B Reactor Museum AssociationB Reactor Museum AssociationAuthor: Michele S. Gerber, Ph.D., Facility Operations Division, Westinghouse Hanford Company, April 22, 1993 (approved for public release May 19, 1993)
From the restart of B Reactor in 1948, through its final closure in 1968, the pile's history was dominated by constant efforts to achieved increased power levels. By late 1956, the World War II power levels had more than tripled, and stood at 800 MW. At that time, B Reactor became the first at Hanford to achieve completion of Project CG-558, a thorough set of modifications and retrofittings designed to allow increased coolant flow to the pile, and thus attain higher power levels. Modifications were made to the pumping and piping systems, electrical systems, and various other components and fittings within the reactor. From these changes, HW operators originally hoped to realize an additional 1,650 - 2,350 MW (constant combined average) of production from B, C, D, DR, F, and H Reactors. However, the power level increases made possible by this project, by Project CG-600 at C-Reactor, and by fuel and tube design improvements far exceeded their expectations. In the 12 years following Project CG-558, other changes in procedures and equipment in the 100-B Area more than doubled the 1956 power level. In early 1964, the pile achieved a power level of 2,090 MW, after which time the AEC changed reactor operating limitations so that they were based not on MW levels but on the bulk temperature of the cooling water exiting the reactor. For the last three and one-half years of B-Pile's operations, this temperature limitation stood at 95°C (as compared to the World War II exit bulk temperature limitation of 65°C).
Questions concerning how to achieve higher power levels, and the consequences of such levels to reactor systems, had intrigued Hanford scientists since World War II. In 1945, HEW physicists speculated as to whether higher power levels would cause more corrosion of aluminum process tubes and fuel elements, more film formation, and/or more stress and deterioration of the masonite in the reactor's biological shields. They even wondered whether the uranium itself, in the fuel elements, would undergo destructive changes in its crystal lattice, thus becoming brittle and crumbly. However, their most crucial queries concerned the influence of increased radiation levels on reactor graphite. Would it swell to the point that it displaced the process channels and rendered the whole pile inoperable? Would the graphite itself "store" energy, and then release it in sudden temperature spikes as the carbon atoms in the crystal lattice of the graphite realigned themselves? Even though HEW scientists had recommended in principle a "cautious advance to high power levels" in July 1945, funding cutbacks that followed the war's end, and then contractor and federal agency changes in 1946 and 1947, prevented any action toward this goal until early 1949.
In April 1949, an incremental test program that would take D-Reactor to 330 MW was undertaken. According to HW management, D-Pile was chosen because, by that time, DR (D-Replacement) Reactor was nearly ready and available in case higher power levels caused irreparable damage to D-Pile. B Reactor, described as the "least distorted" (in terms of graphite expansion) of the three original production reactors, was still needed to produce Po-210. F-Reactor, commented Hanford management, was in the "worst shape of the World War II reactors, and was "the last pile on which any risk should be taken at present." By January 1950, the increased power level experiment at D-Reactor was so successful that it was being operated at 400 MW. By springtime, the AEC was conducting an overall review of the power level and goal exposure program, involving Hanford, Los Alamos National Laboratory, Oak Ridge National Laboratory, and even Canadian reactors at Chalk River, Ontario. Plans were being made to test whether B, D, and F Reactors could operate at 600 MW, without causing undue systems failures or increases in the plutonium 240 (Pu-240) content of the product. High Pu-240 concentration was considered undesirable, since the isotope was prone to produce premature and/or spontaneous weapons detonations.
Increased power levels themselves were easy to achieve, simply by adding enriched U slugs to flatten the reactor both front to back (axial flattening) and side to side (radial flattening). Hanford used three types of enriched elements: "E-metal" (also known as Ike slugs), that contained 1.75% U-235 by weight; "C-metal," a lightweight, low-density, uranium-aluminum slug that contained 4.3% U-235 by weight; and "J-metal" (also known as oralloy), an enriched, uranium-aluminum alloy that contained 7% U-235 by weight. Sometimes, because the enriched metals produced such high, localized heat levels, they would be charged in a "striped loading" that alternated natural and enriched uranium slugs. However, increased power levels presented many additional, puzzling operational challenges in the effects they imposed on reactor systems and components.
By mid-1951, HW scientists were conducting in-depth design reviews to determine the equipment changes that would be necessary to increase reactor power levels to the 600 MW range. They discussed three crucial, potential limitations to augmented power levels. Graphite temperature could prove to be an important limitation, but ongoing research at HW was demonstrating that carbon atoms in the graphite crystal, displaced by irradiation, could realign themselves as they became heated (hence more active) by the addition of larger quantities of carbon dioxide (CO2) into the reactor gas atmosphere. The CO2, stored in low pressure tanks outside the 115-B Building, increased the temperature in the graphite, thereby softening and contracting the embrittlement and swelling caused by irradiation. Such swelling, with the consequent bowing and breaking of process tubes, was especially worrisome to HW operators because tube replacement under such conditions involved "mining" (reaming out) the surrounding graphite. Repeated minings would weaken the mechanical strength of the core itself, thus posing an overall threat to pile life. During 1950 and 1951, many graphite core dust samples were taken usually after the removal of damaged process tubes. The results of these tests began to demonstrate the desired atomic realignments within the graphite, and provided real evidence that the CO2 annealing process worked.
By 1954, the CO2 additions were working so well that B Reactor operated with a gas atmosphere composed of 40% He and 60% CO2, and a full expansion/annealing cycle test was being planned there. Additionally, by taking solid core graphite samples and measuring the stored energy in them, and by applying other physics calculations, HW scientists learned that a sudden and spontaneous release of stored energy was not possible except upon complete loss of coolant. Another potential limiting factor to increased power levels was the substantially higher fuel cladding and tube corrosion rates (and failure rates) then being experienced in the HW reactors (see Sections 3.3 and 3.5). However, again Site operators had reason to hope that improved fuel fabrication methods, improved tube strength, and better understanding of water treatment chemistry would ameliorate these problems (see Sections 3.4 and 3.6).
The most crucial adjustments to allow for safe operation at higher power levels, HW scientists realized, would have to be made in the equipment that brought cooling water into and through the piles. They knew they would have to increase header pressure, in order to push more cooling water through the tubes. Consequently, they would need new, larger front face risers and crossheaders, along with new inflowing check valves, and larger pigtails, Parker fittings, and outlet nozzles. Additionally, new casings and impellers would be needed on the secondary process water pumps in the 190 Buildings, or larger, replacement pumps would have to be installed. All of these improvements would be needed to offset "boiling disease," the Hanford term for a situation wherein steam might form in a process tube. If this happened at higher power levels, greater header pressures would be needed to sweep the steam from the tube (and thus to prevent a localized meltdown). Also, at greater power levels, new Panellit gauges with a finer range of settings would be needed, and the reactor downcomers would have to be strengthened or replaced.
By mid-1953, the 105-B and 105-D downcomers, already operating at 20% - 50% above design capacity, were under intense study. An HW report on them, finished in October of that year, concluded that entrained air in the effluent caused them to vibrate violently. The higher exit water temperatures, and increased flows that would accompany higher power levels, could cause boiling in the exit crossheaders and/or downcomers of B and D Reactors, causing irregular water flow, surging and "water hammer," increased corrosion, and cavitation. Ultimately, they could tear loose, continued the study, and collapse inward due to the negative pressure that existed in them during operations. Such a collapse would send 30,000 to 40,000 gpm of effluent through the riser room and hence to the rear fuel storage basin. Radioactive water then would flood the experimental sample room, cushion corridor, monitor room, and control room, disabling the reactor. The study concluded "that these hazards are considerable, and that some changes in downcomer construction should be made at the B and D Piles." The installation of a downcomer monitor, as well as an alternate piping line coming off the top of the present downcomer was recommended for each of the two reactors.
Also by mid-1953, B Reactor operators realized that the filtration capacity in the 183-B Building would have to be increased, as the filters could handle only a maximum of 38,000 gpm during "difficult treatment periods" (i.e., times of high water turbidity). More important, however, was the need to increase pumping capacity in the 181-B, 182-B, and 190-B Buildings. The current pumping rate was up to 39,000 gpm, but the absolute hydraulic limit, they calculated, stood at 68,000 gpm. By 1954, they estimated that, in order to safely operate at power levels of 800-900 MW, the 100-B Area water system would have to provide for 71,000 gpm process water to the reactor, 18,000 gpm export water (sanitary water to the 200 Areas, supplied by 100-B Area pumping systems), and 4,000 gpm non-process water to service the non-process requirements of 105-B. To achieve such capacity, they proposed replacing with electrically powered units all of the steam turbines that then supplied about 40% of B Reactor's process water, and adding still more electric pumps. Also, in order to push the 71,000 gpm through the pile itself, they recommended replacing the orifice-type flow regulators in the process tube inlets with venturis, devices that would create less pressure drop. They also knew that check valves would be needed at both ends of each front face crossheader, and that the higher operating temperatures associated with increased power levels would require the installation of temperature monitoring devices in the biological shields, more sophisticated exit water temperature instrumentation, and the removal of the aluminum thimbles from the HCRs, VSRs, and the test holes.
In the meantime, as power levels crept upward in B Reactor during the late 1940's and early 1950's, fuel elements ruptures, feared since World War II, became a reality. The first slug rupture at the Hanford Works both occurred at F-Reactor in May 1948. However, the next two such failures took place in B-Pile, in September and November of 1948. In both of the latter cases, the failures were detected by increased activity in the beta-sensitive ionization chambers that sampled sectors of the effluent water, and the reactor was scrammed. The crossheader that had produced the high water activity reading was scanned with a radiation detection instrument to locate a deposit of active oxide in the rear face fittings. In each of the 1948 cases at B Reactor, this method easily located the process tube that contained the rupture. If it had not, according to HW procedures, pile operators would have taken a water sample from each tube in the crossheader. If the correct tube still could not be located, they would have restarted the pile to cause the rupture to deposit more radioactive material in the water, and then would have re-sampled each tube. Both the September and November fuel failures at B Reactor were severe ones that required replacement of the affected process tubes, rear pigtails, nozzles, and thermocouple lines. In the first of these ruptures, the pile was down for two days. Over 520 gallons of water escaped into the graphite, with removal requiring increased gas flow through the reactor over several days. In the second rupture at B-Pile, the entire end cap of the fuel element came off, making it one of the most severe failure incidents at early Hanford.
Overall, the number of fuel slug rupture incidents in Hanford's reactors increased slowly during 1949-1950, but expanded dramatically in 1951. That year, the Site experienced 115 fuel failures, 15 of which occurred in B Reactor in just five months (July through November). In the two July ruptures and the two August ruptures, the pile remained shut down for about 26 hours each time, causing production losses of 400-500 megawatt days (MWD) each. However, in September 1951, anxious to reduce the large production time losses resulting from the increasing numbers of ruptures, HW operators instituted a new procedure known as the "quickie" scram. This technique, a very fast replacement of a ruptured fuel rod that had not stuck in the process tube, had to be performed in under 28 minutes. Otherwise, the xenon-133 [correction: xenon-135] poison that built up gradually as the reactor operated would overwhelm (tamp down) the reactivity of the pile to the point where restart was impossible until the excess xenon-133 [correction: xenon-135] had decayed away (at least 24-26 hours). In the "quickie" procedure, none of the river pumps was shut down, but the crossheader that served the tube containing the ruptured slug simply was valved down. The failed element was displaced into the fuel storage basin, fresh U slugs were charged into the tube, and a "hot startup," involving pulling the control rods out to effect a rate of rise of 100-150 MW per minute, was accomplished. During the first month that the quickie procedure was instituted at HW, it worked in three of the five situations in which it was tried. These three quickies, two of which occurred at B Reactor, resulted in a combined loss of only 1.4 hours of production time and 200 MWD.
From September 1951 forward, the quickie procedure was attempted whenever a fuel element appeared to be ruptured but not stuck in B-Pile and in the other HW reactors. On the average, the technique worked in about half of the cases. For example, in October 1951, B Reactor experienced three fuel failures. The first was a quickie that required 28 minutes of down time and cost 7 MWD. The second, one of the most severe ever to occur at the pile, caused the tube to split and forced the gunbarrel out by 13 inches. Approximately 160 gallons of water leaked into the graphite, both the tube and gunbarrel had to be replaced, and the reactor was shut down for 79.8 hours. The third rupture that occurred in October 1951 also required tube removal, and cost 33 hours of downtime.
Still the number of fuel failures continued to climb. The opening month of 1952 witnessed five fuel failures in B Reactor, two of which were quickies and three of which required tube replacements and outages of more than 24 hours each. The severity of the ruptures covered a wide range, as indicated by rear pigtail readings on the affected tubes that ranged from 1-50 rads per hour (R/hr). All of the failed fuel elements were removed to the 111-B Building, by then known as the Radiometallurgy Facility, where they were cleaned in a 10% nitric acid solution and then examined. In 1954, two extremely severe ruptures, necessitating tube removals, occurred in high exposure thorium slugs (known as "Q-metal" or "10-66" material) being tested in B Reactor.
By this time, fuel fabrication improvements had become a major focus of study at HW. In April 1951, a program was instituted that brought more rigid adherence to process specifications and more rigorous inspections of welds and welding cycles, and machined pieces. Additionally, new development activities looked at redesigned end closures on the fuel jackets, thicker end caps, redesigned slugs with rounded or beveled ends, new methods of bonding slugs within jackets (including diffusion bonding, and a zinc-tin bonding process), downsizing or extending slugs while they were within their aluminum jackets, double canning, improved grain structure of the U (involving various types of heat pre-treatments), and the use of other metals or alloys for the jackets. By 1954, most of these ideas had been discarded, but the Hanford Site had switched from the Triple Dip to a Lead-Dip canning method in fuel fabrication, and was experimenting with "cored" fuel elements (i.e., those having a three-eighths-inch diameter central hole, with aluminum plugs or supports on either end). Another idea that looked promising at the time was a change in "can closure" methods from the standard fusion weld to a point pressure ("cold") weld, because no braze lines would exist for corrosion attack. Also, a method of mechanically bonding slugs was tried, wherein the U core was electrolytically etched to produce a roughened surface. Then, when the core was sized into the fuel jacket, aluminum could flow into the roughened U surface and increase the strength of the bond. Again, none of the fuel fabrication experiments that were new in 1954 were adopted for long term use at HW, except the lead-dip canning process. That process remained in use through 1971.
By mid-1955, Tru-line fuel elements were in use at Hanford. These ribless elements had one male and one female end each, configured to prevent cocking and buckling of charges within the fuel column. At the same time, the use of internally and externally cooled fuel elements (I&E slugs), those having a complete hole through the middle, end to end, was being discussed at least as a concept, as was the possibility of cladding fuel elements with zircalloy-2, an alloy containing zirconium and small amounts of nickel, tine, chromium, and iron. Despite these innovations, the planners of Project CG-558 in 1955 expressed the fear that slug rupture rates would increase by a factor 2.6 times for every extra 100 MW of power that they were about to achieve. In late 1956, as the big project was nearing completion at B Reactor, scientists worried that slug rupture rates, not water pumping capacity, would prove to be the overall limiting factor in production increases.
Along with fuel element failures, the operators of B Reactor had to contend with increasing levels of corrosion of process tubes as they continually raised the pile power levels in the late 1940's and mid-1950's. Excessive corrosion could lead to tube rupture, a condition that was detected when moisture monitors located in the reactor's rear gas plenum registered inordinate wetness levels. In 1951, pile engineers estimated that the process tubes would corrode at a rate of 0.29 mils per month when the exit water temperature was held at 69°C, but that the rate would increase to 0.54 mils per month (nearly a doubling) if the exit water temperature were allowed to reach 85°C. By 1954, the bulk exit water temperature at B Reactor stood at 92°C. By that time, HW scientists had tried inserting zinc extensions into the process tube nozzles, to try to prevent nozzle sticking and corrosion, and they had tried decreasing the amount of aluminum spacers used in the approach to the front nozzles. Both of these experiments failed, and the ruinous corrosion continued.
In 1952, four process tube leaks of over 150 gallons each occurred at B Reactor. That same year, HW scientists tried protective coatings to anodize the tubes against the hydrated aluminum oxide corrosion that formed on the outside surfaces as the result of galvanic action between the aluminum and the wet graphite. Because they knew that this corrosion also contained iron oxide, calcium carbonate, and other trace substances, they chose an "Alumilite"* coating whose main constituent was 15% sulfuric acid (H2SO4). First, the tubes were degreased in inhibited alkaline cleaner, then rinsed in cold water. The coating was applied electrolytically, at a concentration of 12 amperes per foot. After the application, the anodic films were sealed by boiling in either distilled water or a 5% potassium dichromate solution. However, initially good results with a test group of 200 tubes later proved disappointing, and the process was not adopted on a large scale at HW. The corrosion of process tubes continued to be a key issue at the atomic site however, and caused so many tubes to stick in the graphite channels at F-Reactor that several channels had to be capped off in 1953.
That same year, internal tube corrosion came under intense scrutiny. Site scientists theorized that such corrosion was caused by "cavitation" (the formation of unstable vapor bubbles created by higher water temperatures). Pressure from the collapsing bubbles, they surmised, caused tiny implosions that blasted at the internal tube surface and eventually formed holes. With so many issues and types of corrosion under study, and with each HW reactor needing about 200 tube replacements per year, the "P" Department in 1953 began using the 108-B Building, whose tritium separations mission was being transferred to the Savannah River Plant, as an irradiated process tube examination facility (see 1.3.17). The building became known as the Critical Components Examination Facility. It contained floor space for cleaning tube sections with the necessary 10% nitric acid solution prior to examination. Also, a lead-shielded cave in which to conduct investigations of "hot" metal was emplaced in the facility in about 1953. Along with the 189-D Building, the 108-B facility served in this capacity for more than a decade.
In order to reduce the ruinous corrosion and ruptures occurring in reactor process tubes and fuel elements, HW chemists in early 1951 began an intense review of influent water treatments. In 1948, they had reduced the sodium silicate addition to the cooling water to 2.5 ppm, and discontinued it completely in 1950. In 1951, they initiated tests at 100-F Area to determine whether commercial grade aluminum sulfate (alum, also known as paper maker's alum), would perform as a better coagulant in the filtration process than the ferrous sulfate compounds that had been used since World War II. By 1952, they had found that the alum worked well, especially when aided by activated silica (sodium silicate prepared by the addition of sulfuric acid). The main benefit of the alum/activated silica method was that it reduced the iron content of the process water, thus reducing film formation on the process tubes. The need for frequent reactor purging, with the accompanying buildup of film and purge particles on rear face piping, also was reduced. One disadvantage of the caustic alum was that it increased corrosion on the aluminum process tubes. However, the advantages were so overwhelming that the alum/activated silica method was adopted by HW Process Specifications in July 1952.
The 100-B Area did not actually convert from ferric sulfate to alum/activated silica coagulation until early 1955, as a part of Project CG-558. Equipment changes were few. The existing feed mechanisms in the 183-B Building had to be re-set, and the piping that carried the lime addition had to be re-routed because the lime was added at the filter outlet flume (i.e., after filtration) instead of at the raw water inlet line. After the modifications, lime flowed by gravity as a slurry from feeders, then across the sedimentation basins and filters to be introduced at the outlet flumes. A 150,000-gallon capacity, liquid alum storage tank (36-foot diameter, 20-foot high) was located above ground on the north side of the 183-B head house. The above ground acid and sodium silicate storage tanks that served the 183-B Building actually were located just outside the 183-C Building, where the activated silica solution was prepared. It then was hauled to B-Area in tanker trucks, and stored in a 35,000-gallon, above ground tank located outside the 183-B Building. (Later a pipeline was installed from the 183-C activated silica preparation facility to the storage tank outside 183-B.) The sulfuric acid for this process arrived at HW at a concentration of 93%, which was then diluted and mixed into a sodium silicate solution that contained 1.5% SiO2 by weight. The proportions of the mixture were 8.9% dilute H2SO4, 28.7% dilute SiO2, and 62.4% water. This solution then was aged for one hour, then further diluted to 1% SiO2 by weight (to prevent gelling of the silicate), and mixed with the raw bauxite that was the essential ingredient in alum. Together, these ingredients and the chlorine were added in the raw water flume that fed the process water basins. The exact mixture achieved at HW was the result of a trademarked process, and was known as Separan 2610.* It was an organic polyelectrolyte, and served to control the buildup of positive and negative ions (which were subject to activation) in the process water.
The interaction between pH and sodium dichromate (Na2Cr2O7) was another crucial aspect of water chemistry that was examined carefully in the early 1950's by HW scientists seeking to reduce tube and slug corrosion in the reactors. In January 1952 the new "Flow Laboratory" was completed at D-Area, and the first experiments there consisted of feeding water at six different pH levels into slug-loaded, mocked-up process tubes. These tests soon confirmed the knowledge originally gathered in World War II experiments that lower pH levels produced lower aluminum corrosion rates. However, the worry still existed that pH levels below 7.0 would cause the hexavalent Cr+6 (Na2Cr2O7 - the original corrosion control chemical) to reduce to trivalent Cr+3, thus increasing chromate film formation on the tubes. In April 1952, in a bold experiment, the addition of Na2Cr2O7 to reactor coolant water was stopped at HW. Quickly, a "severe pitting attack" beset pile process tubes and fuel elements. The damage from this attack was so severe in all of the reactors that the Site returned to the use of 2 ppm Na2Cr2O7 in mid-1953. A report on this experiment concluded: "Dichromate is not a completely satisfactory inhibitor, though it is far better than no inhibitor under present conditions."
Following this experiment, key questions for HW operators became how far they could lower the pH and still not cause a Cr+6 reduction to Cr+3, how much sodium dichromate could be tolerated by the aquatic life of the Columbia River and how much could be tolerated in the drinking water of downstream reactor areas and the 200 Areas export water. HW Process Specifications lowered the process water pH to the range of 7.2 - 7.4 in 1955, and to 7.0 in 1956. These changes were accomplished by adding more sulfuric acid to the influent water and lessening the lime addition. In the meantime, studies in the tolerance levels of salmon, trout, and other Columbia River species went forward at the 100-F Area fish laboratory. At the 10% concentration level in river water, HW studies found, there was significant total mortality in salmon. At the 5% level, there was significant fry mortality, and at the 3% concentration, there was significant growth retardation. By 1953, aquatic biologists estimated, the approximate sodium dichromate level in the Columbia (in the vicinity of the reactors) stood at 0.8%.
The 1946 U.S. Public Health Service (PHS) drinking water standard for hexavalent chromium (still in effect through the mid-1950's) stood at 0.05 ppm. However the form of Cr+6 in the sodium dichromate used at Hanford was equivalent to 0.7 ppm Cr+6. Therefore, the Hanford form actually carried an overall limit of 0.143 ppm in drinking water. A 1955 study conducted by scientists of the Irradiation Processing Department (IPD - successor to the "P" Department) concluded that, due to channeling of reactor effluent in the Columbia River, it was possible that some 100 Areas and 200 Areas drinking water could exceed the allowable limits for hexavalent chromium during periods of low river flow. As a result, a test that reduced the sodium dichromate addition to the cooling water of 105-D from 2 ppm to 0.5 ppm was conducted in mid-1956. This experiment found that the lower and higher proportions were equally effective as corrosion inhibitors except when process tubes were previously corroded or worn down so that the water between the bottom ribs became excessively hot. In the latter circumstances, 5 - 10 times as many slug ruptures and tube leaks occurred with the 0.5 ppm addition as with the 2 ppm addition. Discouraged, site scientists commented that "it may eventually become necessary to limit the amount of dichromate added to the river because of toxic effects on fish."
Other experiments to vary the influent water chemistry and constituents were conducted during the early 1950's at HW, all seeking to limit the destructive corrosion to reactor process tubes and fuel elements. The push to ever higher power levels continued, as did intense study of every variable associated with coolant flow. Turbidity (suspended solid particles and debris) in the Columbia's water abraded process tubes, causing pitting, corrosion, the formation of iron films, and increased levels of key isotopes of concern (especially Fe-59 and Mn-56). During 1953 and 1954, tests conducted in the 100-D Area Flow Laboratory, using unfiltered river water as reactor coolant, produced "severe pitting" of mockup process tubes and dummy fuel elements "resulting from corrosion-erosion." Another key variable in water treatment concerned the filter media themselves. Mid-1950's tests altered the gravel, sand and anthrafilt proportions in the filters, and found that the use of more anthrafilt and less sand allowed percolation rates above the standard 2.7 gpm/ft2. This information was incorporated into new filter designs that were installed as a part of Project CG-558. During these years, the new water treatment methods and experiments resulted in the use of some different chemicals and tests in the 1704-B process control and sampling laboratories. Somewhat higher proportions of nitric acids, potassium hydroxide, silver nitrate, and iron standard solutions were used. However, most of the sampling and testing chemicals remained the same.
The drive to higher and higher power levels in B Reactor throughout the late 1940's and mid-1950's was accompanied by the need for several safety modifications. One component of the World War II reactor design that was especially vulnerable to increased power levels was the third or "last ditch" safety system. The original arrangement, a large tank of boric acid solution held at the top of the pile ready to pour into the VSRs to shut down operations through neutron absorption in case of an accident that interrupted both the primary and secondary coolant flow, was not designed for the augmented power levels. By 1950, HW operators expressed the concern that, at higher operating temperatures, the boron liquid would "flash" to steam at initial contact with the hot aluminum thimbles that lined the VSRs. If this happened, they reasoned, the solution would boil and there might not be enough liquid left to shut down the pile. Furthermore, the vapor formed from the boron solution might rupture the thimbles, thus wetting the graphite. This risk was considered so severe that operators did not dare to test the third safety system at all after the summer of 1950. As a result, the liquid boron was replaced with 29 "ball hoppers" (one at the top of each VSR channel) that contained three-eighth-inch to seven-sixteenth-inch nickel-plated carbon steel balls). These balls, which also acted to shut down the pile through neutron absorption, would funnel down into the VSR channels via a step-plug assembly, in the event of an emergency or a test. Ball removal then would be accomplished by a vacuum system. In January 1952, B Reactor became the first to be fitted with the new "Ball-3X" system.
Additional safety modifications that were installed at B Reactor to accommodate the power level increases of the early and mid-1950's included pile shield restrainers in 1950, and thermocouples for the VSR thimbles (which were approaching their melting temperature) in 1951. The latter year also witnessed the adding of crossheader pressure monitoring equipment, downcomer repairs, and additional health monitoring equipment for the radiometallurgical examination facilities in the 111-B Building. In 1952, outlet temperature monitoring thermocouples were attached at the downstream ends process tubes of the process tubes in B-Pile, and earthquake detectors (called seismoscopes) also were installed. In 1954, the "rod tip length" (control portion) of the HCRs at B-Pile were increased by ten and one-half inches. The following year, continuous effluent water temperature monitoring equipment was added to the reactor, in order to protect against localized power surges and to keep the reactor within the closer temperature limits imposed by the augmented power levels. The latter modification consisted of resistance bulbs located in specially modified pigtails in each reactor zone, and were used in addition to the regular pile temperature monitors (thermocouples).
During the same years, many upgrades were made to the 107-B retention basin itself, and to the entire 100-B effluent disposal system. In 1948, the decision was made to provide a crib (called an "earth reservoir" at that time) next to the retention basin, for the holdup of unusual (high activity) effluent that was most commonly generated by slug ruptures and reactor purges. At the same time, the amounts of chlorine and calcium hypochlorite being added to the 107-B basin for algae control were increased significantly. The north side of the basin was repaired again in January 1951, and in March and in April 1952, and the portion of the effluent line at the outlet of the 107-B basin was replaced in May 1952. The downcomer at the exit to the reactor was repaired in October 1951, as the vertical baffle and several of the supports had once again cracked or torn loose. In December 1954, a gamma water monitor chamber was installed at the exit to the basin, and the 1904-B outfall line was expanded. A year later, an HW study on the "Columbia River aspects of increased production" concluded that the construction of any more single-pass reactors at the atomic site would limit the full potential of the existing reactors, because the river could not safely absorb all of the effluent that such reactors would generate. The radioisotope levels, chemical concentrations, and temperature increases being sent into the river, this study stated, were so burdensome that major changes in influent and/or effluent treatments, or conversion to a recirculating reactor cooling system, would be needed.
Another solution to the effluent disposal problem, first posed by site scientists in 1954, was the creation of an "inland lake system," wherein effluent would be diverted through a canal from each reactor to an artificial lake system on either side of Gable Mountain. After the routing and travel time had safely reduced the radioactivity and heat levels in the effluent, this proposal theorized, it could be discharged through another canal to a spillway on the river bank. Meanwhile, the partial ground percolation of the effluent that did occur near Gable Mountain, according to this idea, would form a below-ground "hydraulic dam" that would impede the northward movement of contaminated groundwater from the 200 Areas. However, an appraisal of the meteorological consequences of such a system, conducted in late 1955 by HW scientists, found many undesirable effects. Steam fog capable of producing stratus clouds would be formed, and radiation fog potentially could be generated, even on cloudless days, "over an extensive area," affecting plants and vehicular traffic. Radioactive frost, dew and rime would be deposited in the vicinity of the lakes and canals, pushing the loadings on electrical distribution systems, heating and ventilation facilities "beyond design" capacities. In view of these discouraging findings, the reactor effluent disposal problems at HW remained intransigent.
During the same time frame of the late 1940's and early 1950's that safety modifications were being emplaced to allow higher power levels in the Hanford reactors, the AEC's Advisory Committee for Reactor Safeguards (ACRS) was closely scrutinizing overall aspects of operations in the 100 Areas. In the late 1940's, the most pressing issue still was that of the potential discharge of stored energy in graphite, and careful measurements were taken of the reactivity transients and temperature transients within the reactor. However, by the 1950-1951 period, the increased addition of CO2 to the pile gas atmosphere caused annealing of the embrittled and stressed graphite, and thus proved to be a promising and feasible solution. Graphite "burnout" (oxidation) caused by excessive water leakage within the reactor, was considered to be the most serious potential threat to pile life, followed by shield deterioration and tube distortion caused by irradiation. As the graphite expanded and contracted, it contorted the shields, and squeezed and stressed gunbarrels, process tubes, nozzles, and Van Stone flanges. Additionally, the masonite within the biological shields deteriorated due to higher operating temperatures, and had to be continually monitored via a gold foil located in the reactor's Neutron Detection Facility. The relative normalized radioactivity of this foil was presumed to be proportional to neutron leakage through the shield.
By 1952, the ACRS was looking at the potential regional consequences of reactor accidents at higher power levels. They developed a formula to define the "hazard area" around each operating pile, and around the collective HW reactors. This formula defined such an area as a circle whose radius ("r" distance) was equal (in miles) to 0.01 times the square root of the reactor power level (expressed in kilowatts). The 100-B Area was soon to witness the startup of C-Reactor, located just south of 105-B and tied into several of the B-Area water and electrical systems (see Section 5.1). Using current power levels, the ACRS estimated that the "r" distance around the B/C Area would include most of the secondary and central control (buffer) zones located on the north slope of the Columbia River as well as "territory west of the reservation boundary." Higher power levels, the committee realized, would extend this "r" distance, and it insisted on careful scrutiny of all proposals to augment the current operating limits.
At its December 1954 meeting, as power levels continued to climb, the ACRS noted a "small margin between the effectiveness of the control and safety systems and the possible increases in reactivity" in Hanford's reactors. It recommended the formation of a "permanently organized reactor safety committee" at HW, and more studies involving the consequences of complete water loss. Further, it mandated actual experiments, later conducted in the 300 Area, to measure the airborne release (oxidation) of fission products that would accompany the melting of U fuel elements. It also funded offsite research into the development of "reactor fuses," or poisons that could be loaded in to permanently shut down and disable a reactor damaged by attack or earthquake.
Aside from modifications and studies made in connection with safety improvements at HW in the late 1940's and mid-1950's, many changes in B Reactor operations were made in the interests of augmented efficiency. While increasing power levels was the surest way to increase Pu production, Hanford operators also developed many efficiencies that raised total production output. One such change involved discontinuing the use of the 212 Buildings in the 200 North Area for storage of fuel rods after irradiation. As reactor throughput rose, the transport of more and more irradiated fuel rods often bottle-necked pile operations and posed safety issues. In the early 1950's, function of the 212 Buildings was phased out and the practice began of holding irradiated slugs in the reactor storage basins for their full decay period prior to shipment to 200 East and 200 West processing facilities.
At the reactors themselves, time operating efficiency (TOE), measured by total days minus outages and scrams, was the greatest factor that could be controlled or improved by finding quicker, less wasteful ways of performing maintenance and other tasks. By 1954, B Reactor averaged a TOE rating of 82.2%, or about 8,400 - 8,500 operating hours (about 300 days) per year. Another measure of production effectiveness was the level operating efficiency (LOE) rating, an appraisal of equilibrium and non-equilibrium losses. The former referred to the difference between optimum power levels and actual power levels attained while operating. Such losses could be minimized by better flattening efficiency, reactivity transient control, and the achievement of more reactivity in the pile's fringe zones. Non-equilibrium losses were defined as the difference between actual production during startups and the optimum production that would have occurred if the reactor had returned instantaneously to pre-shutdown power levels. Methods to reduce such losses involved minimizing long, "cold" shutdowns, and pushing the rate of rise in the reactor power levels during startups to the maximum allowable HW limit (150 MW/min).
One of the earliest proposals devised by the HW operators to improve time efficiency losses in three reactors was to shorten the time required for charge-discharge operations. They quickly realized the efficiency potential inherent in "segmented discharge," the practice of discharging only the downstream and central U charges in any given process tube and then simply pushing the upstream charges further back in the tube to allow for additional irradiation. This idea was conceived in the late 1940's as a way of achieving optimal goal exposure for fuel charges other than those located in the central reactor zone. In the absence of good flattening efficiency, the upstream charges could receive extra irradiation during a second period in the downstream portion of the tube. Additionally, pile operators also saw segmented discharge as a way to allow them to work on more than one adjacent horizontal row of tubes at once, during "C-D" operations. To achieve both desired effects, HW operators modified the then-current pneumatic C-D equipment (known by workers as the "clam shell") to perform segmental discharge.
By the early 1950's, development tests were going forward at HW on three new methods to improve time losses during "C-D." One technique was to perform "C-D" operations without valving each crossheader. They simply would shut down the reactor, wait until graphite temperatures fell below 100°C, then reduce the total coolant flow to the pile to 2,000 gpm, remove the rear tube caps and perform "C-D." However, this practice was hardly more time efficient then the regular method. Another early 1950's system tried to perform "C-D" while the reactor was operating. It used a pressurized charging machine that maintained process tube contact at all times during the procedure and pushed in fuel elements using a ram that could increase or decrease tube pressure by as much as 50 psi. Such dramatic pressure changes often caused slugs to "backseat," or cock and buckle within the fuel column, thus greatly increasing rupture rates. Furthermore, the use of this system resulted in much spillage and splashing of effluent water on the reactor's rear face, piping and walls.
Finally, in the summer of 1952, a new system for "operational C-D" was first tested successfully in B Reactor. Known as the "postum" system this technique grew out of the occasional need to interject poison elements into the pile after startup, to control localized hot spots and/or to adjust to optimum rod configurations. Also, there sometimes was a need to discharge temporary poison slugs without shutting down the reactor. The postum system was operated remotely, and worked by flushing slug charges down the fuel column one-by-one. It consisted of a pressure-lock charging cylinder that worked in conjunction with special ball valves installed on the front and rear nozzles. An auxiliary, high pressure water addition system equalized the pressure between the charging machine magazine and the process tube before "C-D" operations began. The hydraulic characteristics then were maintained as needed throughout the procedure, in order to avoid the wide pressure swings inherent in earlier operational "C-D" methods. In system operation, the "C" machine first was coupled to the inlet ball valve of the tube being charged. Once the attachment was complete, the tube's Panellit gauge was bypassed electrically, and the auxiliary water supply began to operate. A flow valve in the special nozzle opened increasingly, allowing tube inlet pressure to rise to the necessary level. The rear ball valve then was opened, permitting discharge. Special radiation monitoring equipment at the reactor's rear face then scanned the tube, to make sure it was empty of irradiated charges. The rear ball valve then closed automatically, the auxiliary water system adjusted the pressure to the desired level, the "C" machine was activated, and new fuel was charged. Early in the new system's development, a special riser pipe that was to be used as a discharge chute was dropped from the design. Following that change, the operational poison "C-D" system was so successful that it soon became known as the Poison Column Control Facility (PCCF), and equipment was purchased for other HW reactors.
During 1953-1954, PCCF equipment began to be used for charging and discharging regular uranium slugs in production tests at B and KW reactors. Early problems included slug misalignment and slug damage as the result of impacts with other slugs during loading into the fuel column, higher radiation levels at the rear face that accompanied the discharge of very freshly irradiated elements, and heat distribution imbalances that could occur in central reactor zones as very active slugs were replaced by new "cold" ones. Additionally, fuel elements tended to "cock" (become uplifted on one side or on the upstream end) when they were "flush charged," as they were in operational "C-D." Cocked slugs were more prone to rupture than slugs that were properly seated in the fuel column, as the uplifted portion of them touched the tube wall and formed hot spots. Furthermore, the operational "C-D" equipment required increased maintenance time over regular "C-D" apparatus, and some HW operators mentioned a "psychological factor" involving fear that the procedure was unsafe. Still, the new method was attractive in that it saved much reactor outage time, as about 10% of the fuel elements in a reactor would be discharged every three to four weeks. Also, as the system began to be used for the charging and discharging of regular U fuel, a separate exit water monitoring device was installed on each affected process tube. Therefore, slug ruptures could be detected and flushed out faster, thus limiting their damage potential. A full-pile production test of operational "C-D" equipment was performed at C-Reactor in 1956 (see Section 5.11).
Another new technique tried to improve operating time efficiency at the early 1950's Hanford Works was the use of operating purges. Known as "hot" purges because they occurred while the reactor was running, the technique was tried first in the spring of 1951. Site scientists soon learned that hot purges were very effective in removing reactor films, but produced a five-fold increase in effluent activity over "cold" purges (those conducted while the reactor was shut down). Levels of phosphorus 32 (P-32), Fe-59, and copper 64 (Cu-64) were especially elevated. A mid-1952 study estimated that the conduct of two operating purges per week (Site-wide) would increase the potential dosage rate" of raw river water leaving the HW boundaries by 10% "over that delivered by routine effluent." The dramatic increased in P-32 content, the radionuclide of highest concern in Columbia River fish, the document continued, made operating purges "a possibly undesirable procedure." Furthermore, the irradiated diatomaceous earth slurry could collect and concentrate activity levels on the filter media of downstream reactors and nearby cities using Columbia River water. Due to its particulate composition, according to Hanford's chief health physicist, the effluent slurry might also have a higher biological absorption ratio for aquatic life. Further study of the effects of the new hot purges was recommended.
Throughout the period 1953 through 1955, several different agents were tried in the conduct of operating purges at HW. The concept remained a high priority to site operators, since it achieved impressive increases in purge efficiency, thereby saving time and money. Experiments were conducted using variations in diatomaceous earth composition, and chromic acid. However, the World War II Super-Cel proved to be the agent of choice, except for small specialized purges to reduce zinc-65 (Zn-65) activity on rear face piping just prior to certain maintenance activities. Knowing the effectiveness of Super-Cel in defilming reactor process tubes, HW scientists conducted a thorough review of its effluent and river effects in 1956. They found that the greatest augmentations in effluent activity levels occurred in the non-colloidal solids within the exiting water. Up to 69% of the gross beta activity in purge effluent was associated with the solids and suspended particulates, including 76.3% of the total P-32 activity in the effluent, 98% of the Fe-59, 74% of the Cu-64, 90.7% of the Zn-65, and 55.8% of the arsenic 76 (As-76) -- the isotope considered by the site scientists to contribute the greatest portion of the gastrointestinal tract dose delivered by Columbia River water downstream of the Hanford reactors. Furthermore, this discouraging review determined that less than one percent of the solids dissolved in river water, after agitation up to 16 hours. In view of the continuing desire of reactor operators to use this very effective purging method however, compromises were made. Restrictions were placed on the frequency of purges that could be conducted during autumn periods of low river flow, and a series of experiments designed to entrap the solids content of pile effluent in various filter media was initiated (see Section 4.13).
Another early-1950's innovation in HW operations that significantly improved reactor efficiency, in terms of decreasing equilibrium losses in this case, was the development of the poison splines flexible control system. First tested at B Reactor as early as 1949, the poison splines were long, thin flat, neutron-absorbing strips made of boron carbide (B4C) surrounded by 63-S-T6 aluminum (a stiffer aluminum alloy than the very soft 2-S aluminum that comprised the reactor process tubes). The splines were fabricated by filling a thin-walled (.014-inch thick), five-sixteenths-inch O.D. aluminum tube with the boron carbide grains sized between 30 - 60 mesh, and then passing the tube through a series of rolls to flatten it. The final spline thickness (boron carbide only) was .055-inch, and the total thickness with the aluminum walls was .075-inch. The splines were inserted into ribbed process tubes below the uranium charge, between the two bottom ribs of the tube. Their purpose was to provide additional, flexible reactivity control during operations, and temporary control during outages. Their use increased plutonium production in several ways. They could be used to eliminate hot spots in the reactor, where individual tubes operated at their maximum power level and thus placed maximum power level restrictions on the entire reactor. They also removed the need to use entire process tubes as poison "columns" (locations for rows of poison and dummy slugs), thus freeing those tubes to hold uranium charges. And, the splines could be withdrawn during startups and regular operations, thus eliminating the need to shut down the entire pile to discharge temporary poison columns.
Since the poison splines were to be withdrawn and emplaced during operations, some reactor modifications were necessary to permit their use. A slot for spline insertion had to be cut in each process tube nozzle cap, and then a seal and backing plate had to be inserted. The seal was extremely important, as the splines would be inserted during operations with the full force of approximately 19 gpm of water flowing through each process tube. The material finally selected for the seals was a plasticized vinyl called Arco-flex "B" Molding Compound*, which was shaped as a cylinder that fit into the endcap of the tube nozzle. The seals each were bonded against an aluminum backing plate, and each seal had a flanged edge that served as the nozzle cap seal. Each seal also contained a slot .450-inch wide and 0.75-inch high. Insertion into the process tubes required the use of a water-soluble lubricating oil, but no oil was needed during withdrawal of the splines, as coolant flow through the tubes provided sufficient lubrication. Once a seal was emplaced in a tube, a short section of spline was left in at all times, even when the full-length spline was not in use, in order to prevent water leakage. During use in the reactor, the splines became highly radioactive, with the beta emitting activity in the aluminum being more significant than the alpha emitting activity of the boron. Splines were not re-used at HW, but were chopped into short lengths (about 1.5-inches) as they were withdrawn from the pile, dropped into shielded casks, and buried in the solid waste burial grounds near the reactor.
When poison splines began to be used in the HW reactors, several questions arose involving safety and reactivity. Would the three gpm coolant flow reduction per tube result in insufficient cooling of the uranium fuel charges? Would the splines cause excessive thermal gradients within the slugs, thereby resulting in increased slug rupture rates? And would neutrons stream out of the new gap between the front (upstream) shielding slug and the tube end cap, thereby contaminating the reactor front face area? Many tests were performed to answer these questions during 1955, some of these at B Reactor. Ultimately, it was discovered that the reduced coolant flow caused no problems, and a tapered shielding slug was developed for the far upstream position in the process tube. The question of neutron flux perturbations in the uranium fuel changes was more complicated. Yes, reactor physicists determined, neutron flux became very asymmetrical in the presence of the boron-carbide splines, but the numbers of slug ruptures did not seem to increase and the end result of the flux irregularities became moot.
Aside from modifications undertaken to increase safety or efficiency, other equipment replacements and additions occurred at B Reactor prior to Project CG-558. The HCR hydraulic shim pump drive failed at B-Pile in September 1949, because oil had leaked from the motor into the pumps. The pumps were replaced that autumn. At the same time, consideration was given to replacing all of the reactor's rear face stainless steel nozzles with aluminum nozzles, and to galvanizing the front face nozzles to prevent corrosion. However, that project was not funded. In January 1950, many of the Panellit gauges on B-Pile had to be replaced, as film accumulation in the fringe tubes had increased the back pressure on the gauges to the failure point. Additionally, many production tests were carried out during 1951-1954, to determine the efficiency and operational feasibility of various orifice sizes and designs (including a double orifice assembly), and other flow restriction and measurement devices. The object of these tests was to find ways of reducing the pressure loss at the flow measurement (restriction) point, and thus of maintaining more pressure, with no additional pumping, from the inlet through the outlet end of the process tubes. Eventually, the knowledge gained in these tests was incorporated into the designs for Project CG-558. In 1954, molded washer seals were emplaced at the HCR channel openings in B-Pile, as the older sphincter boot seals had deteriorated due to graphite distortion and the effects of irradiation. That same year, automatic fire detection, sprinkler, and alarm signaling systems were installed in the 100-B Area shops and warehouses.
During the years between B-Pile's restart in 1948 and the major 1956 outage required to finish the Project CG-558 modifications, very few serious incidents occurred involving radioactive contamination or high doses received by workers. The most severe such event took place on December 22, 1952, when irradiated fuel elements in tubes being supplied by one crossheader overheated due to insufficient cooling water flow during a shutdown period. When the condition was discovered, the coolant flow was increased. However, the coolant flashed to steam when it contacted the hot fuel charges, and forcibly ejected seven irradiated slugs into the discharge area. Some of the elements lodged on the rear catwalks of the "D" area, and one "hung up" in a rear pigtail. The radiation field in the rear area reached between 10-20 R/hr, and five employees received overexposures while retrieving the fuel to the storage basin. Most of the other incidents in the 100-B Area were much less serious, and involved simple spills of radioactive material and equipment as transport or excavations were taking place throughout the area. At least one known 100-B burial ground fire occurred in June 1956.
Throughout the years prior to Project CG-558, special irradiation testing at B Reactor and the other HW reactors was held to a minimum. Therefore, unusual wastes from such tests also were minimal. On a regular basis, aside from plutonium and Po-210 production, B-Pile was used to manufacture thulium-170 and iridium-192, industrial isotopes used as tracers for processes both on and off the Hanford Site. In both of the latter cases, the isotope was produced from a target made of its own parent element, canned in aluminum. For a short period near the close of World War II, B Reactor also was used for a small, experimental irradiation of neptunium, for unknown purposes, and for the irradiation of barium in order to make radioactive lanthanum (Rala), a tracer gas utilized by the Air Force to measure the range of explosives. However, the Rala program soon moved to the Reactor Testing Station (now Idaho National Engineering Laboratory) near Twin Falls.
Most experimental work that was carried out in B Reactor consisted of production tests, designed specifically to improve some aspect of pile function. For example, a 1952-1953 test at B-Pile examined whether enlarged "free areas" in the inlet crossheader screens would significantly lessen pressure drop across the crossheader screen assemblies. No measurable effect was found. Another 1953 test measured the performance of HCRs fabricated using a new method of bonding the boron to the aluminum tubing. Use of reactor space for special testing was closely regulated by the AEC, and was allocated in terms on in-hours (ih - or time within the pile). For example, the 1955 allocation for special materials testing at the HW reactors was 210 ih, while 100 ih was allocated for special isotopes testing, and 20 ih was provided for general research. AEC policy specified that TOE and LOE losses due to shutdown time or power level reductions due to testing, and to scrams and unscheduled outages due to experiments, together with the 330 ih (combined testing) limit, could not exceed one percent of the total annual plutonium production level. At B Reactor, tests were conducted in the six test holes, labelled A through F, or in process tubes. Of the six test holes, three had diameters of three and one-half inches (with the original aluminum thimble linings), and three had diameters of one-half inch (with the thimbles). The access areas to these experimental holes, located on the right side of the pile, were designated as the "X-levels" of the reactor. Level X0 was at the main floor of the 105-B Building, level X1 was 15 feet above X0, and X2 was 15 feet above X1.
Additionally, special irradiation samples could be placed in the process tubes of B-Pile, in any of three types of containers. A sample can could be placed four to six inches between regular uranium charges (using dummies to achieve the desired spacing), or a "papoose" could be used for smaller samples needing high heat levels. The papoose was a short, threaded can that was attached to a regular uranium slug, locating the sample less than two inches from the hot uranium. Also, a "receptacle slug," a uranium cylinder bored in the middle to contain a sample that was completely surrounded by uranium, could be used for some tests. Within the test holes and process tubes, experimental conditions could be varied by altering the temperature and/or the composition of the gas atmosphere.
In the spring of 1949, a special pilot program for the separation and production of tritium (H-3) was moved from the Argonne National Laboratory (an AEC site in Illinois) to the 100-B Area at Hanford. The project was code-named P-10, as tritium gas was to be a key component in hydrogen (thermonuclear or "Super") weapons then under top secret development. Tritium was produced from irradiated lithium-aluminum (Li-Al) fuel targets, after previous experiments with lithium-fluoride slugs had caused "pile irradiation difficulties." The internal components of the Li-Al fuel targets were manufactured offsite, but they were jacketed in an unbonded aluminum-silicon can in HW's 313 Metal Fabrication Building. In the reactors, these slugs were surrounded by highly enriched uranium "driver" elements. In the P-10 program, irradiation took place usually in H-Reactor but sometimes in B Reactor. Then the fuel targets were moved to the 108-B Building and charged into a furnace with an inert atmosphere and a stainless steel furnace tube connected to a complex series of glass tubing and flasks fitted with palladium valves. As soon as the furnace was charged, the entire line was pumped down to an extremely high vacuum to remove impurities, and the furnace was outgassed to drive off absorbent gasses. At that point, the actual extraction began, based on the principle that extremely hot palladium allowed diffusion and passage of hydrogen but not of helium and other gasses. The furnace temperature was raised until all of the tritium gas had been driven from the irradiated targets, and the gas then was drawn down the glass lines, through palladium valves, and collected in shipping flasks. Toepler* pumps that used mercury to maintain and change pressure levels, were used to transfer gas within the system.
For the P-10 operation, the 108-B Chemical Pump House formerly used by HW's Power Division was retrofitted. The first floor contained a loading dock, melt room, machinery room, air conditioning equipment room, and miscellaneous other facilities. The second floor housed offices, an instrument repair shop, changing rooms, and a "cold line room," outfitted in late 1950, where unirradiated uranium was used in gas testing of slugs and in various other trials and procedures. The third floor contained the majority of the radiation danger zones, especially the can (fuel jacket) opening room, and the large "hood room" where the five actual process lines were located inside exhaust hoods. Additionally, the third floor encompassed the hood room operating gallery, the instrument development room, the mass spectrometer room, the emission spectrometer room, and the Health Instruments Division station. The fourth floor of the 108-B Building contained the exhaust air system, air filters and scrubbers, and much air monitoring equipment. The principal air monitoring device used in the P-10 program was the Kanne* chamber, an ionization chamber that used vibrating reeds to measure the number of ion pairs that were formed. The 104-B Building was built as a storage facility and, in early 1951, the 1703-B Pile Technology Service Building was constructed nearby. A 3,200-square foot, temporary construction frame structure on a concrete block foundation, the 1703-B Building housed offices, a conference room, and various storage closets when the 108-B Building itself became crowded with additional process equipment. In 1950, as the result of repeated H.I. concerns over the escape of tritium-oxide vapors and other airborne emissions, a 300-foot stack, as tall as the tallest powerhouse stacks at the Hanford Site, was built to replace the smaller, previous stack that served the 108-B Building.
During the latter half of 1950, the P-10 project expanded to encompass fuel slug fabrication facilities, and the delivery of metal line extraction facilities from offsite. The new metal process lines were installed in early 1951, and the existing glass lines were renovated. Hot development facilities aimed at scavenging or reprocessing tritium that had been contaminated with air also were added at that time. The scavenging operation worked by introducing the contaminated product into a bed of powdered palladium that had been cooled by liquid nitrogen, thus allowing it to adsorb large quantities of tritium. The bed atmosphere then was outgassed and heated, and the remaining gas driven off and passed through a palladium valve in a process similar to the original P-10 extraction method. Further expansions of the tritium separation facilities, including a new 109-B Building that would house stripping equipment, shipping container outgassing facilities, a cryogenic laboratory, and additional shops and technical development areas, were proposed as Project P-10-X in late 1950. Additionally, in the spring of 1951, the H.I. Division requested larger laboratory facilities as well as another monitoring station located close to the operating gallery in the 108-B Building. The entire project seemed poised for growth when the tritium mission was transferred to the new Savannah River Plant (an AEC Site in South Carolina) in 1952.
Throughout its history, the P-10 operation was plagued with contamination releases to the environment and to personnel. The primary release mechanism was lost product itself, which simply escaped as a gas. Estimates of such tritium releases vary from about 9,000 to 25,000 Ci, to the much higher 7.2 grams. The second largest source of contamination emitted from the P-10 operation was the mercury used in the Toepler pumps and pressure gauges. It is estimated that hundreds of gallons of contaminated mercury was disposed to the 108-B crib, with subsequent diffusion through surrounding soils and groundwater. Tritium contamination also was released via particles of broken glass and pieces of metal equipment, through the grease and oil that contacted process equipment, via liquid nitrogen that was used in the lost product reclamation mission and in the cold traps of the lines and leak detectors, and through the carbon tetrachloride and other decontaminating agents used to wash down equipment. Additionally, P-10 shipping casks sometimes spread external contamination along travel routes and at the 105-B Wash Pad where they were monitored. Preliminary estimates range from 20,000 to 50,000 Ci, to 16.1 grams buried as solid waste, and from 2,500 to 6,000 Ci, to 0.4 grams released as cribbed liquid tritium wastes. Organic and other chemical liquid waste volumes range into many thousands of gallons.
Early understandings of the entry and release mechanisms of tritium in and through the body, and of internal effects, were so incomplete that Hanford's H.I. Division formed a special P-10 Hazards Control Committee in the spring of 1950. The committee soon introduced 30 rats into the 108-B Building to test the consequences of tritium accumulation in the body. Pigs to test skin effects and plants to explore the effects of tritium on photosynthesis, also were brought into the facility, and further experiments were conducted in the 108-F Biology Laboratory. Quickly, it was learned that tritium, being a form of hydrogen, bonded quickly with the water-based systems of animals and plants, and was a ready carrier of contamination. Repeatedly, glassblowers and other workers at the P-10 facilities experienced over-exposures, and the H.I. Division focused intense efforts in the operations there. The P-10 Hazards Control Committee functioned for 17 months, and performed some of the earliest and most fundamental studies in the world on tritium effects. Still, in its final report, the group admitted to vast unknowns in its knowledge base: "The chief mechanisms through which personnel contamination occurs are still unknown...Much of the technique of control remains to be learned."
Finally, beginning in 1954 and continuing into November 1956, the major modifications of Project CG-558 were emplaced at B Reactor. At the 181-B River Pump House (by then called the 181-B/C structure; see Part V), the six existing 10,000 gpm pumps, along with two similar units obtained from the 181-D Pump House, were converted to 10,500 gpm units. The conversion was accomplished by re-winding the pump motors from 450 horsepower (HP) to 600 HP, and by converting the bowls. The majority of the river water that flowed from these pumps then was re-routed, via a new 48-inch line, so that it bypassed the filter reservoirs in the 182-B facility. (A small portion of the raw river water still was pumped to one side of the 182-B building, to provide export water for the 200 Areas and secondary backup cooling water for B Reactor.) The main bulk of the coolant water then flowed directly from the 181-B Pump House to the 183-B Filter Plant, where filtration capacity was increased by many methods. The openings from the flash mixing chambers to the distribution flume, and between compartments of the mixing chamber, were increased. A series of holes, two feet square, were drilled in the wall between the subsidence basin outlet flume and the filter influent flume, to increase the flow. The existing filter media themselves were changed, to consist of 27 inches of anthrafilt, 3 inches of sand, and 12 inches of gravel. Additionally, the orifice and baffle were removed from the filter effluent flume, and the clearwell water reservoir level was reduced by two feet to increase available filter bed head. All of these changes produced a filter flow increase to 240 percent of the original design capacity, i.e., from 2.6 to 6.0 gpm/ft2/min. The filter flow controls were re-set to operate at a flow rate of 71,000 gpm to the reactor, and a system was installed to convey filter effluent samples from each of the filters and the four flumes to the head house.
100-B Area's main process pumping system originally had supplied the primary process water as well as the auxiliary water systems in the area, including the high head (last ditch) tanks and the thermal loops for shield cooling. However, in Project CG-558, these auxiliary systems were taken off of the main pumps and supplied by a new, 4,000 gpm, motor-driven unit. The existing butterfly valves and orifices that were used to maintain sufficient head for the auxiliary systems were removed and bypassed, thus reducing the head requirements for the main pumps to 150-foot each. A large 190-B Annex was built, and the four original 190-B pumps that supplied water to the 190-B clearwells were replaced by new pumps capable of supplying 15,000 gpm and a 150-foot head capacity each. The existing drive motors were retained for these new pumps, and two supplementary motors, 700-HP each, were installed. The very old, primary, steam-driven pumps in the 190-B Building were retained in standby status for shutdown and emergency use. The existing 190-B pumps for the secondary process water were removed and replaced with eight, electric-driven pumps of 10,000 gpm capacity each against a 1,360-foot head. The new secondary pump drives operated at increased speeds, and included 4,500 HP motors and eight and one-half ton flywheels.
Many changes were instituted in the influent piping systems that led from the 190-B facility to the reactor building itself. The twelve existing, 12-inch lines from the 190-B Building to the reactor valve pit were supplemented by two, 18-inch carbon steel lines, with appropriate valves and strainers. All equipment that supplied solids feed (purging slurry) to B Reactor, excepting the 11,500-gallon mixing tanks, was replaced, including piping, transfer pumps, and injection pumps. Two new pumps were installed, capable of delivering 200 gpm of five percent diatomaceous earth slurry at a rated head of 640 psi. A power-operated, self-cleaning strainer was emplaced in the suction line of each pump. The main process coolant piping underwent the most major revisions. The four existing, 20-inch stainless steel lines connecting the main valve pit headers to the risers were replaced with two, 36-inch carbon steel lines, with all necessary valves and fittings. A 36-inch venturi tube was installed in each line to provide flow measurement for the automatic power calculator. The new venturis smoothed and distributed water flow much more efficiently than had the old in-line orifices. While the older measuring devices had decreased coolant pressure by as much as 40 percent, the venturis performed their function while reducing pressure by only about five percent.
Additionally, the two existing, 20-inch carbon steel main headers were replaced with a single, 36-inch carbon steel header. The four extant, 20-inch stainless steel risers were superseded by two, 36-inch carbon steel risers. And, while the old risers had a five-eighths-inch wall thickness, the new risers were one and one-fourth-inch thick. Five-inch check valves were installed at the end of the crossheaders, to prevent backflow through the inlet system in case of a pigtail failure. Previous to this innovation, a single pigtail failure could have had serious consequences for the entire inlet water system. Additionally, the existing four-inch strainers on the crossheaders were supplanted by five-inch strainers. Also, all front face nozzles, header fittings, and flexible connectors were replaced with larger ones of greater hydraulic efficiency. No changes were made in the piping of the export water system through B-Area, but the old wooden, emergency high tanks were replaced by steel tanks.
The entire effluent system of B Reactor likewise underwent extensive changes. Flow restrictions were placed in the main rear face piping to "pressurize" the flow (i.e., increase the temperature at which the exiting coolant would boil), in order to allow for higher operating and bulk exit water (water that had left the process tube) temperatures. By this time, experiments at C-Reactor had demonstrated that bulk outlet water actually could boil safely, as long as effluent piping was sturdy and reliable. Consequently, in early 1955, HW Process Specifications were changed so that Panellit gauges were set not at the "trip before boiling" but at the "trip before instability" limit. Instability of flow caused by inadequate piping and venting, scientists had learned, was the most crucial criteria in setting bulk exit water temperature limits. Soon afterward, they began to examine the possibilities of capturing steam formed in reactor effluent piping for electrical generation, or of filtering and venting the effluent steam, or of quenching the effluent with cold water so that no steam would form. Also, in Project CG-558 modifications, B Reactor's old downcomer, a 42-inch pipe with a single vertical baffle, was replaced by a cascade-type downcomer with multiple horizontal baffles. It was based in the existing cushion chamber at the minus 20-foot location, and a new, 66-inch effluent line was brought out from this point. The remaining sections of the cushion chamber were abandoned. Additionally, a diversion box was provided near the 105-B Building to collect facility floor drainage. The inlet sluiceways to the 107-B Basin halves were enlarged and provided with larger gates. The outlet weir was lowered by one foot, to prevent the basin from overflowing. No changes were made in the 107-B Basins themselves, but a much larger sedimentation basin (for unusual effluent) was constructed. As a result of the greatly increased flows to and through the 107-B Basins, effluent retention time was lowered to "not less than one hour."
The outfall system that conveyed effluent from the retention basin to the mid-river channel also was inadequate for the increased flow rates that would come as a result of Project CG-558. A completely new, 66-inch outfall line was built downstream of the older 42-inch line, extending 450 feet into the river. This new line drained the effluent basin and all other process drainage, while the old line was maintained to drain the regular 100-B Area sewer. While temporary jetties had been constructed to emplace former outfall lines, the huge new 1904-B pipe was the first to be installed using a current dissipator and a barge-mounted crane. The current dissipator consisted of three, 20-foot sections of interlocking sheet piling embedded in a concrete footing. The barge-mounted crane began the installation process by positioning (anchoring) the current dissipator just upstream from the outfall pipe at the water's edge. Next the barge returned to shore and emplaced the first 150-foot segment of 1904-B piping and lowered it into place on the river's bottom. A diver then bolted the flanged end to the outfall pipe at the water's edge, using 52 bolts. Following the same procedure, the current dissipator was repositioned twice, and the next two, 150-foot segments of outfall pipe were seated. Finally, concrete anchors and backfill were emplaced along the entire outfall length.
In 1955, as a part of Project CG-558, improved Panellit gauges and calibration equipment were installed. Use of the old equipment required that each gauge be removed and taken to a calibration bench for testing, resulting in the loss of 55 hours of production time (even using a full crew to perform the testing) to check the entire reactor. With the higher operating power levels, water pressure was consistently higher and the high and low "trip" points on the gauges were set much closer together. Water pressure surges in the 75-foot line that led from each process tube to its Panellit gauge often caused swings that brought about instrument scrams. The number of such incidents at B Reactor rose from 20 in 1951 to 42 in 1954. Consequently, accurate calibration became even more important, and the gauges needed much attention. Installation of the new calibration equipment involved replacing 2,004 pairs of needle valves (a pair for each process tube) with 2,004 three-way toggle valve/needle valve assemblies. These new valves had the same space requirements as the older valves, so adjacent equipment did not need replacement. However, a pump, piping and valving, and pressure regulators independent of the normal reactor cooling system had to be installed to supply regulated water pressure to the test manifolds. The expenditures for this new equipment were justified because, after installation, the gauges could be tested and recalibrated during regular, monthly reactor maintenance shutdowns.
The higher power levels made possible by Project CG-558 strained the material tolerances in B Reactor's biological and thermal shields, and made accurate temperature monitoring more crucial than ever before. New temperature monitoring devices were installed in the biological shield, consisting of thermocouples within each layer of steel at three points in the far (right) side shield and at one point in each steel layer of the top shield. Additionally, operators began inserting neutron-absorbing poisons in the reactor fringe channels nearest the biological shields, in order to reduce heat stress in the shields. They then compensated for this reactivity loss by enriching other areas of the reactor. Also, a rotating vane, sight-glass flow indicator was seated between the thermal loop and each thermal shield cooling tube, to indicate the coolant flow in each tube. Also, new iron/constantan and chromel/alumel wire thermocouples to monitor graphite temperatures were placed on stringers in process channels located in various zones of the pile. An automatic power calculating system was installed, and all existing instruments whose ranges would be exceeded by the new flow, temperature or power levels were re-worked or replaced to fit the new conditions.
Automatic outlet water temperature monitors were installed on about five percent of the process tubes, and the beta activity monitors that sampled each rear crossheader and riser were replaced with scintillation-type gamma monitors to permit earlier and more definite detection of slug ruptures. These new gamma detectors required a periodic oxalic acid and water flushing (cleaning). All of the new safety and monitoring instrumentation was reassuring to HW operators. A December 1956 Hazards Summary Report on Project CG-558 concluded that "overall safety of reactor operation will be improved...even at the higher power levels anticipated." The report noted that higher power levels would not increase the probability of a "disaster-initiating event," but that they would "reduce the available remedial action time...[and] increase the severity of the consequences. It concluded: "Loss of cooling water (unless the reactor has been shut down for a very long time prior to the water loss) will almost surely result in the large scale release of fission products. This destruction will occur even though the reactor is made sub-critical by insertion of the safety rods immediately upon loss of water."
New circuit breakers and underground cables were installed at the 151-B Electrical Substation, to transmit 13,800-volt power directly to the process pump motors in the 190-B Building. Also, a new, 13.8/2.3 kilovolt (kV), 5,000-kVa substation was installed at the 181-B River Pump House. Switchgear equipment was relocated from the 190-B Building to the 181-B and 183-B structures to supply the pumps of those facilities. Still, the immediate post-CG-558 power needs were such that all three, 151-B transformers needed to be in constant service, using forced air cooling systems (fans to cool radiator oil) "for a substantial portion of the time." The forced air cooling allowed the transformers to routinely carry loads that exceeded their self-cooled rating and, if one transformer broke down, the added burdens on the other two transformers increased to the point where even the fan-cooled ratings were exceeded.
Due to the higher heat levels experienced with the power level increases in the mid-1950's, the aluminum thimbles that lined the HCR, VSR and test hole channels at B Reactor had become in-pile hazards themselves. They were constantly in danger of melting, and, as a result, all thimbles that did not contain self-cooling tubes were removed in Project CG-558. Additionally, new HCR inner tip control sections were installed, using the existing rack sections. The new rod sections consisted of one-piece aluminum extrusions that slid through silicon sleeves mounted on the exterior of the left side shield. The new tips had similar neutron absorption capabilities to those of the earlier HCRs, but with greater flexibility and heat transfer capacity. Also, the existing shield gates over the HCR openings were removed, and the shield gate control lines were used as suction lines for a rod seal leak detection system. After Project CG-558, shield plugs were used when a rod was removed, and a shield was installed over the withdrawn parts of the rods. When not in use, the shield plugs were stored in a pit located at the left side of the reactor. Other miscellaneous equipment modifications made as a part of Project CG-558 included installing a 24-tube PCCF, equipped with the front and rear nozzle ball valves necessary for charging and discharging poison slugs during operations. Also, the charge elevator was "double-decked," and a "metal loader," separate from the "C" elevator, was installed to carry heavy charging tools and fresh uranium charges. This last innovation allowed the average charging rate at B-Pile to more than double, from about 19.5 tubes per hour to 41.2 tubes per hour.
Return to History of 100-B/C Reactor Operations, Table of Contents