How did the UPF design disaster happen?

Uranium Processing Facility

Uranium Processing Facility (B&W Y-12)

The screw-up that forced the redesign of the Uranium Processing Facility cost at least half a billion dollars and likely delayed completion of the new production center at the Y-12 nuclear weapons plant for a matter of years.

It didn’t happen overnight. In fact, it appears that key members of the UPF design team — indeed, whole groups — were in denial about the space/fit issues, even up to the time it became something that could no longer be ignored (when the team was on the verge of announcing 90 percent design completion).

The root-cause analysis, which was conducted by a team of experts from Bechtel and Babcock & Wilcox, found that the problem could be traced back to the decision that froze the size of the building in March 2009 – years before the process and commodity design for the UPF was significantly mature – and a desire to get the project done as quickly as possible.

The report, titled, “Y-12 Uranium Processing Facility Space/Fit Issue Root Cause Analysis,” was completed in December 2012. I recently obtained a copy of the “official use only” report, which provides a detailed analysis of the evolution of the project — all the way back to the 1990s.

“The decision to freeze the building footprint and volume was driven by schedule demands to begin construction of the building in 2012 in order to support operation (Critical Decision-4) in 2018,” the report states. “This required that detailed structural design begin in 2009. Since the design of process equipment was immature, there was a risk that it would not fit in the building. This risk was not fully understood, and there was no process to periodically reassess the risk when it was not resolved quickly.”

The analysts also found that management of the space/fit concerns was “less than adequate” by the project management team, which created a big delay in understanding the true significance of the problem.

The team was “highly motivated to manage the process design within the building volume, and it underestimated both the probability and consequence of fit problems,” the report said.

If the team had managed the risks more effectively and if other indicators of fit problems had been recognized, the whole thing might have been resolved sooner, the review team concluded.

“On UPF, pressure to meet schedule milestones caused the PMT (project management team) to move forward with building design in advance of process design,” the report said. “Although concurrent design of process equipment and the associated building introduced risks that were not recognized as significant until late in the project, it was assumed that process design and associated process equipment could ultimately be made to fit within the building.”

As background, the report notes that the need for a new manufacturing facility was first identified in the late 1990s. “The expectation was that such a facility could be constructed at Y-12 . . . for less than $1.0 billion,” the report said. Even at that time, however, a feasibility study had suggested that the 488,000-square-foot building (size projected at the time) would cost about $2 billion.

After the contractor change in 2001, when the B&W/Bechtel team replaced Lockheed Martin, a preconceptual  package suggested that more than one facility would be needed.

That eventually led to the twin facilities, the Highly Enriched Uranium Materials Facility and the UPF.

The initial configuration for UPF was a five-story, 712,000-square-foot facility that would ma. “The incomplete cost estimate was $2.6 billion,” the report said.

“When the M&O General Manager was presented with a progress report on facility development that indicated the need for additional space in the facility, he relieved the Project Director of her duties,” the report said.

A new team was formed to develop a “more economical solution,” with two design alternatives. The first design was a three-story, 583,000-square-foot facility, with a cost estimate of $1.8 billion. Another “capability-based design” was 309,000 square feet, with a “bottom-up estimate” of $1 billion.

The report said one of the main reason for the big variation in footprints was that the space requirements for each of the process had not been identified, and architects were responsible for determining the building size.

After a series of external and internal review, an “intermediate design” was chosen for a 388,000-square-foot with a cost of $1.1 billion and that design option was included in the Critical Decision-1 package submitted to the Department of Energy in 2006.

Review of this package got delayed for a year because of construction problems with the Highly Enriched Uranium Materials Facility at Y-12. It later was turned back to B&W Y-12 to come up with a revised package, which was submitted in May 2007 and approved in August of that year. “DOE added contingency in the amount of $1.1 billion because it considered this a high-risk project, making the range $1.4 billion to $3.5 billion,” the report said.

“During the next year, the project team continued to develop the design but struggled to fit all process equipment in the building,” the report said.

In 2008, contract awards were made for the glovebox design, structural design and process design via so-called basic ordering agreements.

“The project continued to experience issues with congestion as the design progress,” the report said. “From May 2008 through December 2008, 14 separate risks related to space issues were identified.”

Following a tour of a similar facility in the United Kingdom, the team became interesting in a design concept that used a utility floor over the process floor.

The first risk-management attempt was opened in August 2008. “The risk mitigation strategy was to add mezzanines as necessary to create additional space for equipment,” the report states.

After the building shell’s dimensions were frozen in March 2009 (at 388,000 square feet, with a building height 12 feet lower than a previous version), the problem of fitting the equipment into the space became more of an issue.

“In March 2009, the (3D) model had only been partially populated with equipment and reserved space for equipment, utilities, and access corridors,” the root-cause analysis said. “Consequently, model reviews gave a false impression that there was sufficient space for the utilities, steel, and other components – in fact it gave the impression that the building was voluminous.”

In July 2009, the Plant Design group began to have space issues and commissioned a series of “fit studies,” which focused on specific zones in the process work areas, the report said.

“Throughout 2009 and 2010, some individuals in the Engineering organization felt the building was simply too small,” the report said. “The modelers and layout designers were struggling to fit the internal building steel, equipment and utilities into their allotted space. Model reviews were not always attended by the necessary organizations . . .’’

So, why didn’t things change? According to the report, there was a widespread perception within the UPF design effort that the building size could not be changed. Therefore, most of the energy and attention was given to finding alternatives.

“In August 2009, when the equipment layout was approximately 50 percent complete, the Facility PE (project engineer) made the observation in a project review that the margin for space allocation had been depleted,” the report said. “This recognition was disconcerting because much of the equipment, glovebox, and skid designs were in their infancy. However, the lack of remaining margin was taken as a challenge by the Facility design group and possibly because of previous success and/or overconfidence, not escalated further with the Project Engineering Manager or Project Manager.”

The report said the UPF team adopted a strategy of adding mezzanines as a way of solving floor space issues, although Engineering did not realize that while these mezzanines added floor space they consumed volume in the overhead utility routing zone of the process areas.

“The full impact of the additional mezzanines was not recognized until late in 2011 when the overhead space in several process areas was seriously congested,” the report said.

In May 2011, the Facility Design group discovered that the process heat load had been underestimated by 50 percent. Fixing that issue consumed “significant resources” and also postponed utility layout work for months.

“Once the design team completed the process heat load increase effort and work resumed populating the 3D model with glovebox and skid designs, the severity of the fit issue became clear,” the report said.

The Engineering group concluded in March 2012 that the building needed to change, according to the report.

“In March 2012, as the new Engineering Manager attempted to confirm 90 percent design maturity, it was confirmed that the required equipment would not fit in the building in a compliant manner,” the report states. “After study of all practical alternatives, it was concluded by the Project Management Team in July 2012 that the building had to be enlarged. It was decided to raise the building height increasing its volume by approximately 25 percent.”

This entry was posted in nuclear, UPF, Y-12 on by .

About Frank Munger

Senior Writer Frank Munger covers the Dept. of Energy's Oak Ridge facilities and many related topics — nuclear weapons, nuclear waste and other things nuclear, environmental cleanup and science of all sorts. Atomic City Underground is, first and foremost, a news blog, but there's room for analysis, opinion and random thoughts that have no place else to go.