Projects, Activities, Facilities, Operations, and RORC, 1989-1994
At the job I held before coming to ORNL, I was part of a close-knit rad engineering group that was breaking up with the decline of the nuclear power industry. So at ORNL I hoped to become part of a new work “family”. I was anxious to learn the ways and traditions of ORNL and especially of the ES&H organization. I thought that the first order of business for me should be to read the ORNL rad protection procedures and other such documents to get up to speed on how the site was run. I also wanted to find out who did what functions.
ALARA Program Working Space and Conditions
With the importance laid on meeting the requirements of DOE Order 5480.11, I was surprised when I first arrived at ORNL to find that because Setaro was so busy with an ad hoc project for Swanks, he could not spare time to sit down with me and get me started on my work for weeks. Because he is a nice guy, he kept apologizing for this. Temporarily, I was given a cubicle in a large room, whose other occupants were mostly secretaries and computer specialists. I had very little space to put my things because the cubicle was used by others for storage of extra manuals and other materials. The computer was set up for someone else’s type of work and nobody had time to teach me how to use it. Setaro told me that the ALARA Program staff was supposed to get office space, but none had been found yet; when it was, we would be getting our own furniture and computers with money already allocated in the ALARA budget.
The problem was that the ES&H organization had very little space to call its own and no way to commandeer any more. Swanks, his deputies, and some secretaries had offices; the bioassay and whole-body counting people and the instrument calibration people had specialized lab facilities; and there was one small ES&H meeting and training room. But nearly everybody else had loaned space in the areas they served. For example, each rad tech group or subgroup typically had a single office, not separated into cubicles, in one of the buildings or facilities they served. If a new person or set of persons was added to the ES&H organization — such as the ALARA Program staff — then one of the operating or research (O&R) divisions would have to grant them the use of an office or offices. This could be revoked on short notice by the division, with the ES&H people then having to move — at ES&H’s expense, of course. A new building devoted to ES&H was planned virtually all the time I worked at ORNL, but it was postponed year after year and never built. Because ES&H was always the poor stepchild with respect to space and related funding, we joked that we “got no respect”.
After some months, we finally got some loaned space in a Materials and Ceramics Division (M&C) building. Gheesling and eventually our secretary had offices on the second floor; Setaro and I were two flights down in the basement. The chemical ALARA people he supervised were in another building, since M&C had no room for them. There were hot cells above our basement; one night an M&C hot cell window leaked down into the office next to mine, which belonged to a radwaste division training person, and ruined her work shoes. I believe she had to buy a new pair out of her own budget. The radwaste division, too, as a support organization, was a poor stepchild of the Lab.
Before we got our own secretary, we were assigned to use one of the ES&H secretaries, who charged part of her time to our budget. When I first came, she was in the same building as we were, but once we moved, we had to walk two or three blocks to give her things or to get them. The secretaries had monopoly power over memos and letters, ostensibly to ensure that the proper form was used. So when we wanted to send a memo, we had to walk our computer-typed draft over to the secretary, leave it for her to type up (she was always too busy to do it while we waited), wait for her to call us to say that it was ready (as long as three days), and then go and pick it up. A cost-benefit analyst might say that this was a very inefficient arrangement, especially since our much more highly paid time was being used to perform messenger services, but Swanks blessed it explicitly. Once, I needed to get a memo out urgently, so I typed it up and sent it myself. I copied the form of an earlier, secretary-issued memo and I did get Setaro’s comments on it before I typed out the final version. The official ES&H copy was to be filed by the secretary, so I sent her that copy. When she saw it, all hell broke loose — I had put the date in the wrong place, or some such thing. Setaro apologetically told me that he had been told by Swanks to direct me not to do any more memos myself, but to funnel them through a secretary; I apologized back, but pointed out that if a style sheet were provided to us, we could easily take care of our own memos. Soon an official memo went out from Swanks to all personnel to state that all memos and letters were to go through the secretaries — to maintain format standards. So much for the budget, I thought.
With the advent of E-mail, widespread nonsecretarial use of computers for word-processing, and the shrinkage of secretarial resources, the “Secretaries Only” memo-issuing policy eventually went out the window. But for the year or so that the policy lasted, it was apparent that keeping the secretaries happy was more important to Swanks than supporting productivity. It was an early example of how people who were expected to work quickly and efficiently were nevertheless also expected to accommodate and defer to people who had more clout.
Early Projects: The Sr-90 Source Transfer and the 3001 Canal Project
There were other indications in those early years that fresh ideas, or any idea “not invented here”, were not welcome. This was despite DOE’s newly avowed preference for the sites’ hiring people with power plant or Nuclear Navy experience (DOE thought, correctly, that people with these backgrounds had a more rigorous and compliant attitude toward safety practices). Two early projects, which I will call the Sr-90 transfer project and the Building 3001 Canal project respectively, showed how hard it would be to change the old ways.
The purpose of the Sr-90 (strontium-90) transfer project was to transfer some Sr-90 sources into a hot cell. These sources were small — I don’t recall their size, but they were probably metal-encased cylinders a few inches long — and there were many of them. Setaro assigned Gheesling as the ALARA rep to the project. Gheesling went to talk with the O&R division, which for continuity’s sake I’ll call by its later name, the Chemical Technology Division (“Chem Tech”). He came back with disturbing news: Chem Tech was proposing to dump the sources into the hot cell from an opening in the top of the cell. Thus the sources would hit the in-cell table and bounce off in all directions. They would fall perhaps 4-6 feet onto the table and then several feet off the table, or they might miss the table and fall all the way to the floor. Either way, no damage would likely occur to the sources or the cell from the mechanical impact.
However, there was one caveat. When a hot cell has been in use for a long time and the window has received a lot of beta dose (from high-activity sources in the hot cell), an energy charge builds up in the window. If a window having a significant charge receives a sudden mechanical impact, the window may crack due to the sudden discharge of energy. This can result in a loss of shielding via two mechanisms. If the window breaks so as to create an open crack in it, the liquid between the windows (e.g., zinc bromide) may escape out the crack and the shielding provided by the liquid is lost. If the window actually fragments, then the shielding provided by the glass itself is lost too. In both cases, there is a chance of a suddenly high dose rate and flying glass outside the hot cell. A window crack had occurred by this mechanism years earlier during the time Setaro worked at a California isotopes reactor, and he knew of other instances. Thus he was very alert to the possibility with this hot cell operation. It was possible to pre-discharge the window and thus eliminate the hazard of a potential impact, but it was preferable for Chem Tech to unload the sources in a more controlled fashion. Setaro spoke with Swanks about this and Swanks supported him.
Setaro had come up with the idea of an “ALARA Plan”, i.e., an operational safety plan that focussed on the radiological aspects of a significant project or operation. Such a plan would be written by the O&R division, reviewed by the ALARA Program staff, and approved by the ALARA Program manager. The Plan would include not only the ALARA or dose-reduction aspects of the project or operation, but also what one might term the rad engineering aspects, i.e., how the particular physical equipment and personnel would be selected and used appropriately in order not just to reduce dose but to prevent the lesser accidents. (What I call the “drastic” or “serious” accidents would be covered under the facility safety analysis, which generally did not address the normal or lower-level occupational issues.) Here again I emphasize that the rad tech organization usually concerned itself only with the rad tech aspects of the job, such as monitoring the dose rate and contamination. They generally did not, for example, look at the system drawings to see where the leakage potential was; they would take the O&R people’s word for that. Thus in having a written plan that was reviewed by an engineering-minded staff such as the ALARA Program had, the gap between the tech concerns and the operational concerns would be bridged and there was less likelihood of overlooked weak spots “falling through the cracks” (pardon the pun).
A written ALARA Plan had several advantages. First, having a written document for the ALARA staff to look at provided a way of seeing in writing what was planned and not just hearing it orally; it allowed for the content to be frozen by approval signature and thus to have all parties clear as to what they agreed to; and it provided a place where hazards and risks could be discussed and the selection and use of “controls” — the protective measures — could be justified. It thus differed from a regular work plan, which dealt with all hazards, not just radiological ones, and which usually just stated the controls without explaining or justifying them. Second, having the O&R division write the plan meant that there was “buy-in” from the division and they had to think about and plan the radiological safety aspects of the operation, not just make it up as they went along. The division was also committing to do things according to the plan. Third, by making this ALARA “reviewable document” separate from the Radiological Work Permit (RWP), the ALARA staff would not have to inject themselves into the RWP process, which was controlled and carried out by the rad techs. There would have been resistance on the part of the rad techs and the O&R managers to having yet another signature to obtain on the RWP, which was often issued at the last minute before a project or operation began.
But the disadvantages of the ALARA Plan idea arose from the fact that its use could be limited and conditioned by the O&R people and the rad techs. The workers all had to read and sign the RWP before they did work, but the content of the ALARA plan was usually unknown to them. The rad tech complex leader had to review and approve the ALARA Plan — and thus be a party to the Plan — but the ALARA rep did not have corresponding authority over the RWP and thus the two documents could turn out to be inconsistent if the tech or complex leader writing the RWP did not consult or follow the Plan. And whereas the regular work plan and the RWP had to be formally amended or revised before they could be deviated from (in theory, at least), at some point we began to say, as an accommodation to keeping the work going when unexpected circumstances arose, that the ALARA Plan could be deviated from during an operation if in the opinion of the tech covering the job, circumstances warranted it. And we also said that while the ALARA Plan was a good-faith commitment, deviations from it would not be regarded as violations. (Why we said this will be discussed later with regard to the MSRE project.)
But in this earliest ALARA Plan, for the transfer of the Sr-90 sources, the ALARA Plan was taken to be a firm commitment, with Swanks’ backing. Chem Tech proposed three alternatives, one of them the “dropping from the ceiling” option. After conferring with Setaro, Swanks signed off on the two non-dropping alternatives. Chem Tech agreed to do it one of these two ways.
The lessons we learned from this — or that I thought we learned — were that (1) higher ES&H management would back up the ALARA Program staff when they disagreed with the O&R people, if the ALARA Program staff had solid reasons for their position, and (2) our rad engineering expertise could add value to the safety review process. But the road to earning respect for the ALARA Program would be a rocky one, as was shown by our next big project, the 3001 Canal project.
Building 3000 contained an old reactor no longer in use. The nearby Building 3001 had a fuel transfer canal that had been used in conjunction with the reactor. The canal, a 10-15 foot deep L-shaped trench in the basement, was filled with water and was used as a repository for all sorts of radioactive debris. The 3001 canal project consisted of characterizing, removing, packaging, and shipping out this debris. Although ORNL certainly possessed the expertise to do this with its own people, DOE, in accordance with its policy of spreading some of its work out beyond the (prime) contractors, let the contract to a subcontractor. ORNL was supposed to provide a degree of oversight, both of the operational planning and of the execution, and the subcontractor agreed to follow all ORNL rad protection procedures and policies. They used their own rad techs, but these were vetted by ORNL rad protection personnel and were to submit their measurements on a daily basis to the ORNL rad tech organization. This was, I believe, arranged before Setaro was involved in the planning. Thus we in the ALARA Program had no formally agreed-on oversight responsibilities and no formal reporting of any kind was to be made to us by the subcontractor, except that we were to review and approve the ALARA Plan for the project and have some vaguely defined entrée to the project once it started up.
I went to many meetings during the planning phase of this project in 1991. The subcontractor people were congenial at the start and expressed willingness to do whatever it took to satisfy the concerns of ORNL safety people. As I would see was true of nearly every future project as well, a DOE person attended perhaps the first meeting, but after that was seldom present. It was clear that the daily or in-person safety oversight of the subcontractor was in the hands of ORNL safety personnel. It seemed clear too that the subcontractor, having presumably submitted the low bid, needed to keep costs down in order to make any money. I can’t recall all these years later, but it seems to me that at that time we ALARA Program people were not charging them for our time, but were supported on ORNL overhead and so could spend as much time on the project as was needed, as long as we had sufficient time left for other projects and facilities.
The subcontractor people gave us a initial draft of an ALARA Plan. It was very sketchy because they were still planning the job and had not decided the details of the method, much less obtained all the equipment. I carefully marked up the draft to indicate the areas that were lacking and to ask questions. I made it clear to the subcontractor people that the Plan should cover not only narrow dose-reduction issues, but also radiological factors affecting appropriate choices and use of equipment, positioning and use of personnel, emergency measures, and like rad engineering areas. As the weeks passed, I reviewed four or five revisions of the ALARA Plan, each time marking them up and returning them to the subcontractor. The proposed method changed significantly a couple of times, so new questions were raised. The start date for the operation was fast approaching, so the lack of detail in the plan and the failure to answer questions that I repeatedly asked were beginning to concern me.
A significant omission was the area of contingencies, that is, what can go wrong and what will be done in response. For example, since the method settled on was an underwater vacuuming of the canal floor to get up small bits of debris, there was a possibility that a “hot” (highly radioactive) bit could be sucked up into the vacuum pump, which sat unshielded on a small platform above the canal. The hot bit could stick in the pump instead of passing through it and thus create a high dose rate in the canal area. The area was long and narrow — there were only a few feet for standing on either side of the canal. So although personnel could retreat back along the room and out via the stairs, sooner or later somebody would have to re-enter and do something about the hot pump. I raised this issue at one meeting that Setaro also attended. Everybody just sat there for a moment. I think somebody said that the pump flow could be reversed to chug the bit back into the canal, but somebody else said that if it was stuck, then reversal might not jog it loose. Setaro said, “I know what I’d do — I’d get on the platform and kick the pump into the pool”. This was a terrific solution: they’d have to get another pump and figure out a way to get the submerged pump out and into a shielded container eventually, but the person kicking the pump in would get little dose from the momentary exposure and the dose rate would drop instantly as the bit was shielded by the water. This would allow work to continue in the area while the pump retrieval was being planned. The subcontractor people agreed that this was a practical solution.
However, the subcontractor people did not want to commit to any plan of action for this clearly credible contingency. Indeed, they did not want to address contingencies at all in the ALARA Plan. They did not offer a reason, except for one subcontractor person’s testy “That stuff (contingencies) doesn’t belong in an ALARA Plan!” Setaro had been talking more and more rapidly up this point, trying to persuade them to specify contingencies that could be anticipated and to commit to training the personnel on appropriate response actions. In the spirit of agreement, he therefore answered the subcontractor impulsively, “Yes, you’re right, that doesn’t belong”. I was shocked and disappointed: he had given away the store. That is, we in the ALARA Program had defined what the ALARA Plan needed to contain and we were within our scope to insist that our questions be answered. Later I asked Setaro about this. He agreed that he had misspoken and that contingencies were appropriate in an ALARA Plan, as we had defined it. He implied that he had gotten somewhat carried away in trying to get the subcontractor people to meet us halfway. This explanation was credible because he was a person who was always trying to get folks to come to consensus so as to move things forward.
As time wore on, the subcontractor made change after change in the proposed method, but the changes were not always reflected in the still fairly superficial ALARA Plan. We were provided with all kinds of documents (system descriptions, work plans, etc.) in which the answers to our questions were said to be answered, but the subcontractor people did not want to put the answers into the ALARA Plan. It was clear that this was because they could change the other documents readily, with a minimum of review or even no review by ORNL personnel, but not the ALARA Plan, which they increasingly seemed to regard as unnecessary and a nuisance. I spent hours and hours going through the other documents to find any little kernels of radiological import. This was extremely inefficient, but I thought it necessary to ensure that the subcontractor, of whom I was becoming increasingly mistrustful, did not slip in any radiologically unwise actions. We were not the only discipline with doubts: the ORNL quality assurance person also had concerns about the reluctance of the subcontractor to be pinned down, as she told me near the end of the planning stage. When the start date was upon us, the subcontractor, working through their main ORNL divisional contact, the Engineering Division, insisted that they had provided an adequate ALARA Plan and insisted that it be signed as was. I believe that there was an implication that we in the ALARA Program would be blamed for the delay in starting if we did not sign. I thought that this was blackmail.
Setaro and I then met with Ed Krieg, director of the Engineering Division. Engineering, like ES&H, was a support division for the O&R divisions, and it had its own procedures and policies that the other divisions were required to follow. Engineering controlled most construction at the time, in the sense that the other divisions were required to have Engineering provide services and oversight for major modifications and new construction. Like ES&H, they were often viewed as parasites by the other divisions, who thought that they should be able to perform their own cost estimating, construction supervision, etc., or hire it to be done, and not have to involve Engineering. But Engineering ensured that the often byzantine DOE rules regarding construction were observed and that a certain detachment was maintained regarding, e.g., evaluation of changes in midproject. Krieg was thus in the dead middle of the conflict: sympathetic with our wanting to stand on our expertise and on the authority vested in us, but also sympathetic with the desire of both the division that “owned” the canal and the subcontractor to start the project.
I told Setaro and Krieg that I couldn’t sign the ALARA Plan with its present lack of detail and non-resolution of questions. I also said that the project people did not seem to me to be committing to what they did put in the ALARA Plan, i.e., they would junk any provisions that seemed inconvenient. I thought so because some of the changes in the other project documents were not reflected in the latest version of the Plan. Krieg was of the opinion that the subcontractor people could be trusted to do the right thing, even if they didn’t commit to it in writing. At one point, Krieg and I momentarily got testy with one another. Now, in 13 years of discussions with fellow engineers in my pre-ORNL job in Chicago, I had seen many pairs of engineers, usually of different disciplines, have testy moments in discussions like this. Sometimes I was one of the pair, but usually I was just an observer. From these experiences, I knew that Krieg’s and my discussion would progress to resolution or would at least subside into non-hostile agreement to disagree. Setaro, a nice man who, as I have pointed out, was non-confrontational and anxious to achieve consensus, kept trying to chip in to forestall what he perceived as a major argument. But Krieg and I maintained eye contact and did not let him distract us, and we each expressed our opinion in the matter. Then the tension did indeed dissipate as I had known it would. Setaro was obviously relieved and the meeting ended amicably. Later Setaro told me that he understood my objections and misgivings, but he and Krieg thought they could trust the subcontractor and even if they could not, ORNL would be providing oversight and would catch any deficiencies or deviations. Therefore, he said, he would sign the ALARA Plan himself. I was not happy about that but I knew that his intentions were good.
The project began. In the beginning, everything was done by the book and the ORNL rad techs told us that their opposite numbers were performing well. I was busy on another project, but Setaro and Gheesling went over periodically to see how things were going. Then at some point, visitors were barred, except by permission of subcontractor management, because of the increased contamination potential and the close quarters, which were partly taken up by equipment that now included multiple waste drums. As time went by, problems arose. The pump screen kept clogging up: they had used a biocide to shock the contamination-bearing algae off the wetted surfaces of the canal and the loose algal material was coating the filter as the water was vacuumed up. They were using a robot to locate and pick up larger pieces in the pool, but there were problems with its use too. They had various incidents, including having a quick-disconnect line come loose from its fitting and spray several workers standing on the walkway by the canal. We heard about these things mostly from the ORNL rad techs, since the subcontractor made it clear that we were in the way when we visited.
Later there were more intimations that things were not going well radiologically. The radiation doses were going up and the drums were coming out much hotter than had been agreed to originally. ORNL seemed to have no authority to curb this, or perhaps chose not to use it. I was given to understand by several people that DOE was pushing hard to get the project done despite the problems. I think that DOE was having to pay out more money than planned (perhaps under change-of-scope requests) because of problems such as the algal plugging that were not the subcontractor’s fault. As Setaro remarked, if DOE wanted it to be done, it was done no matter what, because they were the boss.
After the project was completed, an ORNL rad tech who had dealt with the subcontractor rad techs and workers told me that toward the end of the project, the work was racketing along so fast, in an effort to make up for the delays, that mistakes such as the disconnection leak were not infrequent. Setaro too heard these things through his own grapevine. Two of the subcontractor techs who worked on the project eventually came to work for ORNL and one expressed relief to me about the job change: not only were the pay and benefits better, but he felt ORNL was a safer employer to work for.
Setaro eventually commented to me that I was right and he and Krieg were wrong. He said that the oral assurances made to him and Krieg by the subcontractor were not kept and that they had been snookered by the subcontractor. I admired him for admitting this. One of his strengths as a supervisor was that he shared his mistakes as well as his triumphs: he mentored his people so that they would grow not only in technical knowledge but also in what you might call business savvy.
Despite Setaro’s admission, however, the 3001 Canal project seemed a bad omen to me. At my former job (at an architect-engineering firm in Chicago), one of the licensing engineers had a humorous sign on her door: “There comes a time in the life of every project when it becomes necessary to shoot the engineers and start production”. That is, the joke is that engineers want to tinker with and to re-design everything until it is ideally efficient, while as a practical matter things eventually have to get underway (besides which, the money allotted for the design and planning will stretch only so far). I took the meaning of that sign at once, since I was the kind of person who always wanted to know when my deadlines were and to have my “deliverables” ready by then. For my first ten years or so at this former job, our managers upheld the standards to which we were trained to work. Safety calculations were reviewed by peers and managers and engineering designs were subject to signoff by safety groups; disagreements between conflicting organizations were negotiated to resolution and our clients (mainly utilities) usually respected our expertise. But in the last few years I was there, the bottom was falling out of the nuclear power design industry with the decline in construction of new nuclear power plants. Increasingly, we were allotted a fixed number of hours to complete a safety calculation or review, without any input from us on how long it would take to perform the task. This arbitrary allotment was often significantly less than it would take if we performed up to our standards. I recall trying to produce a manual for a utility in only 80 hours. My supervisor for this project was a licensing person, not my usual safety supervisor, and we had not worked together before. When I told him how much of my own time I was putting in, uncompensated, to do even a barely adequate job, he told me, in effect, just to slap it together: “If they give you a Yugo budget, give them a Yugo product”. (The reference was to the low-quality but cheap East European car of the day.)
During the last year or two I was at that former job, there was a lot of anxiety among members of my group because we were being laid off one by one as the company lost work. Those of us still there were often loaned to the licensing section to do licensing work. Everybody had to account for every minute of their time and it had to be charged to a client, not overhead, so many of us were putting in significant amounts of unpaid time. The financial pressure, by forcing shortcuts and depressing morale, affected the quality of the work. So when I decided to find a new job before I could be laid off, I hoped to find a place where this sort of shrinking the job to fit the time allotted and nickel-and-dime-ing in time accounting were not done. I thought I had found that at ORNL. But the 3001 Canal project showed that my work group was relatively powerless and that when push came to shove, the deadline would likely trump everything else, with rules and commitments pushed aside.
In this and subsequent projects, I tried to be sensitive, as Setaro and Krieg were, to the need of the divisions and subcontractors to arrive at a reasonable point of agreement in planned safety coverage and then start the work. Setaro, as in this project, tended to credit people’s stated intentions and to assume that they were gentlemen who would keep their word. But I thought it was essential to require the entity planning the job — the O&R division or subcontractor — to answer the safety people’s questions and not put them off indefinitely, as happened in this project. It was also essential to require them to commit to doing the work the way they said they would, with appropriate renegotiation provisions. Otherwise, safety people’s authority to review and approve the work was just a joke and the commitments on paper were just so many trees killed in vain.
I highlight this conflict between the safety people’s authority and the operational entity doing the work because it is a principal topic of this book. I will be giving more examples in future chapters.
Reactor Work
One strength that I brought to my ORNL work was my familiarity with reactors because of my master’s degree in nuclear engineering and my 13 years in the nuclear power world. There were six or seven old reactors at ORNL, five of which I had something to do with over the years. The main one was the High Flux Isotope Reactor (HFIR, pronounced HIGH-fer). This 100-megawatt neutron research and isotope production reactor was arguably the second in the world after France’s Institut Lauë-Langevin reactor. I was associated with HFIR as the assigned ALARA rep from the ALARA Program and ORP and as a member of the Reactor Operations Review Committee (RORC).
When I came to ORNL in 1989, all of the reactors at ORNL were shut down. This was because of the “coupon incident” at HFIR. In order to measure the potential deterioration of underwater components, some test material — the “coupons” — was used. The coupons were placed at representative areas around the core and associated components to receive dose as the core and components did; a few of the coupons were removed from time to time and sent for analysis. In this way the deterioration of the reactor materials could be monitored by the surrogate coupons. However, at some point in 1987 it was discovered that the coupons had not been removed for quite a long time past their due date. DOE apparently got agitated at this failure to fulfill a safety commitment. ORNL then shut down not just HFIR but also all its other reactors and DOE began an investigation. I do not know all the details of how this transpired because it occurred before I came. Setaro liked to point out that it was ORNL itself, not DOE, that shut down the reactors, but someone else once remarked to me that had ORNL not done so, DOE would have. At any rate, many ORNL people spoke of the shutdown as DOE’s having to exact its pound of flesh, because DOE kept HFIR down for three and a half years, a period I call The Great Shutdown. During this time DOE required a lot of documentation from ORNL to show that ORNL had learned its lesson; a “readiness review” also had to be performed before startup.
The other reactors either never went back up again or went up for a few limited experiments and then were shut down for good. These included the Bulk Shielding Reactor, the Tower Shielding Reactor, the Health Physics Research Reactor, and the Oak Ridge Research Reactor. (The historic Graphite Reactor had been shut down and converted into a museum years earlier.) To a health physicist, these were names to conjure by — all of these had been used to make significant contributions to nuclear experiments, to radiation shielding studies, to training, or to radiation dosimetry. For some functions, one or another of these reactors was unique. But over the years following the shutdown, DOE shortsightedly invoked budget strains as the reason for not continuing to operate these reactors.
Contamination Incidents and The Contamination Control Initiative
Soon after I arrived, an incident occurred at HFIR that resulted in a great deal of trouble and concern. (My memory is fuzzy on how this began, so I warn everyone that what I write here about the start of the incident is not necessarily correct.) A researcher was found not to have “monitored out” (checked his skin and clothing with a detector for contamination) before leaving HFIR. The rad techs thus had to check his car and home with their detectors. They found that a couple of pairs of his shoes were contaminated. Other people were also found to have failed to monitor out, so their homes also had to be checked. DOE rightly became exercised about this and insisted that ORNL start a widespread check of people, items, and areas. This was called the Contamination Control Initiative by ORNL.
The name was misleading because it implied that this was an idea of ORNL’s or that ORNL had wanted to do it. In fact, both the rad protection organization and the O&R divisions regarded the all-ORNL survey as an unnecessary nuisance. However, it did have several beneficial effects. First, the handheld instruments used at ORNL were mostly of types that had been built at ORNL or bought years ago. (The reader should recall the Tiger Team observation about ORNL’s use of ORNL-designed instruments.) These instruments still worked, but their detection efficiency and ease of use were inferior to newer instruments, particularly the “pancake detector” that was standard at nuclear power plants and was becoming widely used at our fellow DOE plants. Because ORNL wanted to show DOE that it was serious about detecting contamination and also because more instruments were needed for the intense monitoring effort, ORNL started to buy pancakes.
Second, ORNL did not have any of the new half-body monitoring instruments, such as the Eberline PCM-1, that could monitor the body top to bottom accurately in about a minute, versus the several minutes or more, inconveniently and with lower accuracy, that was typical of handheld detectors. An Eberline rep sent us a PCM-1B on spec, as it were, for use at HFIR. He wasn’t paid for about nine months for this instrument, and at one point I asked him if he was worried about that. He said that he was not: every site that used one for a while was completely sold on it and came to feel that they “couldn’t live without it”. Not only that, but he usually got orders for more soon after. He was right — everybody at ORNL preferred the half-body monitors and more were obtained in time. These space-efficient, quick-counting, and virtually passive-use instruments were a decided improvement over hand frisking, especially when there was a line of tired people trying to get out at the end of the day.
Third, the all-ORNL survey turned up many contaminated items and areas that had been thought to be clean or nearly clean. This discovery was partly due to the improved monitoring instruments, but it was also due just to checking things that hadn’t been checked in years. Dealing with the items and areas was time-consuming and expensive, but certainly many items past their useful life were properly gotten rid of and many useful items and areas were recharacterized and reposted. Most of all, there was an improved “contamination awareness” at ORNL.
The initiative had one casualty. For a week, workers in the Biology Division were supposed to be checked by a mobile half-body PCM on their way out of their area. One researcher was found to have contaminated shoes and I think clothing too; he had been careless in using radiotracers in his laboratory. The local rad tech complex leader told me that he had found contaminated bench counters and equipment in the researcher’s lab on various occasions, when these counters and equipment were supposed to be clean. I believe he had found the floor to be contaminated as well. So this researcher was already known to be careless compared to other researchers. But he was also found to be publishing papers at what was said to be a very high rate, which we were told indicated a high productivity but also suggested to his management that he was cutting safety corners (and possibly sacrificing accuracy as well). Despite all this, I thought his punishment for being caught with contaminated shoes was too harsh: his division director forced him to retire.
During the Contamination Control Initiative, meetings were held most days in the ORNL shift supervisor’s office. This was attended by some rad protection staff, usually a DOE representative, and some DOE subcontractors who acted as advisors and reviewers of the results. The DOE rep during most of the time was a nice but rather anxious young woman who (as was so often true of DOE people) had a degree in a field other than the one she was providing oversight for. In her case, it was a degree in chemical engineering, while she was overseeing rad protection. At one point, the toilet seat in the ladies’ room at the REDC facility was found to have some fixed alpha contamination on it. The DOE rep got agitated and insisted that this situation be investigated to find out how the seat had gotten contaminated. When she was not around, the others at the meetings — all men except for me — spoke disparagingly of her anxiety about this. I thought she was overly intense about it, but she certainly had a point. The bathroom and especially the toilet should have been squeaky clean in this high-alpha facility and the toilet should have been cleaned often enough that contamination would not become fixed. After a few days, the theory was put forward that the contamination must have come off a workman’s shoe when he stood on the toilet to change a light bulb. There was no real concern expressed by rad protection management about this; it was spoken of very matter of factly, as though it were “just one of those things”. The DOE rep accepted this, but she obviously was not too happy with the attitude that one shouldn’t think such occurrences meant anything in terms of contamination control. She seemed too cowed by the ORNL people’s barely concealed impatience with her to press them regarding how they were going to prevent a similar occurrence in the future.
I attended most of these meetings because my assignment was to prepare the “body sheets”. These were a list of all persons whose skin or clothing had been found to be contaminated during the Initiative. Included were their badge numbers, the location on the skin or clothing where the contamination had been found, etc. (This list — a large pile of paper — was still in my file cabinet when I was laid off in 2000. I would ask every couple of years if I should archive it, throw it away, or keep it, and I was always told to keep it in my own files.) I also had to check with the complex leaders about how the contaminations might have occurred and about missing information. In doing so, I sat down for a long talk with each of them. They were all men at the time and they mostly seemed to me to be capable and cooperative. One in particular was not considered to be a simpatico person by others. But as we sat in his office — through which his techs trooped in and out and in which the phone and radio were continually interrupting — I noticed that he was very efficient in prioritizing the work as it came in and in dispatching his guys hither and thither. They seemed to get along with him also. He observed to me sadly at one point that he was just getting through until retirement: the fun was gone with all the hassles of new DOE requirements. He was patient with me as we went slowly through his complex’s section of the list, which took about 90 minutes with all the interruptions. He obviously could not afford to be incommunicado for the 30-40 minutes that it would have taken without interruptions because he was constantly making judgments about what the techs should do and instructing them, e.g., whom to meet where as per the calls. I wondered later why he was so nice to me (with his reputation for being cranky) and so non-defensive, unlike some of the other complex leaders. I think it was because we were talking “man to man”, so to speak; my questions about how the contaminations might have occurred seemed interesting to him, and he could see that I was taking his observations seriously. It struck me that perhaps a reason for his crankiness elsewhere was his feeling of being burned out and marginalized. I thought it sad that he felt that way, considering the length of his career, the number of people he supervised and thus influenced, and especially his ability to manage a bustling beehive like his complex office.
Tracking out of contamination had occurred at Brookhaven National Laboratory not too long before it occurred at HFIR. There, they had quickly prepared a procedure for checking people’s homes with radiation detectors. I was asked to prepare one for ORNL and I did, after talking with my counterparts at Brookhaven. But as it turned out, my procedure was not used. Because of the scope of the ORNL-wide survey effort, not all of the surveys could be done by the rad techs, so other rad protection people were pressed into service. Thus one day at the entrance to the HFIR site I met one of the (professional) instrument calibration people doing shoe checks. Another day, I found that the junior internal dosimetrist had been pressed into service to do home surveys. The latter, I discovered, was using his own sensible checklist-cum-procedure and he had no idea that I had also produced one. There were several other times during the Initiative when two different people found themselves assigned by different managers to do the same thing. There was thus a “who’s in charge” aspect to the whole thing that I found disconcerting.
This was also one of many times that I or other persons prepared a “product” for rad protection use that was not put to use at all. Another instance of this was in early 1991, when at Setaro’s request I produced a set of vehicle and clothing contamination guidelines as part of the beginning of the Contamination Control Initiative. It was actually a technical basis for determining the actions to be taken at certain levels of contamination. I spend many hours on this, including a lot of my own time. I gave it to Swanks at a meeting Setaro arranged. That was the last I ever heard of it. I believe, from comments made by Setaro, that the rad tech organization preferred not to have a technical basis and instead to use what seemed to be a few criteria chosen on the basis of convenience.
We had some questions in 1990 about how best to monitor for alpha contamination, given DOE’s confusing and hard-to-follow guidance at the time. I called around to most of the other DOE sites to find out how they did it. This was such a hot potato that the people I talked to asked me to tell them what I found out from the other sites. At the suggestion of the inimitable Al Tschaeche, then of Idaho National Engineering Laboratory, I organized a meeting for DOE contractor reps at a national professional meeting. This was not productive of any action item, but it did allow people to express their frustrations and I got to meet some fellow DOE professionals. I was struck by the fact that at the other sites, the senior (professional) health physicists and rad engineers were the ones who were setting monitoring criteria and the like; problem-solving and decision making activities like these seemed to be what these people did all the time, along with participating in work planning and like activities. But at ORNL, activities such as criteria-setting seemed to be functions delegated by Butler to people on an ad hoc basis, including complex leaders whose main duties were supervising rad techs. I had the impression that many of these people were dabbling in such activities, as though the activity were assigned as a perk and not always because the person assigned had expertise in it.
For example, I certainly agreed that the complex leader in charge of, say, the monitoring in a given facility should be consulted about how the area radiation monitors should be set and operated; the practicality and workability of any change had to be considered and the complex leader would be the best person to speak to that issue. But it seemed to me that in the choice of monitors and how they operated, the local rad tech supervisor had a virtual veto power even when he was relatively ignorant of, e.g., setpoint choice. If he declared he didn’t want something, it generally wasn’t put in. This is a somewhat exaggerated statement on my part, since certainly the rad tech supervisors had some things imposed on them and had to put up with things they disagreed with. But it was certainly also true that often a veto or a delay was provided by someone who would not have had such power at other sites.
On 29 December 1992, an incident occurred in which some 17 people became contaminated in the course of a project to install a new liquid low-level waste collection system. Only one was an ORNL employee; the rest were employees of another prime contractor, MK-Ferguson (MK-F), or its subcontractors. The problem was made even worse because almost half of these people had left to go home or go shopping when the contamination was discovered. ORNL rad protection personnel were dispatched to about seven homes, some out of town and even out of state, and to perhaps ten businesses. Note that while ORNL was not responsible for oversight of this project — MK-F was, being a prime itself — DOE apparently directed ORNL to respond in this emergency effort. Although this extreme case of loss of control did not result in a significant dose to anyone, the potential was obviously there. In these early days, nobody was fined and, it seems, nobody was really penalized for the loss of control. This was even more true of subcontractors and prime contractors who did not run the site — DOE seemed to have much lower standards for their safety performance than they did for a site-running contractor like MMES. Thus DOE’s penchant for distributing work among various companies also distributed responsibility and diluted control.
I had written Swanks a memo in 1991 on the subject of the MK-F-ORNL interface. In it I expressed concern about their independence (as a prime contractor) and cautioned that we needed to exert oversight of their activities on the ORNL site and come to an understanding with them as to the ORNL “rad rules”. He never answered. The incident above thus served to heighten my concern about MK-F and about ORNL’s commitment to safety.
In June of 1994, there was another incident involving MK-F. This time an MK-F vehicle was coming out of an area posted for contamination. There seemed to be a lot of mud on its tires, mud that had to have come from the inside of the contamination area. An ORNL rad tech saw this and was bothered by it. As he ate his lunch, he mulled over what he had seen. Finally he couldn’t stand it any more: he sneaked over to the vehicle and checked it with his detector. It was indeed contaminated. ORNL reported this to MK-F and apparently they agreed that the vehicle was contaminated when it shouldn’t be. But MK-F did not report the incident as an official occurrence, as was required by DOE Order. I sent Sims a memo about this, suggesting that the tech be given an award for his watchdog action, which might have prevented a worse incident. I don’t think he was given any recognition by Sims, but I believe that his complex leader, which whom I discussed this, included it in his performance review.
Reactor ALARA Reviews
A notation in my 1991 appointment book says that there was “an urgent ALARA review meeting at HFIR”. I was expected to drop all other work I had planned for that day and rush out to HFIR, which I did. This was only the first of many times that I was called to come on short notice to do a rad review of a HFIR job or to participate in one. I didn’t mind it the first time — after all, things do sometimes come up suddenly. But I became impatient with its happening so many times afterward. This was, after all, a facility that had a daily planning meeting and computerized scheduling — how was it that they couldn’t find time to give advance notice to safety people required to participate in reviews? The answer that was given to me almost every time by the rad techs was that the rad techs themselves had not been told until the day before that the job was scheduled for the next day (although they usually did know some days in advance of the expectation to do the job soon) and that since HFIR was a 24-hour facility, the operations people needed to be free to do new jobs as soon as other jobs were completed. This was a crock, of course, since even then we knew that at other DOE facilities, the operational people were required to “book” reviews in advance. But at ORNL, with its tradition of operational people calling the shots, the rad protection organization did not insist on timely notification.
Harassment of a Radiation Control/Safety Officer
In November 1992 a fellow rad protection person who worked at HFIR told me about an incident involving the Research Reactors Division’s (RRD’s) Radiation Control Officer, who was also its division safety officer (DSO). I later spoke with the DSO about it. Both my source and the DSO said that the DSO had had to remind some craftspeople to put on their safety glasses in their shop at HFIR. They made a running joke of it, saying “Here comes the DSO, time to put on our safety glasses”. The DSO documented various instances of his having to make them put on the glasses and he sent a memo to the P&E (crafts division) DSO, who forwarded it to a P&E manager, who forwarded it to a crafts supervisor. The crafts supervisor took the memo and “chewed on the shop workers” over it; he apparently also posted it on a bulletin board. A few days later, somebody called the telephone company and fraudulently stopped service to the RRD DSO’s house. Someone also put a classified ad in the newspaper offering the DSO’s car for sale and giving his telephone number. In one of these instances, the person was required to give the DSO’s Social Security number, which he did. The DSO’s mother, whose health was not good, lived with him and his wife. So the temporary loss of telephone service was not just an annoyance to the DSO, but a real concern to him with regard to emergencies.
The person who canceled the telephone service could not be identified, but the newspaper had recorded the voice placing the ad (perhaps on an after-hours answering machine). The person’s voice was recognized by his management and he was identified as a union steward. The DSO was quite surprised that the steward was involved: as the DSO told me, this man was a good worker who was an RRD management favorite to work because he would stand tough conditions (such as sweating for long periods in a respirator) without complaining. Others were apparently involved also, but the harasser refused to name them. The DSO considered taking legal action but did not because, he was told, ORNL would deal with the problem. The DSO told me that he was opposed to having the harasser fired; the DSO was a good guy, a Christian person, and as a former high-ranking Navy non-com, he respected the chain of command. Besides, he felt that P&E had not previously taken a firm position on the wearing of safety glasses and had made him the heavy in enforcing safety rules. All he wanted was for the harasser to be firmly disciplined and to be moved out of HFIR work. After some time, however, a crafts supervisor approached the DSO and asked if he could continue to work with the harasser. As the DSO told me, his response was, “Could I work with him? Yes. Did I want to work with him? No. Did I think he should be moved? Yes. One of us should be moved and I didn’t see why it should be me”. Because of the crafts supervisor’s questions, however, the DSO realized that the harasser was not going to be moved, at least not any time soon — and he was not.
As a member of RORC, I was able to find out some other things. The RRD director stated that he had talked to the P&E director (“and even higher”) about this, but he had been unable to have the worker disciplined in any significant way. When the RRD director was asked why he did not ban the harasser from HFIR, as he had the authority to do, he was evasive; he said that if he banned this person, he would be faced with banning “the next one and the next one and the next one”. I believe the actual reason was to avoid union trouble. A rad protection person who did not work at HFIR but had known the harasser for years told me about a talk with the harasser, who said that he would consent to move if he could go to the Isotopes area, but there were no openings there and so he chose to stay at HFIR.
I was not able to find that the worker had been given a day off without pay, had a letter of reprimand placed in his personnel file, or had anything other than a lecture. He was not required to apologize to the DSO and, when I talked with the DSO, the harasser had not done so. He said he was told that the harasser was assigned to “improve” relations between the two divisions, but so far (days later) the harasser had made no overtures to him. Meanwhile, the other people in the craft shop made hostile comments to the DSO, when they spoke to him at all. He confessed that he felt intimidated: after all this, he would inspect the shop after hours, when the workers weren’t there, and he was be much less likely to speak to a craftsperson about a safety issue. He told me that he thought that his director and his manager had done all they could for him.
I was shocked that the harasser could get away with this kind of thing. I wrote Swanks a memo about this, giving particulars. I pointed out that the DSO had documented the instances of safety deficiencies, so it was not just a clash of personalities. I pointed out that Setaro noted that the NRC had censured Tennessee Valley Authority management for not coming down hard on harassment of quality assurance inspectors at their nuclear power plants — a direct parallel with this case. I urged Swanks, as the top representative of ES&H interests, to take up this case with higher ORNL management. I never got a response. I wrote him a followup memo stating that a lawyer friend of mine noted that since the harassment occurred outside work, the company might not be able to do anything legally. Still, said my friend, that did not let P&E management and ORNL management off the hook as far as the safety aspects were concerned: they should have backed the DSO to the hilt from the start. Why, I asked, could the ORNL director or his deputy not appear at every safety meeting held during the next quarter and, thumping the lectern, tell everybody that anyone found egregiously or repeatedly violating safety rules “is going to be sent to my office to explain to me personally what his problem is. And it had better be good!”?
Of course that did not happen. I also brought this matter up at a meeting of RORC. I thought this was an appropriate topic for us to consider. But the chairman refused to have this mentioned in our annual review report (which was in draft at that time). He said that the Department of Labor had come down hard on MMES for harassing a whistleblower and it was feared that the harassers in the DSO case might go to the Department of Labor and complain that they were being harassed as whistleblowers if they were punished. So, the chairman said, they had to save MMES any more potential legal trouble. This seemed backward to me, since the person trying to enforce safety rules was in fact the one harassed. Besides, the harassment actions were actionable from a tort point of view: the DSO’s family safety had been compromised, he had been deprived of a service (telephone) for which he had contracted, and the telephone company and newspaper seem to have been defrauded. That is, while the telephone company likely did not charge the DSO to restore service and the newspaper may have been reimbursed for the ad by the harasser, both the telephone and the newspaper people spent time working with ORNL authorities to sort this all out, time that was undoubtedly not reimbursed by MMES.
There seemed to be a lesson here about protecting people who try to enforce safety rules. I am sure that DOE found out about this incident, either at the time or eventually, but chose to do nothing — and allowed ORNL to do nothing.
The Reactor Operations Review Committee (RORC)
RORC was an independent committee of about 6-8 regular members and 4-6 ad hoc members. Its purpose was to oversee and review RRD operations. It was not a part of any division; rather, its members, all from divisions other than RRD, were appointed by the director of ORNL (later, appointment was instead by a deputy director), so it was one of the so-called “Director’s review committees”. The nuclear safety organization (called the Office of Nuclear Safety, or ONS, most of the time I was at ORNL) served as the RORC support staff, with a member of ONS being an ex officio RORC member. RORC did not directly approve RRD documents and operations but instead sent its recommendations to the head of ONS; he issued the actual approval (or refused approval, if that was the RORC’s recommendation). Usually RRD and RORC would have met, often repeatedly, to hash out RORC questions and objections, usually they would have agreed on the final form of a document, and usually RRD would drop proposals that wouldn’t fly; thus the final RORC recommendation was almost always for approval.
I was a member of RORC from 1991 until I was laid off in December 2000. The job I had before coming to ORNL included doing calculations of source strengths and dose for both ordinary operations and accidents and reviewing all kinds of safety documents for reactors. It required one to be what I call “engineering-minded”: that is, able to understand discussions, documents, and drawings covering all kinds of physical systems and components. For example, one had to be able to read a drawing that showed a radwaste tank system and be able to see if a bypass leg would allow pump leakage to reach a monitor or not. One also had to understand regulatory requirements, which for safety issues are not always easy to interpret. All of this was good experience to bring to the work of RORC. I never felt out of place on the committee in the sense of being out of my element, although sometimes I was in the minority on an issue.
Informationwise, one would think that my work as the reactor ALARA rep would have contributed more to my work as an RORC member than vice versa, but this was not so. In my capacity as the reactor ALARA rep, I had limited or no access to a lot of the information that I was able to find out readily during briefings of RORC by RRD or from documents supplied to RORC by RRD as a matter of course. Some of the committee work was confidential, in the sense that we would discuss problems and issues with RRD and they would be allowed to fix them or resolve them or at least come up with their position on them before we would make a final decision as to whether they were compliant or were operating in an unsafe configuration. Also, in our work on the committee we were responsible to the committee chairman and to the ORNL director, not to our “home” division directors. Thus I did not “copy” my ORP management on RORC memos and on only a few appropriate occasions (as I will discuss later) did I speak with them about a matter of RORC business.
When I first joined RORC, its chairman was a distinguished ORNL old-timer, Don Trauger. His approach was collegial, his conduct of meetings businesslike. More than half of the members seemed to have been on the committee for years and were fairly familiar with all of the reactors. So although our scope included not just narrow safety questions but also such safety-affecting areas as training of reactor operators and emergency response, we seemed to get through a lot in a reasonable amount of time. Relations with RRD seemed amicable; indeed, the RRD people seemed to be somewhat deferential to the committee. For example, both RORC and RRD made efforts to accommodate each other’s schedules, but RRD seemed to exert itself a bit more, as befitted its auditee status.
After a few years, Trauger retired and another longtime RORC member, Howard Kerr, took his place as chairman. Kerr had some ideas about how RORC could become more active in certain areas, notably in oversight of the “little” reactors (i.e., those other than HFIR). But after a year or so, his election to public office precluded his continuing as chairman. Kerr stepped down in late 1996 and R. M. “Mike” Harrington, another long-time member, became chairman. Although Harrington was an able person, he didn’t seem to have the “gravitas” of Trauger or Kerr, which seemed to make a difference in how RORC was handled by RRD in future years.
Overheard Statement at HFIR
In November 1992, on RORC business, I had occasion to visit an RRD engineering group leader. I had a binder or some such that I set down beside my chair in his office and forgot to pick up when I left. I remembered it before I left the HFIR site, so I went back to the group leader’s office to get it. He was meeting with two other people and his door was open. As I approached it, I heard him say, “I’ve covered up for you with RORC and everybody, but you need to…..”. I couldn’t quite make out what they needed to do, but it seemed they needed to clean up their nuclear safety act in some way. As I appeared in the doorway, the group leader saw me and a look of astonishment and dismay came over his face. I pretended not to have heard anything — I just smiled and murmured “Sorry” as I retrieved my binder.
I reported this incident to the rest of RORC at our next meeting. I noted that while I did not know the two people the group leader was addressing, I had seen them around HFIR and thought they were not of RRD but were doing work for RRD. (I think they were in the Instrumentation & Controls Division, which maintained most of the reactor instrumentation.) The RORC chairman was of the opinion that since I could not identify the two people the RRD group leader was talking to, we should not even approach the RRD group leader about this. Still, he asked me to keep an eye open and see if I could find out who the people were. For months, I looked carefully at every less-than-familiar face every time I went to HFIR and even when I was in other ORNL areas where these people might be, but I did not see them again. I reported this to RORC and the matter was then dropped, which distressed me.
The HFIR Iridium Release Event
A significant event in the history of HFIR occurred in September 1993. I will describe what happened in some detail in order to make clear various points, because “the devil is in the details”. Lay readers can skip the details, but I hope they will read the last paragraph of this section, on the lessons learned.
Iridium-192 is widely used in radiographic sources, e.g., to check welds by making pictures like X-ray images. It is also used for cancer treatment by temporary insertion of sources into the patient. At DOE’s Idaho National Engineering Laboratory site (INEL, later called INEEL), the nonradioactive iridium precursors were prepared in disk form. The disks were placed in a stack into very long thin cylinders that were then sealed by welding. Since INEL’s reactor was down temporarily, the cylinders were sent to HFIR to be neutron-irradiated. In the irradiation, the precursor material was activated to form the desired isotope, Ir-192. This had been done various times before at ORNL without incident.
After irradiation, the rods were removed to a shielding cask underwater to cool for a time. This cylindrical cask was then lifted out of the water and set upright (long axis vertical) at the side of the pool in the reactor bay, to sit there for several hours to await transport to an ORNL hot cell. When the time came for transport, the cask was lifted up with a crane and rotated 90 degrees in the air so that it could be moved over to and laid on a waiting flatbed truck, thus to be transported on its side (long axis horizontal). This was the same procedure as had been used various times before.
However, on this occasion, as the cask was rotated, the “chirpers” (alarming personal detectors), the area dose rate monitors, the air monitors, and the evacuation alarm all went off. The craftsmen stood there looking at the rad techs for direction. The lead rad tech had the craftsmen lower the cask back onto the floor on its side while the other rad tech evacuated all workers in the reactor bay. The lead tech stayed for a short time to do a quick survey of the cask and the area; he felt safe doing this since his detector gave him instantaneous readings of the dose rate at his location. He found that the 5 R/hr meter “pegged” at the bottom of the cask; the readings were 3 R/hr at 12 feet from the bottom, 1-2 R/hr at the top, 5 R/hr at 5 feet from one side, and hundreds of mR/hr at the other sides. The reactor shift supervisor came down into the reactor bay. Since the hottest spots were at and near the bottom of the cask, he judged that the best course of action — the best way to leave the cask in a safe configuration — would be to put the cask upright. Aided by the tech, he quickly used the crane to rotate the cask back and set down it upright on the floor, at which point the alarms stopped. The tech’s instrument still indicated significant hot areas near the cask, however, and the readings at the bay door and at the control room side of the window overlooking the bay from an upper story were 40 and 2.5 mR/hr respectively. A teletector (long-reach detector) later showed that the latch area at the bottom read 35 R/hr.
The supervisor’s decision to right the cask, instead of evacuating and allowing the situation to be evaluated by his management and the rad protection organization before acting, was criticized by a non-ORNL guest assessor during the RORC annual review later that year. The assessor felt that the situation was too unstable or unknown for action but was stable enough to allow for withdrawal of personnel. RORC discussed this with him in the course of settling on our recommendations. Several members took the attitude that the supervisor was in charge and as long as he made a reasonable judgment under the circumstances, he shouldn’t be faulted for it. (The term “second-guessing” was used.) I and others took the attitude that (1) the supervisor could see that the dose rates, while significant, were not life-threateningly high and (2) subsequent events showed that righting the cask did simplify the recovery operation.
Today, however, I would side with the guest assessor as regards what the supervisor should have done. Although it was clear that the source was likely the “hot” iridium, the supervisor had no idea what had actually occurred. So although the dose rates were not life-threatening, they could have risen at any moment. Also, there might have been a very significant level of airbornes, which the tech’s instrument would not have indicated. Finally, it was only after the fact that righting the cask was shown to be a simplifying choice — the supervisor could not possibly have known this for sure from what he knew at the time. I still believe, however, that RRD was right in not disciplining him for it — he kept a cool head and did a reasonable job of dealing with the situation. From other experience, I also thought that he was a capable shift supervisor and thus someone whose judgment in many different situations that might arise was likely to be sound.
It must also be kept in mind that the evacuation alarm was so loud as to be painful. The rad tech who stayed in the bay told me that it was hard to think with such an extremely loud, continuous noise close at hand. So I felt that the shift supervisor, in going out into that cacophony to take charge of and evaluate the situation, had had an extra difficulty in making sure things were left in a safe and stable state.
Over the several days after the incident, meetings attended by dozens of people were held to figure out what had happened and to decide what to do. Brief, careful entries into the bay were made by the rad techs to characterize the nature and extent of the radioactivity, using handheld detectors. One problem was that until the cask was removed, the background in the bay was high and so the detection and location of the individual particles were more difficult. A collimated detector was therefore also used, aimed at the bay floor through the reactor bay window. A thermal imaging device was used to locate one particle. (This was credited by an RRD manager as my suggestion. That is, following the observation by someone during one meeting that the particles were thermally hot due to the decays, as well as being radiologically “hot”, I went off and found out what temperature differences these devices could detect and how much hotter the particles were than, say, the concrete floor. Having concluded that the thermal imaging device could work, I suggested its use at the next meeting.) These methods saved dose by allowing remote detection to within inches prior to “nailing down” the exact locations quickly with regular detectors and picking up the particles.
Particles of Ir-192 of various sizes were found lying on the reactor bay floor; by inference, particles were also lying loose inside the cask. The floor particles were collected in various ways, such as by sticky tape mounted at the end of a pole. Eventually the cask was sealed off on the bottom by a metal “diaper” (after a lengthy engineering design and analysis process) and moved to the hot cell for removal and examination of the source rods. All of the rods were found to be intact except one — obviously this rod was the source of the released particles.
What happened? Well, up to this time the welds were not inspected radiographically at INEL. A faulty weld on one rod thus went undetected. A small hole in that weld allowed water to enter the rod during the cooling time underwater. When the cask was set out by the side of the pool, the still-significant heat of decay caused the water in the rod to flash rapidly to steam. The hole in the weld was big enough to allow the initial slow ingress of water underwater, but not big enough to allow the billowing egress of the near-instantaneous steam burst when the cask was out of water. Thus the rod was blown open near the top — the rod material actually ruptured and tore, as was seen later during the examination in the hot cell. Although the heat of the explosion was not enough to melt or vaporize the Ir-192, the force was clearly enough to fragment some of the disks. So particles of Ir-192 were blown out the hole in the rod and into the inside cavity of the cask. Any particles escaping out the gap formed by the ball-shaped entrance shield/door on the bottom were shielded by the cask while it sat by the pool. Thus the rod rupture went undetected at first. But then, as the cask was lifted and especially as it was rotated, the particles already out of and under the cask were exposed and other particles slid out the shield gap onto the floor. It was at this point that the particles had an unshielded line of sight to the various monitors, which then detected them and set off the alarms.
As to why the welds were not radiographed, that was attributed to a perception that there was no need to do so. After this event, they were. But beyond this, the question was asked as to why this type of accident had not been foreseen. The conclusion was that the whole operation had been reviewed in parts by different safety groups across the DOE complex. The irradiation in the pool had been reviewed by one set of people and transport in the cask on a truck had been reviewed by another set. However, the lifting and rotation had not really been reviewed by either because the rods were assumed to be integral from the time they were manufactured until the time the iridium was extracted in the hot cell. This type of accident had thus “fallen through the cracks” of review.
I have to stress here how uninvolved I was in the operational “recovery” planning efforts, even though I was the rad engineer assigned to work with the HFIR people on rad issues. Setaro and Gheesling and I went with Swanks to an early incident response meeting at HFIR and after that I went to all the incident meetings I could find out about. But inevitably I would discover that the HFIR operations people and rad techs had decided to make an entry into the bay late in the afternoon, after the day’s meeting was over. I had trouble getting dose rate and dose information, having to ask for it each time I wanted it and having to wait for it to come, usually orally. Dose and dose rate decisions — ALARA decisions — were made exclusively by the rad tech supervisors and the operations people. The HFIR complex leader, Roger Davis, was a nice guy and was often apologetic about the lack of information. As he had ever since he had been complex leader, he seemed sheepish regarding my exclusion; as always, I was sure he had his orders to operate this way from his rad tech group leader, Dale Perkins. I think that the version of the health physics procedure dealing with the review of radiological work that was then in force gave me the right to be involved, but this was just ignored. I did do a check of the (shielding) diaper proposed by the operations people, but this was done to satisfy what my supervisor and I thought was my responsibility, not because the HFIR operational people or rad techs asked me to. In any case, neither group was interested in the results. Had my results shown the diaper to be inadequate, I believe (from later events) that they would have disdained my work.
I was able to keep abreast of the plans to diaper and move the cask only through my RORC connection, not through my ALARA rep connection. It was very frustrating to be excluded from all these decisions, which I knew from talking to counterparts at other sites would have been made by the rad engineers or senior health physicists at their sites, or at least been made in consultation with them. Even RORC, however, had no real approval power over what was done during the recovery; we were briefed but unlike our reviews of anticipated accidents, we seemed to be included for information only, as a courtesy. I must qualify this statement by adding that had we seen any kind of discrepancy with the HFIR Safety Analysis Report we would have stepped in. But this was not considered a “nuclear” accident in the sense of having anything to do with the reactor per se and so, I thought, it was treated as not really within our scope even though it resulted in the reactor’s being down for about two months and thus constituted a major disruption of reactor operations.
Another reason for RORC’s reticence was that while the irradiation was done at HFIR by RRD, the Ir-192 production program was managed by Chem Tech and the processing of the irradiated rods was done by them in their hot cells. RORC did not wish to be seen as second-guessing Chem Tech in the safety reviews. Chem Tech was also pressing hard for prompt recovery efforts: every day that passed meant that the Ir-192 was radio-decaying and after about two months it would be too “weak” to be of practical use. Thus if the cask and its rods could not be moved to the hot cell in good time, Chem Tech’s program would suffer a big financial loss and ORNL would be looked on as an unreliable supplier.
DOE seemed not to be much in evidence after the first few meetings. I assume that they were kept apprised in side meetings or by phone or E-mail as to what was planned, but one would have thought that they would have made it a point to attend all the incident response meetings so as to keep track of the process in real time. I thought it was as important to understand why ideas were rejected as why they were adopted, especially where the expense weighed heavily in the decision. I saw no evidence that DOE was interested in the doses that the recovery workers were receiving or were anticipated to receive.
At my suggestion, group leader Perkins and I gave back-to-back talks on the radiation protection aspects of this incident at the next annual meeting of the (national) Health Physics Society. I tried to coordinate my talk with his, as Part 1 and Part 2, so we could cover all the territory without repeating one another. But he was not cooperative, so eventually, having blocked out our separate areas of How It Happened (me) and What We Did Rad-Protectively (Perkins), we did our separate things. I was surprised at Perkins’ reluctance to work with me on this. Although as I note elsewhere in this book he was perhaps the primary exponent of non-cooperation between the rad engineers and the rad tech organization, these talks were part of our professional non-ORNL activities and I thought that the subject matter showed ORNL rad protection in a good light. Also, I thought he might have been flattered that I took as the title of my talk something that he had said during the first meeting after the event: when the operational folks were getting bogged down with PhD-ish details of decay schemes and were ignoring the very real possibility of the entire bay’s “going airborne” or at least of getting the whole bay “crapped up” by particles moved around by the ventilation system, he had gotten their attention focussed back on the real issue by exclaiming in his Southern drawl, “This ain’t no science project!” I thought that that expressed the whole rad protection problem in a nutshell, i.e., that from a safety point of view, cleanup while the situation was reasonably stable trumped analysis carried out in minute detail while the situation deteriorated.
The lessons of this event and its aftermath were as follows. First, I learned — or rather, had it emphasized to me since I believed this already — that no matter how complete or how tight you think your safety analyses are, you should be prepared for unexpected adverse occurrences. Second, the rad tech organization was prepared to exclude the rad engineers even in the case of this serious incident, where the expertise of the ALARA/rad engineering group could be assumed to add something, especially, e.g., in checking the shielding calculations done by the reactor people or in suggesting engineering solutions to dose problems. Third, DOE was missing in action. Although I had noticed this before, it was glaringly evident in view of this serious loss of control of radioactive materials.
The Tower Shielding Reactor (TSR)
ORNL is out in the country on its own very large plot of land, but the Tower Shielding Reactor, or TSR, was particularly isolated, on a sub-site with especially controlled access. You entered through a guarded gate and were issued a “muster badge” as a way of keeping track of everybody on site. Then you followed a path over to a concrete pad on which stood two very tall metal towers and one short tower. The control room and office area were underground. In the old days, the reactor was actually hoisted up from the short shielding tower out into the air via cables that ran up and down the two tall towers — hence the name Tower Shielding Reactor. The reactor then irradiated the countryside, so to speak, and the dose rates at various distances through the air were measured. Obviously no people could be outside in the area during that time; to ensure this, roll was taken in the underground area prior to the raising of the reactor. I wondered what they did if a plane went over — when I flew back to Knoxville in the afternoon once and we dropped down for the approach to the airport, I saw below me the two towers. Presumably the planes were high enough that the dose rate was not significant. At some time prior to my coming to ORNL, the hoisting of the reactor out of the short tower was discontinued and the reactor was used only to produce a beam through a side port in the short tower. More or less the same safety precautions were followed, however. TSR went back up after The Great Shutdown because there was an agreement with Japanese researchers to perform some experiments, which the Japanese were mostly funding. But TSR was shut down for good in 1992.
I visited TSR on various occasions during these first few years. The older researchers were involved a lot in the management and operation of the reactor, working with the operators and the reactor manager. Thus the researchers were not just facility users as, e.g., visiting researchers would be, but were also involved in decisions regarding the configuration of the reactor, such as when to change a control plate. The reactor manager was a hands-on worker. One day Setaro and I discussed rad protection at the reactor with him. He seemed to regard safety programs as being unduly picky and intrusive. I asked why they didn’t do a certain safety thing at the reactor; I forget what it was, but it was something that was standard elsewhere. He shot back, “Because we don’t want to.”
While I admired the researchers and this capable reactor manager, it was clear that they were “from the old school”, where the operational and especially the research types believed that they were completely capable of identifying and handling safety issues themselves, without any need for safety specialists. I should say the old professional school, since there was a separate “school” for the nonprofessional types (crafts, technicians, and operators). The old professional school refrain is “We know how to work safely” (an actual quote from a real researcher, although not one at the reactor above); the old nonprofessional school refrain could be phrased as “We ain’t afraid of no stinkin’ gammas”. Either way, the idea is that safety specialists are usually in the way of O&R people and add no value. However, most of these folks would have omitted their assigned rad techs from any list of objectionable safety specialists because they knew the techs and they could see the routine day-to-day survey and documentation support that the techs provided. Also, to many professionals among the O&R people, the rad techs were controllable and did the rad scut work they perhaps would not have wanted to do themselves. But professional or less frequently encountered safety people were objectionable because they were not controllable and they made decisions that the O&R people thought should be theirs.
Although I saw the push-back attitude quite a bit during my early years at ORNL, it usually came out only in isolated ways, e.g., voiced by individuals in the course of a conversation or small meeting, not by an entire bloc of people in a large meeting or as part of an entire project strategy. It was as though it was not politically correct to voice opposition to improving the comprehensiveness or effectiveness of safety coverage by involving specialists, probably because DOE was backing it. I was sensitive to this in the early years because we in the ALARA Program were trying to “win friends and influence people” rather than throw our weight around (such weight as we had). So I followed Setaro’s lead and did not argue with these people; we tried to be conciliatory and stress that we did not wish to take up too much of their time. I still think that this was the right tack to take, but undoubtedly to some people this appeared to be a show of weakness. To such people, if you “pushed back” at safety people and they remained nice to you — and even backed off some — you were smart or even justified to push back even more. As later chapters will illustrate, RRD learned this lesson well: although the crusty old researchers and the reactor manager eventually retired, their successors pushed back in ways the oldies never would have dreamed of.
The Advanced Neutron Source (ANS)
A new reactor was planned for ORNL, the Advanced Neutron Source (ANS). It had originally been called the Advanced Neutron Reactor, or some such name. But due to the strength of the anti-nuclear power plant lobby, DOE renamed it, on the theory that an emphasis on the research aspects of the facility would make it more palatable to Congress. From early 1990 on, I was assigned to be the ALARA representative to ANS. Because there was no rad protection representative on the design team, I was the rad engineer as well as far as reviewing the design and advising as to dose criteria, etc., were concerned. Setaro assured the project management that my time would be covered by the ALARA program budget, on the grounds that this was such a highly important new facility for ORNL. I think that this was necessary in order for my involvement to be palatable to project management, as it appeared to me that they did not want to pay for a rad protection person to participate during the early design process; rather, they wanted to have their mechanical design engineers, many of whom had nuclear power plant design experience, set the dose and dose rate criteria, etc., and then have shielding specialists in another ORNL division do the shielding calculations. They expected to have the rad tech organization provide some input, but only in matters related to tech operations, such as the layout of the tech office. But neither the mechanical people nor the shielding specialists were health physicists and they were generally not knowledgeable about DOE design requirements and guidelines or even about historical doses and dose rates at research reactors. Also, as I noted earlier, the rad tech organization was not generally able to provide input or review the design except within the limitations of tech interest. So it seemed to me that the ANS people were trying to do the radiological design without any rad engineers or professional health physicist involvement.
There were concerns about the budget for this project every year until it was cancelled in 1995 and it always seemed to me that they had to skimp. The skimping and the eventual cancellation were a shame, in my opinion, because the design team was mostly a superior and dedicated bunch of people that had done an excellent job of identifying the operating requirements of the facility and then working to produce a design that would fulfill these requirements. It was an extremely complex design to do and I believe that the team was succeeding at it. Health physicists should note that this reactor, like Canadian power reactors but unlike US power reactors, would have used heavy water (i.e., with deuterium replacing some of the normal hydrogen atoms), so that on the one hand, significant neutron production efficiencies would be realized but on the other hand tritium would be formed due to irradiation of the heavy water. The team paid careful attention to this problem in particular. DOE should have been proud to back the project to the hilt, but instead allowed it to become a political football in the areas of budget and disarmament.
Regarding the budget, there was always an argument in Congress and the federal agencies about “Big Science” projects such as ANS, whose final estimated cost was about two billion dollars. The Superconducting Supercollider project was cancelled a few years before ANS and there were rumors that a powerful congressman whose state lost out on that project had worked to kill ANS out of spite. In any case, DOE did not seem to have tried to “sell” ANS as well as it should have and certainly did not make it a priority as they should have. Not only would ANS have become the premier reactor in the world for neutron research (surpassing such titans as the Lauë-Langevin reactor, the Brookhaven National Laboratory reactor in New York, and HFIR itself in the neutron flux produced), but there were some other proposed functions, such as isotopes production and materials irradiation, that HFIR performed also but that would have been more customized and modernized at ANS.
Regarding the disarmament issue, there was a treaty clause that the US had supposedly agreed to that seemed to preclude designing a new research reactor with the degree of enrichment that ANS would need to have. There was no possibility that this reactor — with researchers constantly coming from around the world to use it and with unannounced inspections by the International Atomic Energy Agency — could have been used to produce material for weapons. But somehow governmental agencies other than DOE managed to veto the use of the higher enrichment. The ANS team had to redesign the core using a lower enrichment, which resulted in a lower projected neutron flux and other less satisfactory operating parameters. Even with this redesign, however, the project was cancelled.
Soon the reactor project was transformed into an accelerator project, the Spallation Neutron Source. There are technical drawbacks to using an accelerator rather than a reactor for doing neutron research that I won’t detail here. Also, the accelerator wouldn’t be as usable (or in some cases usable at all) for the extra functions. But its chief selling points were that it wasn’t a reactor and it was supposed to cost less, which seemed to be all that mattered. I believe that the eventual cost of the SNS will turn out to be the same as or more than the cost of ANS would have been, what with the extra design effort and all.
During the last days of the ANS project, the design documents were put on two compact discs and everybody got a copy. Included were two lengthy rad protection documents: an extensive set of radiological design criteria of which I was the principal author and a shielding document on which I collaborated with one of my favorite fellow engineers, loaned from Engineering to the project. There were many other documents that I thought were of excellent quality, authored by others. I believe that we all did some sterling work on this project and it was a sad day when its end came. Every engineer has worked on multiple projects that got cancelled and were never built, but everybody felt that this one was special. For that reason, I still treasure the ANS mug that everybody got as a memento.
Various Other Projects That Were Never Built
There were various other projects that were never built for what seemed to be specious reasons, although a great deal of time and effort was spent on designing them. One of these was the Medical-Decontamination Facility, the purpose of which was to provide suitably equipped space for medical storage and meetings and also for emergency-level personnel decontamination. When I first came to ORNL, Setaro showed me the two existing personnel decontamination areas. One was a small dilapidated wooden building slated to be demolished when a replacement was built; this building had one room, with a slatted wooden table for intensive decon and a shower for more superficial decon. One drawback was, I believe, that under more recent DOE or EPA rules, the water used had to be collected as radwaste, which meant that it could not go into the sanitary sewer but had to go into a radwaste tank. As I recall, the wooden building had no such tank but rather connected directly to the sewer. The second personnel decontamination area was in the Health (medical) Division itself, a small and extremely narrow room. It seemed to me that if there were a stretcher set in the room, there would not be enough room for anybody except a very skinny doctor or nurse to stand beside the stretcher to decontaminate the victim. Also, only one contaminated person could be accommodated in this room at a time. I was told that DOE (perhaps advised by a consultant or assessment team) found these two facilities to be inadequate. So ORNL had committed to building a new decontamination facility.
The new facility was to be built onto the Health Division offices — thus it was essentially an add-on to existing facilities — and was to accommodate two or more contaminated people. The actual design-basis incident was not specified, but I always thought of it as an explosion that sent contaminated “shrapnel” into the flesh of multiple people. They would need not only surface decontamination, but also medical assistance for lacerations, embedded bit removal, and possibly burns. This facility would be only minutes away from almost anywhere at ORNL, while the world-class DOE-sponsored REAC/TS (Radiation Emergency Assistance Center/Training Site), a radiological-medical treatment and training facility in the town of Oak Ridge, was about 15 minutes away by ambulance. It was important to have a quickly reached treatment facility where injured, contaminated people could be stabilized before transport to REAC/TS.
I was on the design team for Medical-Decon Facility from 1990 until the project was abandoned in about 1992. I wrote radiological design criteria, commented on drawings and system descriptions, and did all the things a radiological engineer would do on a typical non-DOE design team. The team members did not strike me as being different from what I would have encountered back in the power plant industry and we all seemed to get along well.
But DOE at that time imposed a limit of $1.2 million dollars for capital projects. If a capital project cost less than that, the site could build it at its own option (with DOE approval of course); if it cost more than that, it had to be approved by Congress as a line item in the next DOE budget. This generally required at least two years and so delayed things considerably. The $1.2M figure had been the same for many years, I was told, and hadn’t been increased to keep pace with inflation. Thus it was an unrealistic limit, but there was no help for it.
The project team tried gamely to keep the cost below the limit, accepting compromise after compromise. But they were stymied by DOE’s rules for construction estimates, such as the specification of the percentage to be allowed for contingencies over and above the conservative base estimate. The project was subjected to a “values assessment”, in which an independent internal team looked at the structure and its equipment and pronounced judgment as to whether the design was reasonable and cost-beneficial. While the values team’s recommendations were mostly minor, they seemed somewhat out of touch with the project objectives in some major ways. For example, the values team faulted the project for providing “a lavish canopy” over the ambulance entrance (i.e., it was “lavish” to have a canopy at all). The project team pointed out that multiple injured and contaminated persons might be brought by ambulance to the facility in a pouring rain — it was ridiculous to expect the health personnel to unload the injured from the ambulance with rain and wind pelting directly onto the contaminated people, washing contamination and blood off onto the driveway, the health personnel’s shoes, etc. Near the end of the project design, the exasperated project architect brought in two architectural designers to a meeting. His purpose was to have them bolster his contention that the bottom-line figure based on DOE financial calculation requirements was from Fantasyland. These two guests, like him, had previously worked in industry and knew what it cost to build relatively simple structures such as this facility. They agreed with him that the facility could be built as designed for well under the $1.2M. Still, DOE rules could not be bent and since ORNL couldn’t or wouldn’t pursue this as a line item, the proposal to build the facility was abandoned.
Later, the medical folks did get a records storage/meeting building on that very spot. Since it didn’t have a decon facility, it did not have to be equipped with decon equipment, radwaste tank, etc., and came in well under the $1.2 million figure. However, the status of the personnel decontamination facilities — which were so questionable to DOE that the replacement facility had been pitched as the solution –continued to be in limbo. The irresponsible way that the Health Division perpetuated this situation will be related later, but for now, the reader should note that ORNL bought time by proposing and then designing the facility over a period of many months. I believe that this was done in good faith. However, ORNL then dawdled in providing replacement facilities even after the dilapidated wooden building was torn down. DOE surely knew this was the case, but failed to insist on or to facilitate the provision of the promised improvement.
Another design team I served on from 1991 on was for the Waste Characterization and Certification Facility, or WCCF. This was intended to replace a facility that determined the radiological and other characteristics of waste contained in sealed drums before the drums were sent to ultimate disposal sites. The existing facility was located in what was essentially a warehouse away from the main part of ORNL; the people who worked there had to go to a building over the ridge in order to go to the bathroom. But they had modern instruments and were upgrading them themselves in order to make them even more state-of-the-art. The contrast between the inadequate building and the advanced instrumentation was heightened by the fact that one or two waste storage bunkers nearby created a relatively high background at the facility that the staff had to overcome, e.g., by counting a subject drum a longer time. The high background thus made for an inefficiency. The staff, even as it changed over the years, seemed to maintain their optimism regarding getting a new facility and they were always eager to show off their latest new instrument or improvement.
The new facility, after a couple of years of planning, fell victim to the multi-site centralizing drive that I mentioned earlier. It was decided that there would be a three-site version of the WCCF, to be located in an old building at the K-25 site. This might have been a good idea, except for several things. As I found out as a consulting member of the design team for that facility, the K-25 people’s needs and preferences seemed to be driving the design, not the needs of all three sites. Worse, the old K-25 building needed a lot of rehabilitation — there were metal walls that were rusted out at the bottom and had holes that allowed daylight to stream in, among other ills — and the existing layout was not good for the flow paths of the several types of waste containers.
As part of my work on this facility, I identified on the design sketches various radiation streaming paths due to inadequately placed shielding and I calculated that the resulting dose rates could produce doses significant in terms of the design targets. (For you cognoscenti, the doses depended on how many workers there would be and where they were located, which hadn’t been nailed down yet.) But I was told the facility staff would have to live with the results as designed because that was the way the space and utilities were already arranged in the existing building. I was a experienced “shielder” and I was not happy to be overruled by people who knew little or nothing about shielding and dose calculations. But I tried to be a good soldier and to provide identification of the best and worst places for people to be located for extended periods. I realized, however, that I was not the only one who thought that technical considerations were being overridden: several other people on the project told me that there was political pressure to locate the facility at K-25 and to use this old building, because K-25’s longtime mission was over and its staff were scrambling to find things to do. This may be putting it too broadly, but repeatedly our technical observations about the inappropriateness of the building for the purpose were greeted with silence or evasions, a dead giveaway of a hidden agenda. On this project, after a point, the people running the show simply stopped inviting me and some of the others to meetings, I suppose on the grounds that we had contributed all that we could. Still, it meant that we were not able to comment on the final design. I do not know what ever happened with this project.
I also worked on the design team for a new ORNL laundry. The existing one had separate sets of washers and dryers for the nonrad and rad streams (the latter being for the launderable protective clothing). There was also a respirator cleaning area. The existing laundry was seriously short on space and the throughput was therefore limited. There were also detection issues with respect to the incoming and outgoing clothing. I spent a lot of time on the layout of a new laundry in 1990. The laundry would be expensive and thus have to be a line item project with, again, at least a two-year lead time. But it too was never built. Again, the problem of centralization arose, with the ORNL laundry commitment being held up while the powers that be tried to find a place at the Y-12 or K-25 site where the laundry for all three sites could be done, or at least the contaminated laundry. When that finally fell through, DOE budget priorities then did not allow for a new ORNL laundry. Some of the design team, including me, tried to figure out a layout that would use one or two add-in trailers together with the existing laundry, as a “temporary” improvement. The trailers were purchased, but P&E, the division that ran the laundry did not use them as planned: they were apparently able, to a great degree, to make their own decisions about how they wanted to use the facilities once they were set up. Thus the improvements in rad control were not realized. ORNL laundry was being hand-frisked long after most other sites or their laundry subcontractors had gone to automated frisking. (It may still be hand-frisked for all I know.)
Problems In Appling DOE Requirements, Actual and Perceived
A frequently occurring refrain at ORNL was how dumb DOE was, how they imposed arbitrary requirements across the board, with no consideration of how cases differ — in short, how DOE itself did not practice “the graded approach”. There was certainly some truth to this allegation, but I thought that the real problem was that DOE issued its edicts (usually from Washington) and then did not anticipate or closely monitor how they were applied. Thus where DOE might have thought that a particular requirement would be used in the three or four cases they thought of, in practice it might be applied by the contractors to eight or ten other cases that DOE apparently had not thought of and would not have wanted the requirement applied to. If DOE’s field reps had been closely monitoring the various projects and facilities at a site, they would have been able to anticipate these potential misapplications or to detect them as they occurred. They could then have gotten back to DOE-Washington immediately to obtain a clarifying statement regarding applicability or a deviation for the particular case and site, or at least guidance in application. But this process did not seem to occur on a timely basis, if it occurred at all. It seemed to me in these early years that if a site wanted an authoritative clarification quickly, it was best for the site people to call Washington themselves rather than wait while the local DOE people did. It seemed that some of the “rigid rules” that ORNL people complained about were actually “cover your ass” applications in cases where DOE actually did have some leeway in how individual site judgment could in fact be applied, or were applications that were extreme enough under the circumstances that a deviation could be requested from DOE, if anybody onsite want to go to the trouble. Thus the complaint of “rigid rules”, “one size fits all regulation”, that was so much complained about at times seemed a canard to me.
An early example of this was a case where the level of contamination control of an area seemed at odds with the real potential for contamination. The choice of control level is determined by the potential for contamination: the higher the potential, the higher the level of control. Other risks, such as heat stress and water stress, are also to be taken into account, but sometimes are slighted. I learned of this particular case when I was talking to a radwaste engineer whose office was in Building 3000. He explained to me that he had severe enough diabetes that he had had to go on insulin. He kept some hard candy on hand in case he started to feel shaky. But eating and drinking was not allowed in that building. So he would have to go outside to suck on the candy and he had to go across a small parking lot into another building in order to obtain water or a soft drink, which he could not then take back and drink at his desk.
In the so-called Regulated Areas (a holdover term from the old AEC days), eating, drinking, and smoking were forbidden because although there was not any actual contamination in the occupied areas, it was perceived that there was a potential for contamination. In the case of this building, in past surveys a low level of contamination had been found in the “overheads” (the upper areas — either open, as beams, or semi-closed, as attics). Rather than expend a lot of money and effort to clean up “a small amount of contamination” with “limited potential for exposure”, as some ORNL managers would put it, ORNL had decided to control the area at minimal cost by making it a Regulated Area. I would have thought this a practical idea except for two things. First, it was not a limited area of a building but the whole building that was controlled; as the engineer told me, the occupants did not have a building lunchroom, water fountain, etc., like nearly everybody else at ORNL, although strangely enough they did have bathrooms. Second, most of the people having offices in the building had nothing to do with any work going on in the building proper; they were stuck there because their divisions had nowhere else to put them. This struck me as inequitable both for the workers personally and for their divisions, especially since the trekking back and forth to another building every time they needed a drink was inefficient. My solution, had I been asked for it, would have been to reevaluate the potential for contamination. Did contamination ever waft down from the rafters? Was it in fact free to waft or was it “fixed” contamination? Etc. If the contamination might from time to time waft down, what were all these people doing spending most of their 40 hours a week in locations that were not surveyed even weekly? If the contamination was unlikely to waft down, then why was the area below “Regulated” at all? I thought that the answers to these questions, firmly established, could provide a better solution to controlling this building.
I realized that no one would listen to me and I did not have the power to change this situation. What I could do, I thought, was get a special dispensation for the diabetic radwaste engineer. I wrote a memo to my division head, Sims, asking that this engineer be allowed to keep candy and a soft drink or water bottle at his desk and to ingest the candy and liquid as needed. Eventually Sims issued a memo allowing exceptions to the rule, I believe with the proviso that the local rad techs would survey such excepted offices more frequently. In this case, I thought it was not DOE’s “rigid rules” that were to blame, but ORNL’s rigid application of them — quite a different story.
Occurrence Reports
DOE always had an occurrence reporting system, i.e., a formal system for reporting significant incidents and accidents. However, this went into high gear with the issuance of DOE Order 5000.3 (later DOE Order 232.1), in which DOE instituted a fixed-form reporting system that contractors were required to use. Reports submitted under this system were reviewed by DOE people, not only locally but in Washington, and were bounced back to the contractor to be rewritten if they were deemed inadequate or not credible. After several years, the HFIR folks told me that the Washington people were insisting that since these reports were public documents, they should be comprehensible to an intelligent member of the public. Hence such facility-specific terms as “subpile room” and “control rod drive”, which were understood and accepted by the DOE technical reviewers of the reactor reports, were rejected by the DOE people who reviewed the reports for form. The HFIR people had even had reports bounced back for grammatical errors. I agreed with them that using common technical terms without explanation should have been allowed, but — knowing some HFIR engineers’ writing as I did — I had sympathy for the DOE people who were trying to parse some of their sentences.
I thought the DOE technical people sometimes accepted reports that, to a knowledgeable person, had dubious assignments of causality. At ORNL, writing occurrence reports was the responsibility of the division controlling the facility at which the occurrence happened, not of some independent group specializing in investigating these occurrences and writing the reports. Thus the division could control what information went into the reports and how it was “spun”; most importantly, they could control the selection of the “root cause” and the “contributing causes” of the occurrence and thus they could limit any resulting corrective actions. The writing was done by a line supervisor or other line representative, often without safety people’s review of the draft. This was very frustrating to safety people, who had no power to alter what was put into the reports. Although there was a review by, I believe, the shift supervisor’s office and/or the Office of Nuclear Safety, these people would usually take line management’s word for the cause if it was an operational safety issue (and not a higher-level accident analysis issue). Also, if the complex leader or rad engineer disagreed with what was written, he often found out about it after the report had been sent to DOE, when there was little chance that it would be revised and resubmitted.
Control of the reports by the line organization was a significant factor, I believe, in the minimizing or even ignoring of events in which radioactive contaminations occurred or radiological procedures were violated. The division would not want to have to admit they were at fault and in various cases they blamed other organizations for the incident. Two of RRD’s favorite scapegoats were the laundry and the ES&H training organization. For example, I have a copy of a 1990 memo from an ES&H training person to a HFIR manager in which the training person was responding to a memo that the HFIR manager wrote him criticizing rad worker training. Included in the manager’s memo was a copy of an occurrence report in which inadequate rad worker training was blamed for a contamination of most of the reactor bay floor. But as the training person noted in his memo, the worker involved took rad worker training over a year earlier and rad worker training had already been improved substantially since then, such as increasing the time the course took from 8 to 12 hours. Even so, the training person had agreed to use the occurrence as a “lesson learned” case study in the class.
However, he observed, the occurrence report was confusing. The cause was listed as “procedure” but no procedure existed or was planned to be written for use in the training class since the activity during which the contamination occurred (visual inspection of contaminated items) was too infrequent to take time with in a general rad worker class. He finessed this point by suggesting the writing of such a procedure by the reactor people for use in task-specific training. He also noted that a non-ORNL consultant had reviewed HFIR reactor programs, the HFIR log book, and contamination surveys related to the event for radiological implications. This consultant noted that the contamination was not just in a small area, as would occur by trackout on shoes, but over a widespread area. The consultant and the rad tech supervisor thus concluded that the cause was more likely the HFIR pool ventilation pattern. (Note that this was recognized by all rad protection folks as the probable cause even though the RRD folks were apparently in denial for a long time. Eventually, the HFIR pool ventilation pattern was revised to correct the potential for contamination.) So the ES&H training person concluded his memo by saying that ES&H did not agree that inadequate training was the cause of this contamination. However, I believe that no change was made in the occurrence report.
In 1992, the ES&H training head was sent another memo about a HFIR occurrence. This time the sender was a P&E manager. (The P&E craft workers were sent by their division, a support organization, to work in other divisions’ facilities.) The manager told the training head that some of the craft personnel had expressed concerns that rad worker training included “more theoretical or technical information than is needed and does not adequately cover the more basic or practical aspects, such as barricades; signs; protective clothing; respirator use; personal dosimeters; and the requirement for not smoking, eating or drinking in radiological areas”. He noted that “these concerns were expressed again relative to a recent radiological event at HFIR”. This of course was nonsense. I took rad worker training myself every year I was at ORNL and it always included all of the items cited by the manager except for respirator use, which was covered in separate respirator training. The point of the memo seemed to be to support the occurrence report’s blaming of inadequate rad worker training for such obvious and deliberate violations as “smoking, eating, or drinking in radiological areas” or for personnel errors such as contaminating skin or personal clothing in the course of removing protective clothing. The worker involved obviously knew he was breaking the rules, but P&E in this instance was trying to get itself off the hook by blaming the rad training program.
Several times I wrote memos to RRD people and others that pointed out logical or factual errors in HFIR occurrence reports or that asked questions about points in the reports that were not clear. I recall receiving a reply only once. Several other times I tried to phone the author of the occurrence report. More than once, an author did not return my call even when I left multiple messages. This sort of rude behavior was the more blatant because usually the authors knew or could readily find out that I was in rad protection and on the RORC.
I believe, from what people at other sites have told me, that the situation was quite different at their sites: there was a dedicated group that wrote the occurrence reports or at least reviewed them for accuracy and validity, or there was a method to assign an appropriate independent person to write or review a given report. In either case, at other sites line management did not seem to control the writing of reports in which there was a safety issue, as was true at ORNL.
That said, I must point out that despite the blame-avoidance that some divisions occasionally indulged in, in the early years ORNL was demonstrably reporting contaminations that other sites were not, because the Oak Ridge DOE people were requiring it even though other sites’ DOE offices did not seem to. ORNL reports at these levels were in fact in compliance with the requirements of the DOE Order and in time other sites began to have to report more. DOE allowed multiple similar low-level events to be reported in a single “roll-up” report, which saved time; even so, the expenditure of time and effort was substantial for all sites. The HFIR people estimated in the mid-90’s that an uncomplicated occurrence report cost them $2000 (in terms of man-hours of preparation, review, and documentation) and more complicated ones could cost ten or more times that. There was thus a strong motive to avoid having to make a report even if one did not consider the cost of corrective actions.
Efforts To Obtain Outside Technical Information
In 1992 a representative from a vendor company asked to visit ORNL. This company made equipment to “scabble” surfaces — that is, to remove paint and even the first inch or two of concrete by a sort of rapid-impact chipper. The improvement that this and other companies had made was that they coupled the scabbling head with a vacuum system. Thus when contaminated surfaces were being scabbled, the particulates produced did not just blow up into the air but were captured with a high efficiency by a filter connected to the scabbling head with a tube. With this equipment, a lower level of or even no respiratory protection had to be used (compared to other methods) in doing decontamination of walls, floors, and other surfaces. This kind of scabbling equipment was being used at K-25 and we heard that Y-12 was interested in it, but it was not being used so far at ORNL. Because of the hands-on demonstration the vendor rep would be putting on, I invited some operational people and the rad tech complex and group leaders as well. I was happy to see how many people there were from operations groups, but sorry to see that the rad tech organization was not well represented: I think only two came and one of the two, group leader Perkins, made only a brief appearance. It seemed that the rad tech organization, which prided itself on being “all things rad”, did not appear interested in a cost-effective device — an “engineered control”, as the regulatory documents would call it — that provided better radiological control of airbornes. This was just one example of the many times that we rad engineers would try to bring in information about new ways of doing things, only to find them rejected by the rad tech organization (and sometimes the operational people as well) without examination of their merits. There was also the famous Not Invented Here syndrome that, as we saw earlier, caused old-design detectors to be used at ORNL long past the time when other places had gone to pancakes and other modern designs; it was a notable feature of rad work that “the old ways are best” philosophy usually applied where rad control was concerned and innovation was not welcome unless it came from the operations people or the rad techs.
