exchanger into the core. (¢) Toavoid dangerous thermal stresses and abrupt reactivity surges, the control circuits do not permit the pumping of cold feedwater into the heat exchangers until the level is above 509%. (d) To give smooth startup, the fuel injection pump can run only at half-speed until a tem- perature of 250°C is reached. {(e) The fuel feed valve will be closed and the concentration in the fuel system will be lowered by injecting condensate if the core outlet temperature becomes excessive, if the circulating pump stops, or if the power level exceeds normal. (f) The contents of the high- pressure systems will be automatically emptied to the low-pressure storage tanks through the “dump” valves on a signal of extremely high pressure, or a radiation leak into the steam system. Differential-pressure control between the core and blanket systems during dumps is by throttling control of valves from a differential-pressure signal. Inventory systems. To obtain an accurate inventory of the fuel and moderator solutions, the storage and condensate tanks are weighed with pneumatic weigh cells. A pneumatic system was selected primarily because taring can be done remotely by balancing air pressures, and components are less susceptible to radiation damage. Piping to the tanks is kept flexible to compensate for the varying loads which result from thermal expansion. The volumes of the fuel and blanket high-pressure systems have been accurately measured, so that when the pressurizer condensate reservoirs are filled to capacity, the weight of liquid in the high-pressure systems can be computed from the core, blanket, and pressurizer temperatures. These 7-4] THE HOMOGENEOUS REACTOR TEST (HRE-2) 387 weights, combined with the respective condensate weights, dump-tank weights, and experimentally observed holdup in the condensate piping, yield the total liquid inventory of the reactor. 7-4.9 Remote maintenance [24]. Maintenance of the equipment in the circulating systems of an aqueous homogeneous reactor is difficult because of the intense radioactivity emanating from the surfaces following pro- longed operation at high power levels. This problem, more acute than in the case of heterogeneous reactors because fission-product activities and neutrons are not confined to the reactor core, must be successfully resolved in a practical manner if the homogeneous reactor is to be an economical power producer. Practical systems for maintenance and repair must pro- vide adequate personnel shielding to reduce radiations from activated equipment to tolerable biological values while simultaneously providing access to the equipment. Equipment activation. Of the two principal methods by means of which the equipment is rendered radioactive, activation by fission-product con- tamination is the more important in the HRE-2. About 30 w/o of the fission produects is normally removed from the fuel solution as gases (xenon, krypton, iodine) by stripping with Oz, D2, and steam. Of the remaining fission products, 56 w/o are estimated to have sufficiently low solubilities to precipitate from the solution as solids. This is particularly true of the rarc earths, which constitute 28 w/o of the fission products. Tellurium, technetium, and molybdenum, which contribute 12 w/o of the fission products, are assumed to be soluble. Therefore, so far as activation of equipment is concerned, the rare gases { ~309) and the soluble products (~12%), making up 42 w/o of the total, may be considered removed or easily removable from the equipment. Some portion of the remaining 58 w/o will deposit on hot metallic surfaces or be retained in cracks and crevices. It is possible by statistical analysis to estimate within a factor of two the specific gamma and beta activity of the fuel solution. Methods for caleulating this activity have been published [25,26]. The 1nitial fuel ac- tivity, after prolonged HRE-2 operation, is of the order of 25 curies/ml. It should be noted that it is proposed to remove the insoluble fission products continuously with small centrifugal-type separators (hydroclones) in the HRIE-2 fuel processing system. The effectiveness of this procedure in competition with absorption on the container walls, however, is us yet unknown, Induced activation of structural materials by neutron absorption ig also an important contributor of radioactivity. During operation of the reactor, a thermal flux exterior to the equipment items will be present because of neutron leakage from the reactor vessel. Also, delayed neutrons will be emitted in the interior of the piping and equipment from fission products 388 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 in the fuel solution. These neutron fluxes may be estimated, and knowing the surface areas and material constituents of the equipment, the resulting induced activation may be calculated. Methods of calculating induced activation are reported in the literature [27,28]. The effect of induced activation may be controlled to some extent by attenuating neutron leakage from the reactor through an adequate thermal- neutron shield surrounding the reactor vessel. Also, the constituents of the structural materials in the cell may be specified in such & manner that elements of potentially high neutron activation, such as the cobalt normally present in stainless steels, are minimized. Other materials, such as boron, which has a high neutron-capture cross section and is an alpha emitter rather than a gamma radiator, might be used on equipment surfaces to minimize induced activity. Radiation dosages. Design criteria followed in the design of shielding and tooling for remote-maintenance operations are based on tolerance radia- tion levels that are acceptable for normal operations. In continuous work- ing areas, radiation levels as low as 1 mr/hr are specified. For certain in- frequent operations a level of 74 mr/hr is permitted. However, during periods of intense maintenance it may sometimes be permissible to allow personnel to work in an area of higher than normal radiation, but the total weekly radiation dosage must be held to 300 mr or less. In some instances it is necessary to work in relays so that no one person receives a higher than permissible weekly dosage. Proposed maintenance concepts [24]. The concepts which have been pro- posed for the maintenance of equipment in the HRIE-2 have been based on the philosophy of either dry maintenance or underwater maintenance or combinations of the two. In general, it is assumed that the item requiring repair will be replaced with a new item and the faulty component removed to a shielded hot-cell area for maintenance, Tests of underwater-maintenance procedures proposed for HRE-2 using special tools and equipment have demonstrated the feasibility of removing and replacing equipment. In these tests it was demonstrated that con- tamination of the heavy-water fuel solution by shielding water could be prevented by freezing a plug of heavy ice in the pipe before disconnecting equipment. Ireeze jackets are placed around piping at flange disconnects for this purpose. Also in the HRIE-2, the insulation around piping was de- signed to avoid the retention of water after draining the flooded equip- ment. Piping joints. Removal of equipment in the HRE-2 depends on flanged joints adjacent to the equipment which may be disconnected and remade with suitably designed tooling. The joints in the HRE-2 use American Standard stainless-steel ring-joint, weld-neck flanges modified to permit leak detection (Fig. 7-13). These flanges are located so that all bolting is 7-4] THE HOMOGENEOUS REACTOR TEST (HRE-2) 389 Fig. 7-19. Universal jig for locating flange positions on HRE-2 replacement high-pressure valves. accessible to tools operated from above. The number of loose pieces such as bolts, nuts, ferrules, and gaskets is minimized by welding nuts to the flanges such that they become integral with the flange. Remotely operable tools.* Tools used in the IHIRE-2 maitenance pro- cedures are generally simple, rugged units mounted on long handles. They include manual, air-operated, and water-operated socket wrenches for turning nuts, hooks for lifting, clamps, knives, magnets, etc. Some typical maintenance devices are illustrated in Iligs. 7-19 through 7-22 and described as follows. The universal high-pressure valve jig shown in Fig. 7-19 is used for manufacturing replacement high-pressure flanges for the HRIE-2. By placing brackets in appropriate holes, referenced with respect to the 21 high-pressure HRE-2 valves, the position of the flanged pipe ends for any one of the valves may be accurately determined. With this device, the re- quired number of spare valves for the reactor is minimized, and at the same time any valve which fails can be remotely replaced. *Other remotely operable tools are described in Sec. 19.5.6 of the Reactor Handbook, Vol. II, Section D, Chapter 19, ORNL-CF-57-12-49. 390 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 F1a. 7-20. Spinner wrench for rapid removal of HRE-2 flange bolts. The spinner wrench in Fig. 7-20 is used for removing bolts quickly after they are loosened with a high-torque wrench, for spinning the bolts snug before final tightening and, by use of a right-angle bevel-gear attachment (not shown), for handling holts in a horizontal position. It has a chain- drive offset to fit bolts not accessible by direct drive, such as bolts hidden under pipe fittings and bends. The flange-spreader tool (Fig. 7-21) is required in remote operations to spread low-pressure valve flanges to prevent damage to the ring gasket during removal or ingertion of the valves. The tool is shown in place with the flanges spread and the valve assembly moved back on its cell fixture. The hydraulic torque wrench in Fig. 7-22 is used to tighten or loosen bolts on flanges larger than 2-in. pipe size. This wrench is actuated by a hydraulic (water-operated) cylinder and is capable of applying a 2000 ft-lb force moment. Even in the relatively short operating history of HRIS-2 the concept of underwater repairs with tools such as those described above has been proved to be completely practical. Soon after the reactor attained signifi- cant power it was found necessary to replace all the clectrical power wiring —nearly 4000 ft of wire. This was accomplished in & period of three weeks, with the cell completely flooded. During this period the fuel circulating pump was removed to recover a set of corrosion samples. The removal and replacement of the pump required only 100 hr working time, which 7-4] THE HOMOGENEOUS REACTOR TEST (HRE-Z) 391 Fi1e. 7-21. Flange-spreader tool for preventing damage to gaskets during re- moval of low-pressure flanges. Shown in place during maintenance operation on HRE-2. would have been the total shutdown time if the pump replacement alone had bteen performed. Also, the core inspection flange was removed and the core mner surface inspected by means of a periscope. In these three major underwater-maintenance operations no delay or difficulty was experienced. 7-4.10 Containment methods [29]. In designing for the containment of HRE-2, extremely rigid leakage specifications were set, both for the pri- mary piping system and the shielded tank containing the reactor. In the case of the piping, the leakage specification was based on the minimum leak which could be found with the available mass-spectrometer leak- 392 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 N Fic. 7-22. Hydraulic torque wrench for loosening or tightening flange bolts. Shown in operation on HRE-2, detection equipment, namely, 0.1 cc helium (STP) per day. Allowing for several such minimum leaks, the equivalent loss of liquid at 300°C and at 2000 psig could be as great as 5 cc per day, or a total leakage of approxi- mately 250 curies/day. Because of its volume and surface area, and be- cause of the difficulty in measuring small leakage from very large vessels, the leakage rate for the reactor tank was set at 10 liters/min or less, at a test pressure of 15 psi. As indicated in Article 7—4.4, actual leakages were below this value. The problem of leakage from the reactor vessels and piping can be cata- logued according to the various mechanisms through which leakage might result as follows: (1) excessive stresses, (2) defective materials or work- manship, (3) corrosion, (4) nuclear accidents, (5) hydrogen-oxygen ex- plosions, and (6) brittle fracture. Each possibility is examined in detail in the discussion which follows. Ezxcessive stresses. There are many possibilities for the development of excesslve stresses in a system as complex as the HRE-2. In order to reduce the likelthood of failure as a result of excessive stress, a maximum allowable working stress of 12,000 psi was specified for the type-347 stainless steel 7-4] THE HOMOGENEOUS REACTOR TEST (HRE-2) 393 with which the system was fabricated. This permits an additional factor of safety over the 15,000 psi allowed by the ASME boiler code. As required by the ASA code for pressure piping, the reactor piping arrangement was examined for maximum stresses due to pressure, as well as for hoop and bending stresses resulting from thermal expansion. Equipment was also studied for determining the magnitude of thermal stresses caused by radia- tion heating and temperature cycling. Therefore, in order to keep the com- bined thermal and pressure stresses below the maximum allowable working stress, the pressure-vessel wall is approximately 2 in. thicker than would be required otherwise. Heating and cooling rates on the entire system have been limited to 100°F/hr and 55°F/hr, respectively, and the differential temperature across heavy metal walls is kept below 100°F. (The cooling rate is held below the heating rate to minimize rate of flange contraction.) Cyclic temperature stresses at questionable points in the reactor pres- sure vessel and steam generators were explored experimentally. Mockups were fabricated for the testing of the pressure-vessel nozzle joints and the stainless-steel-to-Zircaloy bolted joint inside the pressure vessel. In each test the temperature was cycled from approximately 250°F to approxi- mately 600°F in 1/2 hr and cooled back to 250°F in 1/2 hr. After 100 cycles the joints were found to be sound. The main steam generators were also cycled in similar tests. Several tube joints cracked open during the first 50 cycles. They were repaired and the heat exchangers were subjected to an additional 10 cycles before final acceptance. Defective materials or workmanship. Defective materials and poor work- manship constitute another area which required special attention to pre- vent failures. All materials for the primary systems of HRE-2 were procured to specifications considerably more rigid than those existing in commercial practice. Optional requirements such as chemical analyses, boiling nitric acid tests, and macroetch tests were exercised in all materials specifications. An additional cost, averaging about 10%, was experienced in the purchase of materials under the more rigid specifications. In some in- stances, for example with the heat-exchanger tubes, special ultrasonic and magnetic eddy-current flaw detectors were employed to indicate defective parts. Three tubes which might have later failed in operation were thus eliminated. Dye-penetrant tests were applied to tubing bends and to all welds throughout the reactor to detect eracks and pinholes. None was dis- covered in tubing bends, but many were found in welds, especially in the tube-to-tube-sheet welds. Special attention was given to the welding of stainless steel butt joints, of which there are approximately 2000 in the entire reactor. The inert-gas, nonconsumahle-electrode method was used almost entirely. Welds were inspected to considerably higher standards than required by the ASME code. In addition to being subjected to dye-penetrant inspection, every 394 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 weld was x-rayed. Although the inspection standards were very rigid, only 39 of the welds were rejected, necessitating rewelding. Corrosion. Although corrosion is an ever-present possibility for leakage, the HRE-2 was designed to reduce attack rates so far as possible by keep- ing velocities below 20 fps, temperatures in the range of 250 to 300°C, and lining some surface areas with titanium for additional resistance. Nuclear accidents. Nuclear accidents are likely to be rare in homogeneous reactors because of the large negative temperature coeflicient (0.1 to 0.2% Ak/°C) [30]. As an example, the worst aceident considered in the HRE-2 [31] was one in which all the uranium suddenly collects in the reactor core and results in a reactivity increase of 2.5% Alk/sec. For this rate, starting at a power of only 0.4 watt, the maximum pressure in the pressure vessel would be approximately 3900 Ib, and the pressure stress in the carbon steel shell would be less than 30,000 psi. Hydrogen-oxygen explosions. Since radiolytic gases (deuterium and oxygen) are produced continuously in the reactor, they are the source of an ever-present hazard. Explosions may be expected whenever the deuterium-oxygen-steam mixture is more than 15% gas. For detonations the required gas fraction is greater. The maximum increase in pressure from an adiabatic explosion of hydrogen and oxygen is only a factor of 3 to 8, whereas for a detonation the factor might be 23 for an undiluted mixture. (It is important to note that detonations can occur only In gas channels that are relatively long and straight.) In low-pressure areas of HRIZ-2 the gas was diluted with steam to keep it noncombustible. Furthermore, a pressure of 500 psi is the basis for design of low-pressure (atmospheric) equipment. Even for a detonation the expected peak pressure would be less than the design pressure. Although an explosion can be tolerated in the high-pressure system with little danger of vessel rupture, a detonation probably could not be. A detonation could occur in the gas separator where the gas channel (the vortex by which the radiolytic gas is collected) is long and straight. It is caleulated [32] that a detonation wave traveling longitudinally along the vortex would produce impact pressures of the order of 30,000 psi but that the damage would be limited to the directional vanes inside the separator. Attenuation of the forces by the solution would limit the pressure rise to that resulting from combustion of the gas—10,000 psia, which produces a tolerable wall fiber stress of 35,000 psi. Thus no serious damage is foreseen from explosions or detonations. Brittle fracture. 1t is generally known that ferritie steels are subject to the phenomenon of brittle fracture. Although the likelihood of brittle fracture in the HRI-2 pressure vessel was known to be small, an investigation was made [33] in order to determine the consequences of such an accident as a result of the pressure rise and missile damage. 7-4] THE HOMOGENEOUS REACTOR TEST (HRE-2) 395 A sudden release of the liquid contents of the fuel and blanket systems— 3910 1b of solution at 300°C and 2000 psig—would result in a pressure rise of approximately 30 psi inside the reactor shield. Although it is somewhat difficult to design a large (25,000 cu ft) rectangular container such as the HRE-2 container to withstand a 30-psig pressure, it was even more diffi- cult to design it to withstand heavy missiles. Studies indicated that missile velocities of 50 to 150 fps could be expected. With the HRE-2 vessel, which weighs 16,000 1b, 1600-1b fragments might be expected. Such a mass at 100 fps would lift an unrestrained 5-ft-thick concrete shielding plug nearly 1 ft. For this reason a blast shield was placed around the vessel. The blast shield was designed to withstand the 1250-psig pressure which would result from the 300°C liquid, as well as to absorb the energy of the expanding steam. A 13-in.-thick wall made of type—304 stainless steel was found to be eapable of absorbing 2.85 X 10° ft-1b of energy with 29 elonga- tion. The fragments from the pressure vessel would accumulate this much energy in traveling across an annulus of 4.8 in. To provide an additional factor of safety, the blast shield was constructed to surround the pressure vessel with an annulus not greater than 2 in. A similar shield made of carbon steel was installed on both fuel and blanket steam generators as a final precaution against an explosion which might damage other equipment and the vapor barrier sufficiently to cause a release of activity. The probability of failure of the steam generator shells would not be affected by the presence of neutron radiation because the generators are located in a low flux region. The vapor-tight container. If leakage of radicactive solution from the reactor proper occurs through one of the means mentioned above, an all- welded, vapor-tight tank provides a second barrier to the escape of radia- tion to the atmosphere. Construction details of this tank, which forms the liner for the biological shield, are illustrated in Fig. 7-8. After completion of construction the tank was given a hydrostatic test at an average pressure of 32 psig, and the welded joints were found to be free of leaks. It was then reopened for further installation of reactor equipment. Prior to operation of the reactor at power, the containment vessel was sealed and tested (Article 7-4.4). 7—4.11 Summary of HRE-2 design and construction experience. Fol- lowing the completion of HRE-2, personnel associated with the project recommended design improvements which might be applied to good ad- vantage in future homogeneous reactors. These recommendations are summarized in the following paragraphs: (1) There are many components in the HRIE-2 system which must operate in conjunction with one another for continuous operation of the reactor. Since a failure of any one of these could cause a shutdown of the 396 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 reactor, the probability of shutdown is higher than it would be with com- ponents operating independently. In future reactors an attempt should be made to decrecase the dependency upon the simultaneous operation of essential components. An example of this is a system which can be operated without continuous letdown from the high-pressure to the low-pressure system, so that failure of the low-pressure system components will not necessarily require shutdown of the entire reactor system. (2) The reactor cell for the HRI-2 contains many structural compo- nents made of carbon steel. I'looding the cell with water and removing radioactive contamination by use of acids will eventually do serious damage to all these carbon steel surfaces. Wherever possible in future reactors, corrosion-resistant materials should be used for structural components, or those components should be protected by coatings which are resistant to radiation, as well as to acids and water. (3) In laying out equipment and piping, provision should be made for walkwayvs, stairs, and ladders to provide convenient access during con- struction without having to elimb on the process and service piping. These walkways, ladders, and stairs may or may not be removable after com- pletion of construetion. (4) Many items installed inside the reactor cell of the HRE-2 might have been located outside the radioactive area. These include the cold traps, which could be located in one of the shielded waste-system compart- ments, the space coolers, and a number of other nonradioactive components. (5) In the IHIRT-2 the sampler 1s located adjacent to the cell shield, and it is necessary to remove the sample from the sampler and carry it to the analytical laboratory in a shielded carrier. The risk of contamination can be greatly reduced i the hot laboratory is built around the sampler so that samples may be analyzed without removal from the shield; however, this would add to the cost of the facility. (6) Access through the containment seal, which consists of welded pans, can be made easier through the use of mechanical seals with organic gaskets. These would not be subject to severe radiation damage since the coutain- ment seal 1s itself shielded from primary radiation. (7) Clontrol of all operations affecting the reactor should be located in one central control room. This includes the controls for the chemical processing system, as well as those for the reactor system and all main con- trols for the steam turbine and steam auxiharies. This does not mean, how- ever, that all instrumentation must be in the control room, but only those things which are necessary to keep the reactor operator informed and to provide him with primary control of the operation. (8) In the HRE-2 refrigeration system a petroleum distillate, “AMSCO 125-82," was circulated inside the reactor cell as a secondary coolant be- cause no primary low-pressure refrigerant could be found which would not 7-5] THE LOS ALAMOS POWER REACTOR EXPERIMENTS 397 decompose under irradiation to release chlorides or fluorides detrimental to the stainless steel piping. Carbon dioxide was not selected because the piping would have to be designed to withstand about 1500 psi. A solution to this problem would be to use copper for all tubing and equipment in con- tact with the refrigerant, in which case I'reon could be used inside the cell. 7-4.12 HRE-2 construction costs. These costs are summarized in Ta- ble 7-7, including that portion of the costs due to the requirement for pre- venting the possible escape of radioactive material from the reactor to the building or surroundings. 7-5. Tur Los Aramos Powrr REACTOR INXPERIMENTS (Larre I anp 2) [34, 35]* 7-5.1 Introduction. Homogeneous reactors fueled with uranium oxide dissolved in concentrated aqueous solutions of phosphoric acid have a number of advantages compared with dilute aqueous uranyl solutions for certain power reactor applications. These applications are based on the fact that concentrated phosphoric acid solutions have high thermal sta- bility and low vapor pressures. This makes it possible to operate at rela- tively high temperatures without creating the excessive pressures en- countered with dilute agueous solutions. These temperatures are high enough to take advantage of the back reaction for recombination of gases produced by radiolytic decomposition and to eliminate the need for an external gas system or an internal catalyst. A phosphorie acid solution reactor, therefore, can be a sealed single vessel with no external compo- nents except a circulator for the steam system. In addition, the high hydrogen density and the relatively low neutron-absorption cross sections of the phosphoric acid system permit the construction of small, compact reactors with a low inventory of fuel, which may be ideal for remote package power plants. Other advantages include the possibility of continu- ous removal of fission-product, poisons and a strong negative temperature coefficient of reactivity. Although the advantages of uranium oxide in concentrated phosphoric acid solutions present a strong case for such a fuel system, the disadvan- tages, though few in number, are ponderous. In fact, one disadvantage, the highly corrosive nature of the phosphoric acid to most metals, may well outweigh all of the advantages. The only metals known to be suitably re- sistant to phosphorie acid in the coneentration and temperature ranges of interest are gold and platinum. Although with proper design the cost of using such materials in a reactor can be kept within reason, the problem of providing and maintaining an impervious noble-metal barrier between the *Based on material prepared by members of the Los Alamos Scientific Labora- tories. 398 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [CHAP. 7 TaBLE 7-7 HRE-2 CoxstrucTioN Cost* (To MAY 1056) Reactor (HRE-2): Design Reactor $549,000 Structure and services 185,000 Instrumentation 126,000 Total design cost (labor OH @ 659) $860,000 Reactor installation Building modification 58,000 Reactor cell and shielding 375,000 Reactor components and piping 1,400,000 Reactor controls (instrumentation and valves) 470,000 Reactor steam system and cooling system 161,000 Miscellaneous service piping 69,000 Maintenance tools 54,000 Waste and off-gas system 78,000 Miscellaneous and spare parts 208,000 Division supervisory labor 120,000 Total installation (labor and material) $2,993,000 Total reactor cost $3,853,060 Containment cost estimate for HRE-2: 1. Additions to shield design, required only to meet containment specification $75,000 2. Cost of sealing conduits and wiring 7,000 3. Cost of leak-tight closures on process service lines 23,000 4. Blast shields on reactor components Pressure-vessel $35,000 Two steam generators 10,000 45,000 5. Radiation monitoring of service lines 10,000 Total cost $160,000 *Excluding $800,000 cost of supporting research and development labor. tBuilding housed HRE-1—initial cost, $300,000. 7-5] THE LOS ALAMOS POWER REACTOR EXPERIMENTS 399 TaBLE 7-8 ProrerTiES OF LAPRFE-1 aAnp LAPRE-2 FurL SYSTEMS LAPRE-1 LAPRE-2 Type of uranium oxide U0s U0, Weight percent H3PPO4 53 95 Vapor pressure, psig 250°C 400 190 Vapor pressure, psig 350°C 1800 300 Vapor pressure, psig 450°C 4500 800 Overpressure gas 02 H, U235 concentration, g/liter 111 77 H concentration, moles/liter (25°C) 90 62 H /U235 190 190 Fuel volume ratio (430°C to 25°C) 1.33 1.16 Temperature coefficient of i reactivity (per °C) —8X 1074 —5X 1074 fuel solution and all structural metals with which it might come in contact is extremely difficult to solve. This problem was not solved in the first ex- perimental reactor to use phosphoric acid system, the LAPRE-1; however, considerable improvements in fabrication and testing techniques have been developed for use in the construction of LAPRE-2. Of the wide range of possible combinations of uranium oxide, phosphoric acid, and water, two systems appear to be of special interest. The first, which consists of UOj3 dissolved in an aqueous solution containing between 30 and 60 w/o0 of phosphoric acid and pressurized with an oxygen over- pressure, was used for the first Los Alamos Power Reactor Experiment (LAPRE-1). The properties of this solution are given in Chapter 3 and summarized in Table 7-8. The vapor pressure of this solution at the design operating temperature of LAPRE-1 (430°C) is 3600 psi. The second solution (to be used in LAPRE-2) consists of UO32 dissolved m a 95 w/o phosphoric acid and pressurized with a hydrogen over- pressure. As seen from Table 7-1, the vapor pressure of this solution at 450°C is only about 800 psi. Because of its reducing properties, the cor- rosive behavior of this solution is such that anything below hydrogen in the electromotive series is only slowly attacked. Since the attack rate is pro- portional to the position of the metals in the series, possible materials in decreasing order of usefulness are platinum, gold, carbon, silver, and copper. Although neither silver nor copper is attacked at a rapid rate, both metals undergo mass transfer, and suitable means for inhibiting this must be found before they can be considered practical for lining the reactor. 400 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHaP. 7 Solution Steam Discharge Line Storage Emergency —iL Dump Tank Transformer Feed Water Solution Pump Transfer Pump & Line Exhaust & Vent Valve Control Auxiliary Equipment Room Control Console - Y : — And Racks Reactor ! m J L Vessel 1 Lead and Water Shield Walls Fia. 7-23. Plan view of LAPRE-1 reactor area (courtesy of Los Alamos Scien- tific Laboratories). 7-5.2 Description of LAPRE-1. The first experimental reactor, LAPRE-1, was housed in a cell at the Los Alamos site which had been built previously to handle highly radioactive materials. The cell was modi- fied by supplementing the shielding with additional concrete on one side and lead on the other sides and ceiling. A stainless-steel wall was also put across the reactor end of the cell to permit filling with water for a neutron shield. Equipment such as the solution transfer pump and sampling system were located in the cell outside the water-filled portion, while other auxili- aries were located outside the cell. These included a 12 gpm, 4000-psi feed- water pump, three 6-in.-diameter, 32-ft-long underground solution meter- ing and storage tanks, and an emergency dump tank. Valve controls were located on the wall of the cell. Figure 7-23 shows a plan view of the LAPRE-1 reactor area. Figure 7-24 shows an artist’s sketch of LAPRE-1 in which the essential features of the reactor are indicated. As will be seen from this sketch, the 15-in.-ID> eylindrical reactor contains a critical zone above the cooling coils and a reservoir region below. This latter region, which contains a hollow cylinder of boron to prevent eriticality, provided a storage place for excess fuel and avoided the necessity for transferring the highly radioactive fuel solution in and out of the reactor as its temperature changed. The 7-5] THE LOS ALAMOS POWER REACTOR EXPERIMENTS 401 4.~ Steam Qutlet g Control Baffle Critical Region ’ P MHeoters & i Insulation Heat change: Boron Pump " Impeiler E , . o " Fic. 7-24. Los Alamos Power Reactor Experiment No. 1 (courtesy of the Los Alamos Scientific Laboratories). large negative temperature coefficient of reactivity made it impossible to compensate for the excess reactivity of the cold critical reactor and it was not considered practical to change the fuel concentration as is done in the HRIE-2. In the LAPRE-1, shim control was achieved by means of the thermal expansion of the fuel. This was accomplished by initially filling the reactor with the cold phosphoric acid fuel solution to a level about 7 to 8 in. above the cooling coils. At this level the reactor would become critical on removal of the control rods, the solution would heat up, expand, and gradually fill the eritical zone. On cooling, the reverse process took place, making the reactor self-compensating. The reactor vessel was so designed that at 430°C, the operating temperature, the fluid level would reach a point Just above the flow baffle. In the actual experiment the reactor was maintained at zero power while the temperature was raised 402 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 Water Steam Manifold Mqifcid Heat Exchanger Graphite # Reflecto Platinum Draft Tube {Critical Region) ®._ BeO Source .- Startup Source Reflector Shim - Fue! Transfer Fuel ! tine Storage § . i Tank Fic. 7-25. Los Alamos Power Reactor Experiment No. 2 (courtesy of the Los Alamos Scientific Laboratories). to 240°C by using electrical heaters located around the exterior of the vessel. This was done to minimize the positive rcactivity change resulting from starting up the fuel circulating pump, which caused a rapid injection of lower temperature fuel solution from the reservoir into the core. Heat removal from LAPRIE-1 was accomplished by circulating the fuel solution at 600 gpm over the cooling coils by means of the centrifugal sealed rotor pump shown in Fig. 7-24. The 12-gpm feed-water stream, heated to 800°F in passing through the cooling coils, was simply discharged to the atmosphere through a quick-closing throttling valve and steam silencer as a means of dumping of heat. The possibility of fisston products escaping through the steam discharge line was prevented by installing a 70-sec holdup line, suitably monitored for radioactivity, between the reactor and throttling valve. In case of accidental stoppage of the pump during full-power operation, the reactor could be shut down with the control rods. Provision for dis- charging the fuel solution by hand or through a rupture disk was included for added safety. The entire LAPRE-1 vessel and cell components exposed to fuel solution were gold-plated, with the exception of the rod thimbles and draft funnel which were made of, or coated with, platinum to minimize neutron ab- 7-5] THE LOS ALAMOS POWER REACTOR EXPERIMENTS 403 TaBLe 7-9 CHARACTERISTICS 0F Lo0S ALAMOS PoweEr ReacTor EXPERIMENTS ! LAPRE-1 LAPRE-2 Power level, Mw heat 20 0.8 U235 in core, kg 4.1 4.1 U235 total, kg 8.5 6.5 Fuel temperature, °C 430 430 Truel pressure, psig 4000 700 Steam temperature, °F 800 600 Steam pressure, psig 1800 600 Cost of components $250,000 $120,000 sorption. The 22 stainless steel cooling coils, 50 ft long, were clad with 6 mils of gold. The characteristics of LAPRE-1 and -2 are summarized in Table 7-9. 7-5.3 Operation of LAPRE-1. Final tests on the LAPRE-1 system were made with a 0.51 47 TO3 in 7.25 A H3PO4 fuel solution. Data were obtained at room temperature in terms of control-rod position at delayed critical versus volume of fuel injected into the system,* and results were interpreted in terms of a simplified calculational model to obtain control rod worths. Tor the five control rods, four located on a 31%-in. radius and one central rod, measurements yielded a total worth of 6.39,. The latter results were in good agreement with period measurements at cold critical. Also inferred from the data was an effective delayed neutron fraction of 0.0001. With the reactor filled to a predetermined level at room temperature, initial heating of the system was achieved by means of electrical heaters disposed around the outside of the vessel. Heat was applied until a core temperature of 240°C was attained; heating by nuclear power was then initiated. With nuclear power a core temperature of 340°C was achieved, at which point the circulator was turned on. Only during the forced con- vection phase of the operation was a uniform temperature distribution established in the fuel solution; previonus to starting the circulator, the *Detailed description of the nuclear data obtained during the test is set forth in “Control Rod Worths vs. Temperature in LAPRE-1" by B. M. Carmichael and M. E. Battat, TID-7532 (Pt. 1), p. 125, US.A.E.C., Technical Information Service Lxtension, Oak Ridge, Tenn. 404 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 averange core temperature wus higher than the corresponding reservoir temperature. Data were obtained during the course of the experiment in terms of control rod positions versus core temperature at delayed critical. For the operation without the circulator, thermocouples located at various points in the system provided data from which average core and reservoir tem- peratures could be estimated. With the circulator in use, the data for control rod position versus temperature were used to infer the temperature of the solution level at the top of the baffle. Above this point a further increase in height did not contribute to the reactivity; hence the system was solely governed by the temperature cocfficient of reactivity. Based on the caleulated negative temperature coefficient for the system (~8 X 10~ 1/°C") it was found that the rods were worth 30 to 40% more at 390°C than at cold eritical. A maximum temperature of 390°C at an operating power of about 150 kw was achieved in the experiment. No steam data were obtained during the test because of a failure in the heat-exchanger assembly after a number of hours of operation. Inspection of the assembly after a cooling-off period indicated that a rupture in the gold cladding of two of the lead-in tubes was responsible for the failure. This rupture was probably due to a bonding between the gold cladding of the tubes and the gold plate of the hallle, in conjunction with vibration of the tubes resulting from the opera- tion of the circulator. Although no further tests with LAPRIS-1 were conducted after October 1956, careful inspection of the reactor parts after disassembly provided information valuable to the construction of LAPRI-2. 7-5.4 Description of LAPRE-2. The second power reactor experiment, LAPRE=2, was constructed in an underground steel tank 50 in. in di- ameter and 20 ft long located at a site approximately 100 feet from LAPRI-1. A sketch of the reactor is shown in Ilig. 7-25. Because of the lower vapor pressure of the LAPRE-2 fuel solution the design and method of controlling LAPRI-2 differ from LAPRE-1. As shown n the figure, the reactor vessel has a simple autoclave shape with relatively thin (5/8-inch) wulls, This compares with the 3-in.-thick vessel required for LAPRE-1. The cooling coils in LAPRE-2 are located above the reactor core, which has the advantage of providing a noneritical region for excess liquid in case of overfilling. Cooling is accomplished by natural convee- tion circulation of fuel solution over the cooling coils. Fuel circulation is aided by an mverted platinum cone in the core region. A 1-ft-thick graphite refiector surrounds the vessel. The inner 6 in. of this reflector can be moved slowly to adjust the reactivity at varying fuel concentrations. Cold eriticality is achieved by slowly filling the reactor with fuel solution 7-5] THE LOS ALAMOS POWER REACTOR EXPERIMENTS 105 from a water-cooled copper tank pressurized with hydrogen. The line connecting the tank and reactor is gold clad (0.010 in.). After the reactor becomes critical with cold fuel solution, further additions are made until the desired operating temperature of 430°C is reached. A separate pancake- shaped cooling coil, located just above the main tube bundle, serves as a level indicator. A thermocouple in the discharge of this coil indicates whether there is power removal and thus if the coil is immersed. All parts of LAPRE-2 are completely clad with fifteenn mils of gold to provide 1009, protection of the structural metal from the fuel solution. This gold, as well as the platinum and noble metals used in LAPRIS-I, are recoverable after conclusion of the reactor experiments. The 600-psi, 600°F steam produced in the LAPRE-2 is of somewhat lower quality in terms of power production than LAPRIE~1 because of the lower operating pressures. However, at a design level of 1.3 Mw heat, the reactor could produce 250 kw of electricity which is acceptable for a remote power station. The characteristics of LAPRE-2 are summarized in Table 7-9. Con- ~truction of the reactor proceeded through 1957 and was essentially com- pleted i early 1958. 406 EXPERIMENTAL REACTOR DESIGN AND CONSTRUCTION [cHAP. 7 REFERENCES . A. M. Wrinserag, Whither Reactor Development, Inter-American Con- ference, Brookhaven National Laboratory, May 1957, Nucleonics 15(8), 99 (August 1957). 2. C. P. Baker et al.,, Waier Botler, USAEC Report AECD-3063, Los Alamos Scientific Laboratory, Sept. 4, 1944, 3. F. L. BentzeN et al., High-power Waler Botler, USAEC Report AECD- 3065, Los Alamos Scientific Laboratory, Sept. 19, 1945, 4. L. D. P. King, Design and Description of Water Boiler Reactors, in Pro- ceedings of the International Conference on the Peaceful Uses of Atomic Energy, Vol. 2. New York: United Nations, 1956. (p. 372) 5. C. K. Beck et al., Nuclear Reactor Project at North Carolina State College, Nucleonies 7(6), 5-14 (1950); and Further Design Features of the Nuclear Reactor at North Carolina State College, USAEC Report ALCU-1986, North Carolina State College School of Engineering, January 1952. 6. R. N. LyoN, Preliminary Report on the 1953 Los Alamos Boiling Reactor Experiments, USAEC Report CF-53-11-210, Oak Ridge National Laboratory, Nov. 30, 1953. 7. J. R. Dietricu et al.,, Design and Operating Fxperience of a Prototype Boiling Water Power Reactor, in Proceedings of the International Conference on the Peaceful Uses of Atomic Energy, Vol. 3. New York: United Nations, 1956. (p. 56) 8. L. D. P. King, personal communication, January 1952. 9. Reactor News, Nucleonics 15(2), R10 (February 1957). 10. S. E. BeavL and C. E. WiaTers, The Homogenecous Reactor Experiment— A Chemical Engineering Pilot Plant, Chem. Eng. Progr. 50, 256-262 (May 1954). 11. C. E. Winters and A. M. WrINBERG, Homogeneous Reactor Experiment Feasibility Report, USAEC Report ORNL-730, Oak Ridge National Laboratory, July 6, 1950. 12. C. E. WintErs and A. M. WrINBERG, A Report on the Safety Aspects of the Homogeneous Reactor Experiment, USAEC Report ORNL-731, Osak Ridge National Laboratory, Aug. 29, 1950, 13. C. L. Secasgr and F. C. Zarr, HRE Design Manual, USAEC Report TID-10082, Oak Ridge National Laboratory, Nov. 18, 1952. 14. S. Visner and P. N, Havnexreicu, HRE Experiment on Internal Recom- bination of Gas with a Homogeneous Catalyst, in Nuclear Scr. Technol., USAEC Report TID-2506(Del.), Oak Ridge National Laboratory, 1955. (pp. 73-90) 15. R. B. Bricas and J. A. SwartouT, Aqueous Homogeneous Power Reactors, in Proceedings of the International Conference on the Peaceful Uses of Atomic Energy, Vol. 3. New York: United Nations, 1956 (p. 175). S. E. BEarL and S. Visner, Oak Ridge National Laboratory, 1955. Unpublished. 16. D. O. CampBELL, Decontamination of the Homogeneous Reactor Experi- ment, USAEC Report ORNL-1839, Oak Ridge National Laboratory, June 1956. 17. C. E. Winrters, Oak Ridge National Laboratory, 1952, Unpublished. 18. 5. E. BeaLn, The Homogeneous Reactor Test, in HRP Civilian Power Reactor Conference Held at Oak Ridge, March 21-22, 1956, USAEC Report T1D- 7524, Oak Ridge National Laboratory, March 1957. (pp. 11-47) REFERENCES 407 19. 8. E. Beary, The Homogencous Reactor Test, in HRP Civilian Power Reactor Conference Held at Oak Ridge National Laboratory, May 1-2, 1957, USALEC Report TID-7540, Oak Ridge National Laboratory, July 1957. (pp. 11-46) 20. 8. E. Beary, Homogeneous Reactor Experiment No. 2, 4th Annual Conference of the Atomic Industrial Forum, October 1957. 21. 3. E. BeaLy, in Homogeneous Reactor Project Quarterly Progress Report for the Period Ending Jan. 81, 1958, USAEC Report ORNL-2493, Oak Ridge National Laboratory, 1958, 22. I. Seiewak and H. L. FarkexBerry, The Homogeneous Reactor Test Mockup, in Proceedings of the Second Nuclear Engineering and Science Conference, Vol. 2, Advances in Nuclear Engineering. New York: Pergamon Press, 1957. (pp. 62-69) 23. D. S. Toowms, JR., Instrumentation and Controls for the HRT, Nucleonics 15(2), 48-52 (1957). 24, 8. E. Beawn and R. W. JurcenseEn, Direct Maintenance Practices for the Homogeneous Reactor Test, Fourth Nuclear Engineering and Science Con- ference, March 1958 (Paper 82); and USAEC Report CF-58-4-101, Oak Ridge National Laboratory, April 1958. 25. T. RockwerL LI, Reactor Shielding Destgn Manual. New York: D. Van Nostrand Co., Ine., 1956. 26. N, GrasstoxE, Principles of Nuclear Reactor Engineering. New York: D. Van Nostrand Co., Inc., 1955. 27. C. D. Borp and O. Sisman, How To Calculate Gamma Radiation Induced in Reactor Materials, Nucleonics 14(1), 46-50 (January 1956). 25, 5. GrassToNE, Principles of Nuclear Reactor Engineering. New York: D. Van Nostrand Co., Inc., 1955. 20. 8. E. Beavr, Containment Problems wn Aqueous Homogeneous Reactor Systems, USAEC Report ORNL-2091, Oak Ridge National Laboratory, Aug. K, 1956, 30. P. R. KasTEeN, Operational Safety of the Homogeneous Reactor Test, USAEC Report ORNL-2088, Oak Ridge National Laboratory, July 1956. 31. 8. E. Bearr and S. Visyer, Oak Ridge National Laboratory, 1955. Unpublished. 32. S. E. BeaLL and S. Visner, Oak Ridge National Laboratory, 1955. Unpublished. 33. P. M. Woon, A Study of the Possible Blast Effects from HRT Pressure Vessel Rupture, USALEC Report CF-54-12-100, Oak Ridge National Laboratory, Dec. 14, 1954. 34. D. Fromax et al., Los Alamos Power Reactor Experiments, in Proceedings of the International Conference on the Peaceful Uses of Atomic Energy, Vol. 3. New York: United Nations, 1956. (p. 283) 35. R. P. Hammonp and L. D. P. King, Los Alamos Homogeneous Reactor Program, in [T RP Ciwilian Power Reactor Conference Held at Oak Ridge, March 21-22, 1556, USAEC Report TID-7524, Los Alamos Scientific Laboratory, March 1957. (pp. 168-209)