Here are just a few of the things we learned today at the Defense Nuclear Facilities Safety Board hearing in Knoxville, Tennessee.
• With the UPF approaching 90% design completion, and more than half a billion dollars spent, a decision has been taken to go back to the drawing board to re-design the facility because the designers noticed they can not fit all the equipment planned for the UPF in the space allotted and, in addition, the facility’s seismic and fire safety precautions were inadequate.
• NNSA is using the UPF, a thermonuclear bomb plant, as a kind of $7.5 billion dollar classroom; they are “learning lessons” as they move through the process. Among the lessons they have learned so far:
1. DOE Order 1189 that requires preparation of a Preliminary Safety Design Report early in the design process in order to save money and integrate safety into the design is a good idea. Ignoring it was a really bad idea.
2. The Safety Board’s concerns about safety integration, first noted in 2009 when NNSA changed its safety plan, was spot on. Ignoring them was another really bad idea.
3. There is a reason DOE structures planning through a series of sequential decision milestones called Critical Decision 0, Critical Decision 1, Critical Decision 2, etc. It makes sense to move a project through those discrete stages.
“If we could reset the clock,” said John Eschenberg, who is the UPF Project Manager now, but was not back then, “We would not have suspended work on the PSDR. The decision to merge Critical Decision 2 and 3 was not the best decision.”
• Even though industry best practices follow a careful Technology Readiness Level scale to assess the maturity of new technologies to prove, one step at a time, their likelihood of being successful, NNSA feels that a “high degree of confidence” is a reasonable substitute for actually proving a process in a production setting, so a technology that has reached TRL 5 can be included in the design as if it had reached TRL 6. (Never mind that their most mature new technology, microwave casting, failed its TRL 7 test when it was inserted in the production line at Building 9212 last February and had to be sent back to the manufacturer for core component replacement. No lesson to be learned there, I guess).
• In the process of dealing with pesky space/fit issues, NNSA has decided not to include dismantlement and disassembly operations in the UPF; any transition from old buildings to the UPF is at least 10-20 years down the road and possibly more. (One Safety Board member noted that if they wait 20 years to move out of Buildings 9204-E and 9215, where dismantlement and machining take place, those buildings will be older than Building 9212 is now. No lesson to learn there, either.)
• Though Chairman Winokur asked a really hard question, he gave them two chances to answer it. Once he asked NNSA’s Don Cook and once he asked B&W’s team of managers, “Why did this happen?” referring to the decision to skip the PSDR at Critical Decision 2. He totally stumped them; they had no answer.
Chairman Winokur: Why? When there was this well-defined set of requirements, why move to CD/2 and CD/3 without developing a PSDR? Why did this happen? Don? Do you have any sense of that?
Don Cook: I’ll try to come directly to your point. Across the nuclear security enterprise, we had experienced some funding problems. We were dealing with a 20% cut. There was a lack of funding, and priority issues, and staffing, and they all came together.
and later, to the B&W Y12 panel
Chairman Winokur: You were headed to a final design this month, but now we are hearing of a re-design. It seems to me there is some disconnect there. But what was going on?
Chairman Winokur: We heard six months ago that you were going to be at CD/3 by now. I see one of you shaking your head. Can you help me understand?
B&W guy: From overall, our goal was to be at 90% design by the end of the month. We were headed there into the spring. But at 70% design completion we ran into the space/fit issue. Now we had to step backwards. We won’t know how far back for about three weeks. We will report back to you then.
[It should be noted that in July, NNSA said the design was at 75% completion, which calls the “last spring” timeline into question. OREPA has been saying the rush to build was putting the cart before the horse and could lead to expensive re-design problems for six months because we believed the General Accounting Office was right. No lesson to learn there, I suppose.]
• After five years of design work, a Preliminary Safety Design Report, Revision 1 (the hastily prepared Version 0 was rejected by NNSA and the Safety Board), was finally delivered last week. Unfortunately, it is for the old design UPF and not the re-design UPF. (Lesson, anybody?)
• No matter how many times they were asked (and they were asked several times) no official from NNSA or B&W Y12 would tell the Board how far along the design was now.
“Where are you right now?” asked the Chairman, “Ninety percent?”
“I’ll take that question for the record,” responded Eschenberg. “We’ll be better informed in 20 days.”
Everyone on the team had been well-briefed.
B&W Y12 team member: “I do believe we know where we’re headed.”
Safety Board: “When will we know? “
B&W Y12: “In three weeks, we’ll know better.”
Round one of the hearing was a face off between the Safety Board and the National Nuclear Security Administration about the proposed Uranium Processing Facility, the $7.5 billion+ bomb plant being designed, make that re-designed, for the Y12 Nuclear Weapons Complex in Oak Ridge, Tennessee, to replace aging Building 9212 where the NNSA produces thermonuclear secondaries for US nuclear weapons under the Life Extension Program.
The NNSA was represented by Dr. Don Cook, NNSA Deputy Administrator for Defense Programs, Robert Raines, Steven Erhart, and, locally, John Eschenberg, UPF Federal Project Director and his deputy, Teresa Robbins.
The entire DNFSB was out in force”Chairman Peter Winokur, Vice-Chair Jessie Hill Roberson, John Mansfield, Joseph Bader, and, at his first DNFSB public meeting, new Board member Sean Sullivan.
Chairman Winokur opened by rehearsing the formalities. Then he reviewed the need for the UPF and the Safety Board’s history of the planning, arriving eventually at the core concern: the NNSA has not yet adequately integrated safety into the design of the UPF. He noted that even as concerns were raised and NNSA claimed to share the concerns, it allowed B&W Y12 to continue with the design.
Then he got specific. He said several recent management decisions added to the Board’s concerns:
1. the reorganization of the federal office overseeing the project (this refers to the decision to merge management of Y12 in Oak Ridge with the Pantex plant in Amarillo into one contract);
2. a major change in the project’s scope and execution strategy (this refers to NNSA’s decision to drop dismantlement, disassembly, surveillance and machining from the scope of the UPF, kicking that can down the road ten to twenty years);
3. the redesign of the UPF structure and some process systems in response to equipment spacing and fit issues (this refers to the discovery, as the design was nearing the 90% completion stage, that not all the equipment would fit); and
4. the development and use of unproven technology in the facility.
Then Steven Stokes, Lead Board member for Nuclear Facilities Design and Infrastructure Group laid out the history of Board interactions with the design of the UPF. He noted the Board was encouraged in 2007 that the UPF was integrating safety into the design; they wrote a letter that raised six issues and asked NNSA to resolve them in 60-90 days. One of them remains unresolved today”the issue of adequate federal staffing. As the project progressed from 2007, design changes were introduced, and management decisions were taken that compromised safety integration, leading to now.
In April 2012, the Board wrote a letter to NNSA concluding that safety had not been adequately integrated into the design of the UPF. They noted the following deficiencies: the need for a post-seismic confinement strategy; the need to thoroughly evaluate unmitigated hazard and accident scenarios, the need to identify safety-related controls to protect the public from fires; and the need to calculate reasonably conservative radiological dose consequences for accident scenarios that may require consideration of safety-class controls.
Mr. Stokes said recent decisions by the NNSA, including the decision to defer portions of the original UPF project, introduce safety related risks that will challenge the project’s ability to integrate safety into the design. He noted other DOE project teams have been unsuccessful in attempting similar complex execution strategies, citing the Waste Treatment Plant at Hanford which began construction before the design was completed, only to discover safety-related issues late in the design process.
The Safety Board is also concerned about the new management structure (folding Y12 and Pantex management into one contract) and the possibility of a new team winning the new contract and taking over the UPF project mid-stream. (This is not an unlikely prospect given the NNSA’s need to demonstrate its displeasure over the Transform Now Plowshares security breach at Y12 in July.) To manage all the transitions, NNSA has a scheme where responsibilities are handed off as the project moves through phases, leading to the possibility of confusion about who is accountable.
The Safety Board, seated behind tables on their raised dais, peppered the NNSA officials, lined at tables on a slightly lower platform, with questions about who was in charge, and how responsibilities were to be handed off, and who would resolve issues. Some of the questions were tricky.
Q. Who is primarily responsible for integrating safety and design?
Eschenberg. Mr. Raines
Q. Did I misunderstand? Because I didn’t hear Mr. Raines saying that.
Raines. It’s a shared responsibility
Cook. The group here is a team. If the issue comes back to safety, we meet as a team to determine a solution.
In reponse to questions about staffing, NNSA said it had insufficient staff in the past but was ramping up, with a focus on gaining expertise in ten critical skill sets over the next 3 months, bringing the total to 22 (15 NNSA Full Time Equivalents and 7 contractors). The critical skills yet to be assembled include criticality safety, chemical safety, fire protection, structural engineering.
One interesting note that peeped its head out in the course of an answer. The first-scope project of site preparation will be done by the Army Corps of Engineers. If you have a long enough memory, you may remember the Army Corps of Engineers was asked last year by OREPA to review the NNSA’s “Wetlands Assessment,” the part of the UPF construction project, announced after the completion of hearings on the Environmental Impact Statement, that said nine separate wetlands would be disturbed or destroyed to construct a haul road to reach a concrete batch plant. When we first asked the Army Corps about the possibility of a hearing on the wetlands plan, the woman in Nashville was enthusiastic in her response and said she would get back to us. But she never did. Hmmmmm”¦
As the Safety Board discussed staffing issues and the capacity for NNSA to provide adequate oversight of the contractor’s plans, one DNFSB member asked, “Do you have to have a PSDR before you reach the final design activities for safety controls?” There was a long pause before the answer came: “Yes, sir.” The follow-up: “Yes. You need all the hazard analysis completed. Is it difficult to get the contractor to deliver the safety basis on time?” Answer: “There has been a problem with deliverables, but the contractor is improving. We believe we now have a credible schedule for deliverables.”
At one point, a Safety Board member asked “Doesn’t it trouble you that these refined hazard analysis”this discussion is coming this late in the process?” Answer: “There were avenues and opportunities to do things different. What gives me comfort is that we are in the throes of detailed design, an iterative approach, with a robust back and forth to come to a conclusion on the best approach. You ask if we’re late? I say we’re just in time.”
Later a question that drew an extended silence before answer:
Q. Does dealing with safety issues this late in the process pose risks?
A. Yes, but more of a cost risk than a safety risk.
Sean Sullivan, newest Safety Board member asked a really hard question. It didn’t sound like a stumper, but he drew blanks from the NNSA panel. He asked John Eschenberg if DOE guidance on technology readiness was being applied to the entire scope of the UPF (including the deferred parts) or just the part for Building 9212. Eschenberg said they wanted to be well informed and they have a lot of information at TRL 5 and blah, blah. Sullivan tried again:
Sullivan: But my question was is the Guidance being applied to all the facility or just the 9212 parts.
Eschenberg: We hold true to the principles of the guidance. In this particular instance, it is a value judgment”¦
Sullivan (believing persistence would pay off): Dr. Cook. Is the Guidance being applied to 9212 or to the entire scope?
Cook: Your concerns, we have. Our priority is to build the entire building and move 9212 in. We’ve requested to Congress a great deal of money, part of it is for risk reduction. The only current technology that is not, in our view, at TRL 6 is Advanced Integrated Machining (AIM) capability. As for the priority of the move, it is 1. 9212; 2.”¦
Mr. Sullivan, apparently taking it as a lesson learned, gave up. I give him 3 points for trying so hard, and an additional point for learning reasonably quickly.
Round Two was DNFSB and the B&W Y12 management team, James Haynes, Mark Seely, John Gertsen, Kevin Kimball, and Brant Morowski. We could only see their backs and I couldn’t even tell who was talking all the time.
B&W mounted an early defense of their ineptitude, explaining to the Board that their experience (180 years between them!) had left them ill-prepared for doing safety analysis on this kind of project. Their experience was in assessing hazard at existing facilities, and no major production facility like this has been built in forty years. So they had to change their mindset and bring in new people. They did not mention that they spent half a billion tax dollars in the meantime.
One concern of the Safety Board was the new PSDR, Revision One, just delivered last week. After nailing down the history of the PSDR, one Board member asked if it was a PSDR for the original design, the current design, or the new re-design. The answer came back: the current design, but it’s not so different, safety-wise, from the re-design.
Q. So Revision 1 of the PSDR incorporates safety even with the redesign?
A#1. That is correct.
A#2. No, that is not correct.
Q. Let me get clear. You submitted the PSDR for the old design.
A. The processes are the same. What’s not included is the change. We will take the building fit solutions and modify the safety design strategy.
Q. So when will we get PSDR, Rev. 2?
A. We plan a “limited scope PSDR” to support construction.
Q. Is it worth our time to look at PSDR Rev. 1, because I don’t want to stay up tonight reading this”¦
A. Yes. The processes and the functional classifications are not changed.
The same train of thought chugged along on the “Hazard Evaluation” track. The Safety Board wanted to know which design they applied to.
Q. If I look now, am I looking at the Hazard Evaluation Studies for the two year-old design or the new re-design?
A. The part about fire suppression is all new.
One Safety Board member noted fire suppression concerns have been a long-standing issue with the Safety Board and gently chided B&W’s team. “We hear your measured responses and your careful reasoning. We’ve heard it before. We want implementation.”
The Board asked about B&W’s assessment of the reason for the re-design and their level of confidence in the new design. The answer was they had included a contingency in the original plan, using mezzanines. But they exhausted even that space and still had a problem fitting everything in. Hence, the re-design.
By now the hearing had gone on three and a half hours without a break. The B&W team was dismissed, and the hearing moved directly into public comments.
Lesson learned: To serve on the Safety Board, you have to have a large bladder.
There was about an hour of public comment. A parade of local government officials pimping for the UPF with nary a word about safety specifics, although it sounds as though they are all for safety as long as it doesn’t impede the flow of money. And half a dozen OREPA members spoke, all in favor of the Safety Board taking a No Compromise approach to safety issues and challenging NNSA’s capacity to manage the project.