A diamond in Dogpatch: The 75th anniversary of the Graphite Reactor Part 1: The War Years
The Graphite Reactor and its companion plutonium separations facility in Oak Ridge played an extraordinary role in the success of the Manhattan Project’s atomic bomb design and production effort.
By Sherrell R. Greene ton Pile, the Graphite Reactor and its a humorously cynical allusion to the hill- companion radiochemical separations billy town made famous in Al Capp’s Li’l he old corrugated steel building building comprised the Manhattan Proj- Abner comic strip. As it turned out, Dog- sits quietly on Hillside Drive in ect’s plutonium Pilot Plant, or Clinton patch wasn’t remote enough. The decision Tthe very heart of Oak Ridge Na- Semi-works. (The formal name of the U.S. to move the Pilot Plant to Clinton was tional Laboratory’s (ORNL) original cam- World War II atomic bomb effort was, of accompanied by an even more disruptive pus. A modern day visitor to the lab can course, the Manhattan Engineer District. decision to move the planned plutonium easily miss Building 3001; most do. But Here we will employ Manhattan Project, production operations from the Clinton inside this old weathered building sits the the name adopted after World War II.) The Engineer Works to an even more remote Oak Ridge Graphite Reactor. Inside this Pilot Plant wasn’t even supposed to have site, soon to be identified as Hanford, in building, between November 4, 1943, and been built in what is now Oak Ridge. The Washington state. November 4, 1963, magic happened. The scientists at the University of Chicago’s As originally conceived, the Pilot Plant Graphite Reactor is one of the original and Metallurgical Laboratory (or Met Lab) had four objectives: (1) to serve as a tech- best preserved facilities of the Manhattan who conceived it in the summer of 1942 nology development and demonstration Project National Historical Park. Sitting assumed that the Pilot Plant would be platform for the much larger production alone in the operating gallery of the reac- built in the suburbs of Chicago. Howev- piles and plutonium separations facilities; tor on a rainy afternoon, you can almost er, as the potential risks of working with (2) to supply gram quantities of plutoni- hear the voices of Arthur Compton, En- atomic “piles” (a term coined because they um urgently needed to enable progress rico Fermi, Eugene Wigner, Alvin Wein- were quite literally carefully constructed in bomb design; (3) to serve as a training berg, and Glenn Seaborg. Today, as we piles of graphite blocks and uranium) and center for the staff of the larger produc- celebrate its 75th (diamond) anniversary, plutonium processing operations became tion facilities; and (4) to provide a devel- the Graphite Reactor remains a true time more widely understood, leaders of the opmental platform for the operating pro- capsule from the dawn of the nuclear age. Manhattan Project overrode the desires of cedures required for the plutonium pro- Originally known as the X-Pile, the Compton and his team at the Met Lab. duction facilities. Despite these sharply X-10 Pile, and (more widely) the Clin- Within a few weeks of the appointment focused design objectives, the Pilot Plant in mid-November 1942 of E. I. DuPont and, in particular, the Graphite Reactor Sherrell R. Greene (
38 • Nuclear News • November 2018 www.ans.org/nn Safety Rod Graphite Thermal Column 7-Foot-Thick Concrete Shield Channel Scanning Device Bridge Tubes
Airflow Baffle
Experimental Hole Fuel Loading Channels Through Shield Control Rod Location
Loading Elevator
Air Exhaust Uranium-Graphite Lattice
Air Inlet Drawing: DOE Graphite Reactor design features
The Pilot Plant The Graphite Reactor and its compan- ion radiochemical processing facility were the first progeny of a shotgun wed- ding between the Met Lab and DuPont. The Met Lab was a combative bride. DuPont was a reluctant groom. General Groves held the gun. After initially refus- ing to have any responsibility for the Pilot Plant, DuPont eventually surrendered to Groves’s relentless prodding, agreeing by January 1943 to assume responsibility for the design and construction of both the Pilot Plant and the plutonium produc- tion facilities. DuPont, however, would not agree to operate the Pilot Plant. After initially refusing to operate anything in Dogpatch, the Met Lab and the University of Chicago finally agreed to assume oper- ational responsibility for the Pilot Plant. The division of design responsibilities between the Met Lab and DuPont for the Pilot Plant was a masterpiece of obscu- rity and a testament to General Groves’s matchmaking skills. DuPont would “de- sign” the Pilot Plant, and the Met Lab would “check the design and be respon- sible for the adequacy of the information on physics and chemistry.” It was clear, however, that DuPont was in charge. The Graphite Reactor design that
emerged from the DuPont/Met Lab design Photo: DOE team (and that of its companion radio- The Graphite Reactor and its control room
November 2018 • Nuclear News • 39 water-rich barytes-haydite concrete. The outer dimensions of the shield—the visi- ble boundaries of the reactor—were about 47 feet long, 38 feet wide, and 32 feet high. Perhaps the most prescient design feature of the pile (located in Building 105, now 3001) was the access provided through the shield into the reactor’s core for instruments and experiments. There were dozens of openings of various sizes (up to 5 feet square) in the pile’s shield on all sides and the top, the intent being to provide access to the diverse nuclear and thermal environments afforded by the reactor’s power level and physical size. Among the more interesting pile access design features were the two “animal tun- nels” in the top face of the shield. These openings hosted an ingeniously designed carriage and gate device that allowed spe-
Photo: ORNL cially machined graphite trays containing Mouse irradiation tray rabbits, mice, or other small creatures to be inserted and withdrawn during pile chemical processing facility) was elegantly surized accumulator system to shut down operation for the study of radiation effects simple, yet amazingly flexible. Wigner and the reactor in case of an emergency. Con- on mammals. Weinberg at the Met Lab were responsible trol rods entered the pile at right angles to The Separations Facility (Building 205, for the nuclear design of the 1-MWt reac- the fuel channels. Two horizontal borated- now subsumed in Building 3019), the de- tor. (Wigner, in later years, would credit steel regulating rods could be operated sign of which lagged that of the reactor Weinberg with having “singlehandedly” manually or automatically to provide fine by a few months, was the world’s first ra- created the nuclear design of the reactor.) control of reactor power. Four borated- diochemical processing “hot cell” build- The thermal design of the reactor seems steel “shim” rods (so called because they ing. Most of the basic design features of to have been shared between the Met Lab were normally only partially inserted into modern radiochemical processing and and DuPont’s design team in Wilmington, the core) could be automatically insert- nuclear fuel reprocessing facilities were Del., with DuPont leading the mechanical ed to shut down the reactor, or manual- first employed in Building 205. The facili- design and the two groups checking each ly inserted to compensate for reactivity ty utilized a hybrid plutonium separations other’s work. changes that the regulating rods could not “flowsheet” in which bismuth phosphate The reactor was a 24-foot cube con- accommodate. The reactor was surround- served as the carrier in the plutonium ex- sisting of 675 tons of graphite moderator ed on all sides by a 7-foot-thick laminated- traction steps, and lanthanum fluoride as blocks arranged in 73 layers. The blocks concrete radiation shield made up of the carrier in the concentration and iso- were machined on-site at Clinton. Thirty- layers of standard concrete and a special lation steps.[1] It contained six processing six rows of diamond-shaped, hollow fuel channels (a total of 1,248) on 8-inch centers traversed the pile from front to rear. These channels were loaded with 4-inch-long by approximately 1-inch-diameter cylin- drical aluminum-clad metallic natural uranium fuel “slugs.” The fuel slugs would be pushed into the fuel channels from the front “charging” face of the pile and pushed out the back of the channel after the desired irradiation time, falling into catch buckets located beneath the pile in a water-filled basin. Cooling air was drawn through the fuel channels by fans and then exhausted up a 200-foot stack. The cooling system was equipped with two electrically driven 30,000-cubic-feet-per-meter (cfm) fans and one 5,000-cfm steam-driven standby fan. The reactor’s safety and control systems were a study in simplicity, redundancy, and reliability. Four gravity-operated borated-steel safety rods were suspend- ed above the reactor, ready to drop into the core when automatically or manually triggered. Two vertical channels were also
available to receive borated steel shot that Photo: DOE/Ed Westcott could be fed either by gravity or by a pres- The Graphite Reactor and Separations Facility under construction in the summer of 1943.
40 • Nuclear News • November 2018 www.ans.org/nn A diamond in Dogpatch: The 75th anniversary of the Graphite Reactor Photo: DOE Graphite Reactor log book, noting, on the left-hand page, first criticality on November 4, 1943. “cells”: one for the dissolution of irradiat- the Clinton Laboratories in August. Mar- challenges and the wartime competition ed uranium slugs; four for plutonium re- tin Whitaker, who had moved to Clinton for supplies, the DuPont construction covery, purification, and waste recovery; from Chicago, was the first director of team turned Building 105 over to the Met and one double-sized cell for the storage the Clinton Laboratories and would serve Lab operations team on October 16, only of contaminated equipment that failed in through 1945. 172 days after ground was broken for the operations. Each of the four plutonium Construction of both Building 105 building. Building 205 was turned over recovery cells contained a precipitator, a and (especially) Building 205 proceed- to the operations staff five weeks later, on centrifuge, a catch tank, and a neutralizer, ed in parallel with their design. Record- November 26.[1] along with the required piping for drain- setting rains in the late spring and early Fuel loading of the Graphite Reactor be- ing and waste management systems.[1] summer turned the entire site into one gan at 4:30 p.m. on November 3, 1943, and The cells were separated from each other enormous, sloppy mess, and worker turn- the reactor achieved criticality at 5 a.m. the and from a centrally located, common op- over was a constant problem. It wasn’t a next morning (just 192 days after excava- erating gallery by thick concrete walls that pleasant place to work. But despite the tion started), with Whitaker, Compton, provided both shielding from radiation and protection in the event of chemical explosions. All of the operations in each cell were remotely controlled from pan- els in the operating gallery. Liquid waste was routed through underground piping to underground storage tanks. Building 205 was connected to Building 105 via a flooded underground tunnel containing a monorail crane that ferried buckets of ir- radiated fuel slugs from the reactor to the processing facility. The building incorpo- rated a 200-foot exhaust stack, where di- luted dissolver gases were exhausted to the atmosphere. Excavation for Building 205 began on March 9, 1943, and for Building 105 on April 27. The first group of scientific -per sonnel moved from Chicago to Clinton Photo: DOE/Ed Westcott in April. The on-site operating and tech- Building 105—the Graphite Reactor (at left)—and Building 205—the Separations Facility nical staff were formally organized into (at right)—in 1943.
November 2018 • Nuclear News • 41 and Fermi looking on. Seaborg would ar- dubbed “Thin Man,” purportedly named polonium-210—Successful detonation of rive a bit later that morning. The Pilot Plant for President Franklin Roosevelt. The first the atomic bomb depended on the avail- had come alive. If, as Compton said of Fer- batch of irradiated fuel slugs from the ability of a small number of neutrons to mi at Chicago Pile 1, the “Italian Naviga- Graphite Reactor was received in Build- initiate the chain reaction at a precise point tor” had “landed in the New World,”[2] the ing 205 on December 20, 1943,[1] and the in the assembly sequence as the fissile ma- Graphite Reactor and the Pilot Plant would first plutonium product was produced by terials were brought together.[5] Accord- be the vessels to carry the first explorers February 1, 1944. Gram quantities of plu- ing to Manhattan District History, Project beyond the shoreline of that New World to tonium were being shipped from Clinton Y, by the late summer of 1943, weaponeers reconnoiter its riches. by mid-March 1944.[4] at Los Alamos were becoming increasing- Emilio Segre received the first shipment ly interested in polonium-210 as a possible The handmaiden of Los Alamos of plutonium at Los Alamos from the Pilot initiator material.[5] Small quantities of The Graphite Reactor began operation Plant in early April. Within several days, the isotope were being produced by chem- almost 11 months before the Hanford he determined that plutonium produced ical processing of radioactive lead residues 100-B reactor achieved first criticality on in a reactor had higher concentrations of from the radium industry, but it was clear September 27, 1944. Those months would Pu-240 than cyclotron-produced pluto- that the only practical path to production have a far greater impact on the U.S. nium and that the resulting spontaneous of the required quantities of polonium World War II atomic bomb program (on fission rate of reactor-bred plutonium was was by neutron irradiation of bismuth.[7] both bomb design and bomb development so high that the gun assembly approach The only option for these irradiations was schedule) than anyone could have imag- would not work.[5] Thin Man was dead. the Graphite Reactor. As a result, in Janu- ined that crisp fall morning in November This finding, along with Snell’s confir- ary 1944, the Graphite Reactor began irra- 1943, when Oak Ridge became mankind’s mation at Clinton that natural uranium diating bismuth slugs that were shipped to second beachhead in the New World. The could not be employed to fuel an atomic Monsanto Chemical Company’s Dayton, Pilot Plant would become the handmaid- bomb, settled the baseline approach for Ohio, facility, where the polonium was en of Los Alamos’s bomb design effort as the U.S. atomic bomb program on gun separated from the bismuth and shipped Robert Oppenheimer’s team accelerated assembly of enriched uranium and cata- to Los Alamos.[4] Polonium produc- its efforts in the spring of 1944 to solve lyzed a desperate effort to develop an al- tion did not commence at Hanford until the riddle of the bomb design. The typical ternative assembly method for the pluto- March 1945.[8] dynamic of this relationship was that Los nium weapon. ■■Producing the first U-233—In addition Alamos would encounter a challenge that Segre’s finding precipitated a major cri- to all of their other priorities, Oppen- could be overcome only via the use of the sis in the plutonium bomb development heimer and Edward Teller at Los Alamos Graphite Reactor’s neutrons, or radioiso- project at Los Alamos. Oppenheimer had become interested in the possibility topes produced at the Graphite Reactor/ briefly considered resigning. Cooler heads of a uranium-233 bomb in the summer of Pilot Plant. prevailed, and Los Alamos launched an 1943.[5] They reasoned that U-233 might The more important contributions of all-out effort to develop the implosion- provide an easier second path to a bomb the Graphite Reactor to the design and de- assembly approach, in which a subcritical than Pu-239, and they were concerned velopment of the atomic bomb include the configuration of fissile material is rapidly that Germany might be pursuing a U-233 following: compressed into a supercritical configura- weapon as well. Thus, they requested that ■■Verifying that natural uranium couldn’t tion via the detonation of a surrounding the Met Lab launch a program to provide fuel an atomic bomb—As World War II mass of high explosives. Los Alamos’s re- gram quantities of U-233, using its “sur- proceeded in 1942 and 1943, the question markable success in developing this tech- plus” neutrons available in the Graphite of whether natural uranium could be em- nique delivered “Fat Man,” the plutonium Reactor. Within two weeks of the reac- ployed to build a functional atomic bomb weapon—supposedly named after Win- tor’s startup in November 1943, 80 cans became a matter of urgent interest. If it ston Churchill—that was dropped on Na- of thorium carbide had been loaded into could, anyone with a sufficient supply of gasaki on August 9, 1945. the Graphite Reactor in response to Los uranium ore could develop a bomb. Could ■■Visioneering Fat Man—As document- Alamos’s request.[9] Before the end of a fast neutron chain reaction be sustained ed in Manhattan District History, Project the year, however, General Groves would in a large mass of natural uranium? The Y—The Los Alamos Project, the “RaLa” order that work be halted on the chemi- Snell Experiment, conducted by Arthur (radioactive lanthanum-140) experiments cal separations process required to isolate Snell in the Graphite Reactor in early 1944, conducted by Los Alamos “became the the U-233 from the irradiated thorium proved that even 35 tons of natural urani- most important single experiment affect- targets. He judged that the probability of um metal could not sustain a fast neutron ing the final design of the bomb” (i.e., the success with U-233 was much less than reaction. This confirmed Snell’s previous Fat Man plutonium bomb).[6] The experi- with Pu-239, and in any event, there was finding from an experiment done with 5 mental method developed by Los Alamos no hard evidence that Germany was pur- tons of natural uranium in Chicago’s Cy- for studying implosion dynamics relied suing a U-233 weapon. clotron.[3] No one was going to build a on the use of the intense gamma radiation ■■Producing tritium for Teller’s “Super”— bomb by piling up natural uranium. emitted by RaLa to visualize the response In May 1944, Oppenheimer, at the ■■Killing Thin Man—By the spring of of a mock plutonium core to the converg- prompting of Teller—who had been work- 1944, Los Alamos’s baseline design for ing shockwaves from detonation of the ing on his concept for “the Super,” or both the uranium and plutonium bombs surrounding explosives.[5] Thus, it was thermonuclear fusion bomb, since early employed the “gun assembly” approach, soon after Segre’s discovery in April 1944 1942—requested that the Graphite Reac- in which two separated subcritical mass- that Los Alamos presented the Clinton tor utilize some more of its “surplus” neu- es of material are rapidly assembled in a Laboratories with an urgent request for trons to provide Los Alamos with a small manner similar to firing a projectile from the production of radioactive barium-140, quantity of tritium (bred by bombarding a gun. The gun-assembled uranium bomb from which Los Alamos would synthesize lithium-6 with neutrons) for characteri- was dubbed “Little Boy” and was the the RaLa. Barium-140 production in the zation studies.[5] The atomic alchemists weapon that would be dropped on Hiro- Graphite Reactor began early in 1944 and at Clinton turned to the task, and the first shima, Japan, on August 6, 1945. The gun- continued until 1946.[4] tritium was produced at the Graphite Re- assembled plutonium weapon design was ■■Lighting the fire: The production of actor before the end of June 1944. Tritium
42 • Nuclear News • November 2018 www.ans.org/nn A diamond in Dogpatch: The 75th anniversary of the Graphite Reactor production did not begin in the Hanford piles until September 1946.[8] Visiting the Graphite Reactor Of all these contributions to the U.S. The Oak Ridge Graphite Reactor is one of eight Signature Facilities of the Man- atomic bomb design effort, only the sec- hattan Project National Historical Park. It is open for public bus tours from March ond (early plutonium production) was through November. Those wishing to visit the reactor must register in advance at planned when the concept for the Graph-
November 2018 • Nuclear News • 43