Author: Michele S. Gerber, Ph.D., Facility Operations Division, Westinghouse Hanford Company, April 22, 1993 (approved for public release May 19, 1993)
The Hanford Site's 100-B Reactor was the first full-scale nuclear reactor to operate in world history. It was built by the Manhattan Engineer District (MED) of the Army Corps of Engineers, with the aid of the DuPont Corporation, to produce plutonium for the United States' military effort in World War II. Like the two other 100 Areas at the Hanford Engineer Works (HEW - World War II name for the Hanford Site), 100B Area was comprised of 685 acres of land enclosed by 4.1 miles of fencing. It also contained 4.25 miles of broad gauge rail track and 6.75 miles of gravel roads. Descriptions of the 100B Area also can be found elsewhere, but a brief listing of the ancillary buildings and structures is appended.
B Reactor itself, along with its influent and effluent systems, operated as follows: Water to cool the pile was withdrawn from the Columbia River by pumps located in the 181-B Building (River Pump House). The 182-B Building (Reservoir and Pump House) provided reserve (secondary backup) water for cooling, water for steam condensers, and raw water for the 200 Areas (known as "export water"). Before Project CG-558, a series of major modifications that took place in the 100-B Area from 1954 through 1956, the 182-B Building also supplied water for the 183-B Filter Plant. After passing through the 182 Reservoir and Pump House, B-Pile's cooling water was routed to the 183-B Filter Plant and Chemical Treatment Facility (known as the head house), where a number of chemicals were added to purify and ready the raw water for reactor use. (For discussions of the influent treatment chemicals, see Sections 2.12, 3.6, and 4.14.) The water then was passed through gravity filters consisting of sand, gravel, and anthracite coal (known as anthrafilt) and was stored in underground "clearwells." From there, it was pumped through underground piping to the 189-B Deaeration Facility, where dissolved gases and entrained air introduced in the filtration process were removed. It was believed by early engineers that the presence of such gases in process water could affect the heat transfer capacity of the coolant, but this problem turned out to have only minor significance. The process water then was pumped to four 1,750,000-gallon steel storage tanks in the 190-B Process Pump House, where sodium dichromate (Na2Cr207) was added to inhibit corrosion on the reactor's process tubes. Due to the rush of wartime construction, the underground piping was not "puddled in," nor provided with sand bases. Therefore, over time, according to a later Hanford report, it was rendered "susceptible to breakage." Twelve sets of steam and electric pumps (a pair of each in each set), also located in the 190-B Building actually pumped the ready water through the reactor (105 Building).
Electrical power for B-Pile came off of the Grand Coulee Dam-to-Bonneville Dam grid, via Midway Substation and the 151-B substation. One 20,000 kVa transformer and one 15,000 kVa transformer were located at the 151-B Building, and emergency electrical power was available from a steam turbine generator located in the 184 Power House. This generator supplied power to steam turbine pumps for the secondary coolant system, located in the 181-B, 182-B, 183-B, and 190-B facilities. The 184-B Building also supplied office heat and other heating needs through overhead steam lines that looped throughout 100-B Area. The furnace ash and fly ash from the steam-driven pumps were dumped to a pit beneath the coal furnace, and then sluiced to a trench running beneath the building and emptying into the sump. The wet ash subsequently was pumped from the sump to the 188-B ash pit. "Last ditch" (tertiary backup) cooling water was available from two wooden, high water tanks (187 Structures) that held 300,000 gallons each, and from the export water lines to the 200 Areas. Additionally, water from the 182-B and 183-B Buildings could be diverted and pumped to the reactor for emergency cooling purposes. Emergency power for reactor instrumentation was supplied from a gas-powered backup generator located outside the pile building.
B-Pile rested on a 23-foot thick concrete foundation topped with cast iron blocks that served as a thermal shield. The 105-Building walls consisted of reinforced concrete in the lower portions and concrete block in the upper portion, varying from three to five feet thick. The roof was composed of precast concrete roof tile, except over the discharge ("D") area enclosure (the rear face) and the inner horizontal rod room (the HCR access area). Over those areas, the roof was composed of six-foot thick reinforced concrete.
The reactor core itself consisted of a graphite "stack" that measured 28 feet from front to rear, 36 feet from side to side, and 36 feet from top to bottom. The stack was pierced front to rear by 2,004 process channels that held the fuel elements. Usually, 32 fuel elements or slugs made up the active fuel charge in each process tube. Nine horizontal channels for control rods entered from the left side of the reactor, and 29 vertical channels for safety rods entered from the top. Additionally, six test holes leading from the right side of the pile existed for the irradiation of experiments, foils, counters, ionization chambers, and special samples. The horizontal control rod (HCR) and vertical safety rod (VSR) channels, as well as the test holes, were lined with a thin sheet of aluminum known to Hanford workers as a "thimbles." At the front and rear of each process channel penetration, a seven and one-half foot long, carbon steel entry and exit sleeve known as a "gunbarrel" served to transfer the weight of the thermal shield to the biological shield. It also protected the graphite during charging ("C") operations, maintenance activities, and other manipulations. The ends of each process tube flared into Van Stone* flanges, to facilitate a close fit and interface against the gunbarrels. Additionally, an asbestos gasket lay against each Van Stone flange, between it and the stainless steel nozzle that projected from the front and rear of each process tube. The nozzles connected to the larger coolant delivery and exit systems.
The graphite core was surrounded by a cast iron thermal shield layer that varied from eight inches thick at the reactor sides, eight and one-eighth inches thick at the top, ten inches thick at the top and rear, and ten and one-fourth inches thick at the bottom. Cooling for the top, side, and bottom shields was provided by circulating water tubes embedded in the blocks. The front and rear shields were cooled by regular reactor coolant flow that passed through the process tubes. The entire thermal shield was surrounded on all sides except the bottom by a 52-inch thick biological shield that consisted of alternate layers of masonite and steel. While the thermal shield absorbed and converted to heat nearly 97% of the gamma energy produced by the fission process, the biological shield absorbed the fast neutrons that passed through the thermal shield. The biological shield slowed the fast neutrons to intermediate flux, and absorbed the released nuclear energy. The entire reactor block then was enclosed in a welded steel box that functioned to confine the inert gas atmosphere within the reactor. Neoprene* expansion joints were placed on the corners of the block to allow for thermal expansion, and expansion bellows were located at each process tube opening. The bellows served as gas seals as the process tubes expanded and contracted with temperature and with the distortions of the graphite. Additionally, each process tube penetration through the biological shield was surrounded by a series of circular, cast iron shields known as "shielding doughnuts." The test holes had removable lead rods for plugs.
The atmosphere of B Reactor originally was composed of helium (He), an inert gas selected for its high heat conductivity (removal) capacity. Early HEW operators estimated that the formation of one gram of plutonium-239 (Pu-239, also known in Site codes as "product" or simply as "49") liberated approximately 80-million B.T.U. (British Thermal Units) of energy, the equivalent of 1,000 kilowatt days of energy. The helium removed heat, moisture, and foreign gases from the pile, and also served as the detection mechanism for water leaks within the pile. Sampling tubes were located in the gas plenum between the rear-face biological and thermal shields. Water leaks in the core flashed to steam, and were detected by measuring the amount of water vapor in the gas sampling tubes. Helium arrived at HEW in rail cars, was unloaded into high pressure storage tanks at the 110-B Building (Gas Storage and Unloading Station). Then it was transferred into low pressure tanks for makeup in the 115-B Building (Gas Purification Facility), located adjacent to the reactor building. The 115-B Facility contained apparatus for circulating the gas, and three silica gel towers that dried the gas as it passed through them. Equipment also existed in the 115-B Building to purify the He by pressurizing and refrigerating it, and then passing it through activated carbon. Underground piping connected the 115-B and 105-B Buildings.
Process cooling water was pumped to B Reactor from the 190-Bxx Building through twelve 12-inch pipes, located in an underground tunnel, to a common header in the 105-B Building valve pit. In the valve pit, the flow was split into two 36-inch headers that ran to the pile itself. At the base of the reactor face, the water was routed upward through 36-inch risers, and was then distributed across the front face through 39, four-inch, stainless steel crossheaders. These small crossheaders, in turn, supplied the cooling water to a row of individual process tubes via coiled lengths of one-half-inch aluminum tubing known as "pigtails" (so designated because their expansion coils resembled pig's tails). The crossheaders were connected to the pigtails with devices called Parker fittings.* After passage through the process tubes, the cooling water (now effluent) exited the reactor through a flow system that essentially duplicated the influent piping in reverse. Pigtails on the rear of the process tubes fed the effluent into crossheaders, which in turn connected to two vertical risers (one on each side of the rear face). These risers routed the water into a common crossover line that discharged into a downcomer. B Reactor's downcomer was a 42-inch pipe containing a vertical baffle to slow the water's flow. From it, the coolant cascaded into the cushion chamber, a cement enclosure lined with cypress planks at the bottom of the downcomer, and thence, by gravity, through a 48-inch, concrete, underground line to the 107-B Retention Basin. After a variable holdup period, the effluent was released to a point nearly 400 feet into the Columbia River via a 54-inch concrete, underground and underwater pipe.
Reactivity control at B Reactor originally was maintained by the nine HCRs, the 29 VSRs and, in emergency cases, by a liquid boron solution that was stored in a pedestal tank located about five feet above the top of the reactor. The function of the HCRs was to control the equilibrium and transient power levels of the reactor during routine operations, and to maintain the desired neutron flux distribution. The HCRs each were about 36 feet long, with the poison (neutron absorbing segment) being about 29 feet, four and one-half inches. They were self-cooled, via aluminum tubing that transported ten gallons per minute (gpm) of water through each rod and that was fastened to the rod with pressed and sintered boron-carbide and aluminum rings. Two of the rods were electrically driven, and seven were hydraulically-driven. The latter were known as shim rods, and were used to achieve ongoing operational control and desired fluctuations.
The VSRs were 39-foot long, stainless steel sleeves with three-sixteenth-inch thick, boron-stainless steel sleeves inside. The outside diameter (O.D.) of the VSRs was two and one-fourth-inches. Each VSR was inserted and withdrawn from the reactor via two separate cables wound around a winch located 40 feet above the top of the reactor. In cases of automatic shutdown ("scram") of the reactor, the electromagnetic clutch holding each rod in the out position would be de-energized (de-magnetized), and the rods would free-fall by gravity into channels penetrating the reactor. The VSRs needed no cooling system, as they were not in the active portion of the reactor during operations. The "last ditch" safety system, the boron solution, was connected to each of the 29 VSR channels via one-half-inch pipes. The aluminum thimble lining served to protect the graphite from contact with this wet solution.
The safety, process control, and temperature instrumentation that monitored B Reactor originally was quite simple. Three manually operated, pushbutton safety devices existed to insert either just the HCRs, or both the HCRs and VSRs, or the liquid boron poison. There also were two automatic circuits that would shut down the reactor without deliberate human action in what were called "instrument scrams." One device simply inserted the HCRs and VSRs upon loss of electrical power. The other consisted of four ionization chambers (neutron flux monitors) located in four zones of the reactor. When the power level exceeded a certain pre-set limit as recorded on picoammeters in the control room, the circuit would "trip" (activate) and automatically insert the HCRs and VSRs. Soon after startup however, the operators realized that a key indicator of whether or not the coolant was flowing smoothly and uniformly through the reactor was needed. A Panellit* gauge, which measured coolant pressure by sensing the amount of flow passing through an orifice, was installed at the inlet of each of the 2,004 process tubes. Each sensor was attached via a single hydraulic line to a pressure monitoring gauge in the control room, and was set at both a high and low trip point. The initial usage period of the Panellit gauge system, in early 1945, was very complicated. Three operators were needed to scan the 2,004 gauges and to alert the control room supervisor to near-trip conditions during each startup, as coolant flow pressure could fluctuate in quirky and non-uniform ways during the temperature and power level increases of a startup. During B Reactor's first startup after complete Panellit installation, eight automatic scrams tripped by the pressure settings occurred in one shift. However, increased experience with the system soon enabled HW operators to eliminate many of these scrams.
In addition to the safety or shutdown circuitry with which B Reactor was equipped, there was a need for ongoing measurements of neutron flux, numerous types of temperature and pressure measurements, and activity monitors that would indicate increased radiation levels in various areas. No one instrument had enough range to measure neutron flux all the way from shutdown (background) levels to the approximately 1,000,000,000,000 (one trillion) times background levels experienced during full power operation. Therefore, the reactor was fitted with a sub-critical monitor (low-level neutron flux monitor) and a galvanometer system to measure mid-range neutron flux. The high-level neutron flux monitor, discussed earlier, measured full power operations and was capable of initiating an automatic pile scram if its pre-set limits were exceeded. Effluent water temperature was monitored via a separate thermocouple attached to each process tube outlet, and gross beta effluent activity was measured at the reactor's rear face by ion chambers positioned to receive a small water sample from each end of the rear crossheaders. The graphite moderator temperature and the thermal shield temperature were measured by thermocouples embedded in various sectors of the graphite and shields. However, when the thermocouples in the graphite began to fail, they were replaced by eleven thermocouples inserted into pile process channels on graphite bead "stringers" (lengths of narrow, sliding tracks). Deterioration of the biological shield, measured by neutron leakage out of the shield, was gauged via a gold foil that was placed just outside the shield. The relative normalized radioactivity of the foil was then counted and calculated, and was presumed to be proportional to the amount of leakage through the shield. Water pressure was measured via many gauges located at the top of the inlet risers, the crossheaders, the top of the downcomer, in the HCR coolant tubing, at the high tank and in the export system piping.
B-Pile's total reactivity coefficient was measured by abruptly lowering the power level by 25 megawatts (MW) in order to determine the temperature transients between the graphite and the uranium fuel, and then to compare these transients with the power level changes. Such a computation was possible because the graphite temperature (measured in the thermocouples) took over an hour to drop to a new low following the 25 MW power decrease, but the uranium temperature (measured by cooling water temperature) dropped to new equilibrium in just a few seconds. As B-Pile's power level increased continually throughout the years after 1948, many new instrumentation devices were added to the reactor system.
Return to History of 100-B/C Reactor Operations, Table of Contents