Before the plans for the separation plants for the plutonium process could be completed, we desperately needed a supply of uranium that had been exposed to radiation—that is, uranium which had been subjected to neutron bombardment in a pile and thus contained plutonium. This was a compelling reason for the decision to build the small semi-works plant at Clinton.
The Clinton plant had to be designed and built with a minimum of preliminary study and thought—far less than would have to be devoted to the reactors in the final production plant. Our major consideration was speed. Although the works would be relatively small, we had to apply the same basic principles and incorporate in them many of the facilities necessary for the large-scale production and separation of plutonium that we would use later at Hanford.
It was different in a number of details: The Clinton pile was air-cooled while those at Hanford used water; it was designed to operate at a rate of 1,000 kw, a small fraction of the amount used by the piles we built later at Hanford. Air-cooling was used because it made for simpler, and therefore more rapid, construction. While air-cooling was entirely practical for a pile of this size, it could not be used for the larger reactors because of heat transfer problems. Later, it was found possible to increase the power of the Clinton pile through a number of modifications.
At the time the design of Hanford was begun, plutonium had been separated from uranium only in infinitesimal quantities and by laboratory methods. The most feasible process was largely a matter of opinion and the equipment needed was even more conjectural. Yet design had to be carried forward on the basis of what information we had.
Because of the overwhelming importance of time, we had to ignore the normal methods of orderly development. Design, procurement and construction had to proceed concurrently with the selection of the separation process, with its development, and with the growth of the required basic scientific knowledge. The primary purpose of the Clinton semi-works, which it achieved, was to produce a sufficient amount of irradiated uranium to give us the essential information upon which to base the Hanford separation process. This despite the fact that the Hanford installation had to be designed and much of it built before we had the benefit of any of our experience at Clinton.
Within du Pont, all work on plutonium was placed under the jurisdiction of a special section of the Explosives Department, known as TNX, headed by Roger Williams. The company’s Engineering Department was delegated responsibility for engineering under TNX for whatever design and construction might be required. The Chief Engineer was Everett G. Ackart; the head of construction, and later Assistant Chief Engineer, was Granville M. Read; the head of design was Tom C. Gary; and, as I have said, Crawford H. Greenewalt represented du Pont at the Metallurgical Laboratory in Chicago. Greenewalt’s was a particularly difficult and important assignment, for he served as the bridge between the hard-driving, thoroughly competent, industrial-minded, scientific engineers and executives at Wilmington and the highly intelligent and theoretically inclined scientists at Chicago. This meant he had to shuttle back and forth between Chicago and Wilmington and later Hanford, easing tensions and calming tempers and, at the same time, seeing to it that needed scientific decisions were promptly reached at Chicago. He was eminently qualified for this liaison assignment, for he was a well-trained chemical engineer and had had more than twenty years of experience in the du Pont organization. He did a superb job.
Information on research developments at Chicago was supplied to du Pont, largely through Greenewalt, in several ways. Copies of all research reports were sent to du Pont. Du Pont people made frequent visits to Chicago and to Clinton to discuss the latest results of research, to review the preliminary designs and to obtain specific detailed information. Some du Pont people were stationed at Chicago and Clinton throughout the design period to provide a link between theory and practice. Every major decision involving details of layout and process had to be concurred in by Compton’s people, to ensure that the plant would be built on the basis of the latest and most reliable technical knowledge. This was constantly changing. In addition to the usual final review and approval by the district of all drawings, those that dealt with the process details were formally approved by the Chicago laboratory before they were released for construction.
Security necessitated special handling of the multitude of drawings, reports and correspondence that passed between the various offices; it also made communication by telephone and telegraph difficult, so we had to take extra care to avoid administrative delays. To speed up construction, working drawings were broken down to disclose as little information on the over-all project as possible. By this means, it was possible to treat many of them as unclassified. Such measures were particularly necessary when drawings were sent to subcontractors or used by construction workers at the site.
Even before du Pont came into the project, the laboratory at Chicago had begun work on the preliminary design of a helium-cooled pile that would operate at a much lower power level than that then being considered for the main plant. After receiving a detailed report from the laboratory, du Pont continued work on this design during December, 1942, and January, 1943; at the same time it began an intensive study of the relative advantages of helium- and water-cooled piles.
Our plans for basing the Hanford design upon helium-cooled piles were finally abandoned in February of 1943 when everyone agreed that helium-cooling, while possibly more attractive from a theoretical standpoint, would present a great many practical difficulties in handling and purifying the large volume of gas that would become radioactive and which, because of its radioactivity, would make the maintenance of equipment difficult, if not impossible. A major consideration in dropping the helium-cooled pile was the problem of designing and maintaining the necessary pressure-sustaining enclosure for the pile. Among other difficulties was that of loading and unloading such a unit under pressure. When it developed that the water-cooled pile would be easier and cheaper to design and build, all work on the helium method was stopped.
Although at the time the Hanford site was selected we had expected that the piles would probably be cooled with helium, we had also recognized that water-cooling might be used, and had sought a site which would be suitable if this proved to be desirable. Later, after we had decided on water-cooling, we discovered that not only was it necessary to have large quantities of cold water, but that its purity was of the utmost importance. We were just lucky that the Columbia River water did not contain dissolved chemicals in sufficient amounts to necessitate more than normal treatment. Even so, as the first pile was being finished, fears arose that if we were to keep the reactors working, we might require cooling water of a purity equivalent to that of distilled water. For that reason, we considered building a large deionizing plant for the second reactor so that at least one of the three piles would certainly be operable at all times.
I was discussing the advisability of this with G. M. Read of du Pont, one night at Hanford, when Dr. Hilberry came into the room. I asked him for his views, and he replied that he did not think we would need the deionizing plant, but if we did, we could not do without it. I turned to Read and said, “Go ahead and build it.” Hilberry then asked what it would cost and I told him that it would be somewhere between six and ten million dollars. He replied, “I’m glad I didn’t know that when I gave my opinion.” Such quick decisions were not too frequent and they were always preceded by as much research, study and thought as could be devoted to them without delaying the completion of the project. Nevertheless, there were many decisions that had to be made when the unknown factors far outweighed the known. We built only the one deionizing plant and fortunately never found any need for it, for had it proved necessary, we would have had to build two more in a hurry, and would have lost considerable production while they were under construction.
We expected the cooling water leaving the pile to contain radioactive materials with relatively short half-lives. (The half-life of a radioactive isotope, or material, is the time required for it to lose half of its radioactivity.) The design provided for conducting this water underground to a retention basin for its final radioactive decay. We took this step to avoid any possibility of injury to fish in the Columbia River. We were certain the dilution would ensure the safety of the human population downstream.
Shortly after the Hanford site was selected, I had talked to Robins, who had built the fish ladders and elevators at the Bonneville Dam, and outlined the measures we were taking to protect the salmon in the Columbia River. He made a lasting impression on me at that time when he said, “Whatever you may accomplish, you will incur the everlasting enmity of the entire Northwest if you harm a single scale on a single salmon.” As it turned out, we did not.
The concrete side walls of the retention basins were designed to extend high enough above the ground to prevent anyone within a critical distance from being exposed to radiation. To avoid any turbulence in the river, the discharge lines were brought into the main stream at an angle to provide for a converging flow, and, to prevent fish from swimming up the discharge pipe, a minimum velocity of over twenty miles per hour was planned. In addition, all effluent was monitored continuously by instruments to make certain that the radioactivity was at all times within entirely safe limits.
There were four outstanding factors that controlled reactor design. These were: first, the hazards to life and health if the radioactive gases should leak out or if the shielding failed mechanically or was insufficient, thereby exposing a portion of the area to radioactive emission; second, the amount of heat liberated while the piles were operating at capacity, which might result in a spontaneous temperature rise beyond controllable limits, if pulsations or interruptions in the cooling water should occur; third, the characteristics of the specific materials within the pile that required cooling water of the highest quality obtainable; and, fourth, the completely unknown effects on the strength of construction materials of the continuous high bombardment of neutrons.
To obtain a maximum yield of plutonium, processing of the first uranium through the pile had to be completed several months before we began separating the plutonium from the leftover uranium and the other fission products. Each pile unit was made of carefully machined, very pure graphite blocks with built-in aluminum tubes which were charged or loaded with uranium in the form of small cylinders or slugs. Since the piles were water-cooled, we were greatly concerned about the effects of corrosion, for it was estimated that the failure of only a small percentage of the tubes could cause the complete failure of the pile.
All design was governed by three rules: 1, safety first against both known and unknown hazards; 2, certainty of operation—every possible chance of failure was guarded against; and 3, the utmost saving of time in achieving full production. The complications were many, for many pieces of equipment weighing as much as 250,000 pounds each had to be assembled with tolerances more suitable for high-grade watchmaking. It was through the assistance and the strength of the industrial companies of America that du Pont was able to solve the hundreds of difficult design and material problems that had to be mastered.
It is hard now to realize how difficult some of our developmental problems were. The aluminum tubes illustrate how complicated even the most simple item could become. Seven months of persistent effort, principally by the Aluminum Company of America, were required to perfect a metal of the proper characteristics so that a satisfactory tube could be produced in quantity.
Although experimental development was started promptly it was only a few months before the first pile operation started that we were able to perfect the very special techniques required for the canning of the uranium slugs which went into the reactors.
The shielding for the pile was another problem. Ten months of work went into this before we could even begin to build it, with three more months before the first unit was completed. First, scientific requirements had to be adjusted to available, usable materials; then the program had to be outlined for design and procurement. In the course of this, a special high-density pressed-wood sheet was developed in collaboration with an outside supplier. Then special sharp tools and operating techniques were required to cut the various shapes from the standard manufactured widths. Next came the procurement of the thousands of tons of steel plates and the millions of square feet of pressed wood. At the same time the very detailed specifications for the assembly, prescribing the closest of tolerances, were written. Some sixty manufacturers were invited to bid and refused, presumably because of the complexity of construction and the close tolerances required, coupled with the short time available to develop sound fabricating techniques; but after methods were developed and prototypes fabricated at du Pont’s shops in Wilmington eventually satisfactory suppliers were found.
We also encountered an extremely difficult problem in the welding of the steel plates surrounding the piles. This work had to be almost perfect. An average superior job would not do. It took months to perfect the techniques required. We created a special super-classification of welders with premium pay. To hold this classification, a welder was required to take special training. He then had to take practical examinations at regular intervals to make certain that the quality of his work would remain up to what we needed.
For one part of the reactor ordinary materials had to be converted into unusual shapes with extremely close dimensional tolerances for the sizes and weights involved. There were a number of most unconventional assemblies, such as the control and safety rods, special instruments, slug-charging and discharging mechanisms, a heavily shielded elevator cab, and the entire cooling system. These items all required the utmost in careful design and diligent inspection of every minute detail.
In order to insure an adequate water supply for each pile, completely independent water facilities were provided, each with duplicate lines. At the same time, all individual units were cross-connected. Arrangements were made for driving the water pumps by either electric motors or steam turbines, so that in case of a power failure from either source, a safe amount of water would still be provided. In addition, there were emergency elevated tanks with automatic cut-ins, in case the normal supply failed. These elaborate precautions were necessary to permit curtailed operation for the time needed to correct any source of trouble.
The piles themselves were surrounded by heavy shields of steel, pressed wood and concrete, to protect the operators from the extreme radioactivity that accompanies the formation of plutonium. The energy of this radiation is equivalent to that of hundreds of tons of radium.
Each pile was located within an area of one square mile. At first they were six miles apart. If additional piles became necessary, we planned to intersperse them between those already in existence, so that the distance between them would then be three miles.
Design of the equipment for the chemical separation plants had to keep abreast of, and in some cases ahead of, the development of the process itself. Fortunately, the two separation processes that seemed to offer the best prospects employed virtually the same equipment and piping layouts, so that it was possible to go ahead with a design the first stages of which would be suitable for either process. Almost every one of the major design decisions for Hanford had to be made before the Clinton pile was in operation.
Originally eight separation plants were considered necessary, then six, then four. Finally, with the benefit of the operating experience and information obtained from the Clinton semi-works, we decided to build only three, of which two would operate and one would serve as a reserve. I should like to point out that these separation plants were designed when we only had sub-microscopic quantities of plutonium. Here again, each plant was provided with its own water system and steam plant and the other service facilities needed for independent operation. Each plant was a continuous concrete structure about eight hundred feet long, in which there were individual cells containing the various parts involved in the process equipment. To provide protection from the intense radioactivity, the cells were surrounded by concrete walls seven feet thick and were covered by six feet of concrete.
In use, the equipment would become highly radioactive and its maintenance and repair would be difficult, if not impossible, except by remote control. Consequently, periscopes and other special mechanisms were incorporated into the plant design; all operations could thus be carried out in complete safety from behind the heavy concrete walls. The need for shielding and the possibility of having to replace parts by indirect means required unusually close tolerances, both in fabrication and in installation. This was true even for such items as the special railroad cars that moved the irradiated uranium between the piles and the separation plants. The tracks over which these cars moved were built with extreme care so as to minimize the chances of an accident. Under no circumstances could we plan on human beings directly repairing highly radioactive equipment.
After being discharged from the reactors, the uranium slugs were kept under water continuously, then sent on the specially designed railroad cars to an isolated storage area. There they were immersed in water until their radioactivity had decreased enough to permit the separation of their plutonium content by chemical treatment.
Following the removal of the plutonium the residues were still highly radioactive. They still had to be handled by remote control. These waste materials included the process solutions from the separation plant. They were finally placed in steel-lined, reinforced concrete tanks buried in the ground to guard against their possible injurious effects. Special provisions also had to be made to take care of the heat that was constantly being generated by this waste material. In the beginning sufficient waste storage capacity was installed to handle one year’s operation, but as the piles continued to operate, additional storage had to be provided. We had always thought that it would be possible by intensive research to eliminate much of this radioactive problem in the future. We also hoped to recover the uranium remaining in the existing solutions and to reduce the bulk of the radioactive waste materials, thus making them easier to handle.
In designing the plants at Hanford and elsewhere, all possible care was taken to safeguard the health of the people who would be working in them.1 Inevitably, this consideration enormously increased the time—which was most important—and the cost, but we strictly followed the policy that where knowledge was lacking, every imaginable precaution must be observed. While placing great reliance on the Chicago laboratory for the adequacy of design from the standpoint of safety, du Pont did not delegate its responsibilities in this area. It sent its own medical and health personnel to Chicago for indoctrination and training, and also borrowed from the University an experienced roentgenologist and a health physicist. Thus, it was able to strike a proper balance between operating needs and safety requirements.
Experience gained at the Clinton laboratories indicated that the danger to anyone outside the immediate operating area would be much less than we had originally feared, but that the danger from the toxicity of the final product was considerably greater. Not too much was known about the human body’s tolerance for neutrons. But their danger was realized and all the necessary steps were taken to ensure the safety of everyone who might be subjected to them.
Radioactivity was always a serious and extremely insidious hazard, for without special instruments it could not be detected. There were three types of radiation involved: alpha, beta and gamma. Alpha and beta rays are high-speed particles of very short range and limited penetrating power. Gamma rays are long-ranged and possess great penetrating power. They produce changes in the human body by ionizing the atoms within the body, thereby destroying or damaging body tissue. Those body cells which multiply rapidly, such as bone marrow, are most easily affected by gamma radiation, while the slower-growing cells are relatively unaffected. Beta rays affect primarily the tissues that are close to the surface of the body. Gamma rays, on the other hand, affect all body tissues and an overdose was certain to be disastrous, for no medical cure for its effects was known.
The National Advisory Committee on X-Ray and Radium Protection had established a tolerance dose for gamma rays at one-tenth of one roentgen per day. Because this was not definitely known to be safe, the tolerance dose at Hanford was set at one-hundredth of a roentgen per day. This dose could be absorbed in a short period of time, or over an entire day, so long as it was not exceeded within a period of twenty-four hours. It was calculated that one foot of lead, seven feet of concrete or fifteen feet of water would provide adequate protection from the maximum radioactivity to be expected during the operation of one of our reactors.
Each part of the plant where radiation could be a factor was studied separately and designed to make certain that the safety precautions were adequate. For example, a great majority of the process pipes in the separation plant were buried in concrete and designed to prevent the escape of radioactivity along straight paths. After operations started we found that the safety measures had been wisely taken.
In addition to all the other precautions, du Pont designed a control system for the piles that we thought would be safe no matter what happened. It consisted of three distinct elements: first, the control rods could be moved either automatically or manually into the side of the pile; second, safety rods were suspended above the pile so that, in an emergency, they could be instantaneously released; and third, as a last resort, arrangements were made to flood the pile with moderating chemicals. This last device was designed for remote operation from a shielded control room. If this safety device had to be employed, the pile would no longer be usable.
Besides the two main features of the plant—the piles and the separation works—we built a special pile to test the graphite, uranium and other materials that would be used in the main reactors. Similar to them in principle, it was operated at very low power so that it did not require extensive cooling and shielding, and its construction was therefore much less complex even than the pile at the Clinton semi-works.
In the same area that contained the test reactor there were facilities for extruding the uranium billets into rods and then machining the rods into slugs and finally for canning the slugs. There were also extensive laboratories and a semi-works separation plant for the study of the separation process, using radioactive materials.
To save time, all the uranium rods needed for the first loading were extruded off the site, but practically all machining and canning were done in this area. Because of our lack of knowledge and our feeling that the loading operation might be quite hazardous, a number of new instruments were devised and installed. Since even the machines and tools for their manufacture did not exist, we had to design and to assemble many of them at the plant and to provide for necessary maintenance, including periodic rebuilding. For this purpose we set up a large instrument shop.
We were fortunate that Hanford was served by adequate electric power. Grand Coulee was able to make 20,000 kva available immediately and could supply our entire power needs by September, 1943. Bonneville had no surplus power at the time, but new generator installations were in progress. Although the bulk of this increased capacity had been committed to the aluminum industry, we were certain that if an emergency should cut off Grand Coulee, we would receive an adequate amount of power from Bonneville.
Within the Hanford site, we had to build over fifty miles of 230,000-volt transmission lines and four step-down substations. Because of the copper shortage, the War Production Board thought that we should use aluminum cables. We encountered considerable difficulty and delay in obtaining any decision from the WPB on this matter and it was not until July, 1943, that procurement could begin. Then only the most vigorous expediting enabled du Pont to obtain the material in time to meet its construction schedules.
Added to all these problems was the very urgent one of providing adequate living accommodations and the essential community services for all the Hanford workers. The small village of Richland had been acquired as part of the site and was used as a base upon which permanent living quarters could be built. Provisions were made for the permanent housing of some fifteen thousand people. The main administration buildings, the central service facilities for the plant, and all the other structures normally required by such a project were located in the village. Among other things, we had to build laboratories, storehouses, shops, change houses, fences, electric, steam and water lines, sewers and storage tanks, as well as hundreds of miles of roads, railroads and distribution lines.
In all that has been written about the Manhattan Project, little attention has been given to what life was like for the thousands and thousands of people who worked at one or another of our wartime plants. Life in each had its own unique aspects but certain factors were common to them all—isolation, security restrictions, spartan living conditions, monotony. It was perhaps hardest, in many ways, on the women. Hanford affords a good illustration of what I mean. Here we employed several thousand women in various capacities—as file clerks, stenographers, secretaries and so forth.
To look after their welfare and to see that they were as content as possible under the circumstances, we were fortunate enough to secure the services of a remarkable woman, Mrs. Buena M. Stein-metz, then Mrs. Maris and the Dean of Women at Oregon State College. It was her responsibility, as Supervisor of Women’s Activities, to deal with the problems that are bound to arise when a large population of women are housed as a group in an isolated area under rugged conditions, with few of the amenities of normal life.
Admittedly, our concern with morale was not entirely altruistic, for a stable clerical force was essential. We simply could not afford a constant turnover. The trouble was that employees found it easy to get jobs in Yakima, Seattle and other near-by cities where living conditions were far pleasanter. The turnover hazard started on arrival.
Recruited for a highly paid wartime job in the “great Northwest,” and transported at considerable expense, often across the whole United States, many of these girls and women arrived at Hanford with unrealistic expectations. Disillusionment sometimes set in almost as soon as they got off the train at the railroad station in southeast Washington, in the middle of the night. Weary from long travel in day coaches crowded with wartime traffic, expecting to get into their quarters at last and have a bath and a good night’s sleep, they found instead that they still had another forty-five-mile trip by bus to Hanford. There they had to bunk the rest of the night in a reception center, without their luggage and with the barest comfort facilities, and then face a day of employment routine before they could seek the relaxation of bath and bed. For many of the new employees the greatest disappointment was to find that Hanford was out in a sagebrush “desert”—when they had dreamed of Washington’s famed evergreen forests.
To offset their natural disgruntlement, Mrs. Steinmetz held an induction session for each group of new arrivals to impress them with the overriding importance of the work being done at Hanford and to help them realize that in their jobs they would be contributing directly to an unprecedented effort to bring the war to an end. Since most of the women had someone they loved in the armed forces, this was an effective appeal. Inevitably the effect was not permanent in every case, but whenever one of the women decided to leave the project, Mrs. Steinmetz would endeavor to find out why she was leaving. In this way she picked up many clues that were useful in bettering the morale of the other employees.
In addition Mrs. Steinmetz served as a kind of wailing wall, maintaining day and evening office hours during which the women could come to her with their complaints and troubles and sorrows, knowing they would find comfort and sound advice. Also, in each of the barracks in which the women lived there was a responsible hostess, or “house mother,” who gave them a further feeling of security and support.
The du Pont administrators at Hanford, she recently wrote me, “were unlike any team of administrators I have ever known in their complete commitment to the job to be done. Despite the pressures on them and the burden of making vital decisions daily, their doors were always open to me, and they apparently always had time to consider any problem I thought important enough for their attention.”
What never ceased to amaze her, she added, was the promptness with which they acted. For example, the area between the gate in the barbed-wire fence that surrounded the women’s compound of twenty or so barracks, and the barracks themselves, was covered with river gravel as a protection against mud and dust. But the gravel in short time wrecked the shoes of anyone who had to walk on it frequently—and shoes were rationed. After a while there was a ground swell of protest. On one of Read’s visits to Hanford he asked Mrs. Steinmetz, “What do you need?” “Sidewalks or asphalt,” she answered, thinking that sometime they might get something. To her astonishment, trucks were spreading asphalt the next day. A small thing, perhaps, but it kept us from losing employees we could not afford to lose.
Concerned also with a wholesome program of activities for the families of men employees, who lived in trailer housing, Mrs. Steinmetz on another occasion arranged for the use of an old vacant church building in Hanford as a community center, and asked to have some necessary repairs made. Then she called a mass meeting of all the women at the project who would be interested in either Red Cross or Girl Scout work. The turnout was large, and hardly had the meeting got under way than a crew of workmen swarmed over the building; some of them mounted to the roof and proceeded to rip off the old shingles and clap on new ones, while others descended to the basement with a portable concrete mixer and poured a new floor for what was to be the community center kitchen. In spite of, or perhaps encouraged by, this unexpectedly prompt response to their needs, the women carried on amidst the din. Incidentally, the Red Cross and Girl Scout programs organized that day are still in existence at Rich-land.
Wartime communication was inadequate at best, and one of the greatest values of the Red Cross at Hanford—and the main reason that Mrs. Steinmetz persuaded it to organize a chapter there in the first place—was that it gave the employees, far from their homes as they were, an assurance that important messages from or concerning members of their families, in the services or out, would reach them promptly. It was a link with the outside world, and kept them from feeling totally cut off.
Another cause for dissatisfaction at first was that there were no clothing shops at Hanford, and only one inadequate beauty salon for thousands of women. To remedy this state of affairs, a good women’s apparel store was persuaded to install a branch, and a special five o’clock bus was put on the route to the town of Pasco, forty-five miles away, which made it possible for the women employees to have dinner in town occasionally, do some shopping, get their hair done, or see a movie and so forth. There was a late return bus for women only, the need for which Mrs. Steinmetz determined by making the late bus trip back to Hanford herself.
There was a small library on the site, and of course regular services were held for both Protestants and Catholics—those for Catholics being conducted in a large circus tent that had been put up to serve as a theatre for moving pictures. But restlessness arising from plain boredom was not easy to cope with at best. One popular measure was the arrangement made with the commander of the Navy Air Force contingent at Pasco for any of the women who wanted to attend the regular dances there, with Navy buses supplying the transportation.
From the problems of reactor design to the health of fish in the Columbia River and the condition of women’s shoes covers a considerable range of problems, and obviously they were not of equal importance. But they all mattered in the job we were trying to do.
1 See Appendix IV, page 420.