Memoirs of a start-up

All startups resemble one another, but each is unique in its own way. So it has been with the startup of the DUF6 Project and its plants in Paducah, Kentucky and Piketon, Ohio. The two plants are the first of their kind in the US, designed to process the DUF6 tails generated by decades of uranium enrichment at the plant sites and at Oak Ridge, Tennessee (this material has been moved to the Piketon facility).

The plants were constructed and operational testing was initiated by the previous contractor, UDS, under contract to the Department of Energy. Following the procurement for the start-up and operational phases, Babcock & Wilcox Conversion Services, LLC (BWCS), assumed management at the end of March, 2011, after a three-month transition period.

We have more than 700,000 tons DUF6 in approximately 58,000 steel cylinders to process. We expect the processing to take up to 25 years at Paducah, where there are 37,000 cylinders, and 18 years for Piketon’s 21,000 cylinders.

Given the immensity of the work and its schedule, one might reasonably ask, “Why do it at all?” The answer is relatively simple: risk reduction. The material is unstable, and the steel cylinders are subject to corrosion. They can contain the material for decades, as they have, but not permanently. Congress, responding to environmental concerns from the states, mandated that the DOE convert the material to a more stable form for its beneficial reuse or ultimate disposal.

On an annual basis, we inspect 25,000 of the cylinders. Some few have minor repairs and require a monthly inspection. When we have finished our job, the cylinders will all have been refilled with the more stable uranium oxide, and will be suitable for transportation. By the end of the project, they will likely have found a new home at one of three potential permanent disposal sites.

No start-up ever runs the way you hope and want it to run. This is especially true of new types of facilities, and the DUF6 Project was no exception. But it is only in hindsight that you learn what you didn’t or couldn’t know (which is not to say that we are through learning).

The first issue we faced was that our plants and technologies had been scaled up from bench for the first time. Production under the license had previously been limited to single runs, whereas we were operating seven full production lines (four at Paducah and three at Piketon) with associated systems.

Conception to production usually involves several iterative steps, but there was an urgency to get these plants up and running to meet local environmental compliance orders. Things that might have been learned in sequential steps were not illuminated in a single giant leap forward to production.

We, as well as our customer, expected technical and managerial challenges that are likely to exist with any start-up. These were amplified by the pressure from the US Congress to move to full-scale production rapidly. It is always wise to expect the unexpected with the start-up of a first-of-kind facility . . . and perhaps that is the most important thing we’ve learned.

When we first looked at the plants, we thought most things were as they should be. No one at that point had any reason to believe otherwise. Appearances can be—and were—deceptive. In fact, this was one of our major lessons.

We had a contractual production target that we thought was within reach, only to find that it was a much bigger stretch than we anticipated. But . . . we have overcome many of the obstacles, and plans have been made for changes that will lead to—and have already led to—higher levels of production.

Perhaps we shouldn’t have been surprised. Similar start-up and operational complexities are part of the history of the nuclear enterprise, including nuclear materials processing. No nuclear operation has gone from start-up to maximum or optimal capacity and availability in less than several years—more often, decades—of refinement.

We are currently in production and have been since August 5, with all seven lines running exclusive of maintenance and repairs. Short scheduled outages allow us to perform corrective maintenance that cannot be done with the lines running.

In FY 2012 (which ended Sept. 30), we safely processed over 6000 metric tons of DUF6 and shipped more than 1.1 million gallons of aqueous hydrofluoric acid offsite. Much of this quantity was achieved in the last months of the year as both technical and managerial challenges identified earlier were resolved.

Some of the challenges we have addressed:

• The system as a whole needed to be more integrated on every level. While all of the pieces worked as tested, the interactions produced by continuous production and facility operations stressed every element in new and arduous ways. Many elements failed to perform as anticipated under those conditions.
• Safety is our first priority. We believe everyone should go home at the end of their workday in the same condition as they arrived. We made a continuous effort to put our money where our mouths were on safety, even to the point of calling safety pauses and investigations under deadline conditions. Building on the safety record of the previous contractor, we finished four years without a lost time accident.
• At start-up, there were major parts shortages at both plants, particularly of vulnerable parts. Bridging this gap became and remains a high priority as we learn more about which items are fragile under operating conditions. We learned that having an efficient supply chain is critical to success. We continue to work on it.
• Although many parts were fine in situ, they later showed real problems and vulnerabilities under long-term operating conditions. These could only be identified when the units and plants had significant run-time.
• We discovered materials issues with two major and critical parts designed for long-term use. They were significantly degraded within 30 days because of the corrosive nature of the oxide, or because of increasing, unrelievable differential pressure. Parts constructed of new material (in some cases meeting the original design specification) are replacing these items. Current indications are that the new material delivers the desired performance.
• Some important subsystems interacted with the plant systems under operation in such a way that their failure caused shutdowns. Reliability was one part of this; reengineering was needed. Design changes were sometimes necessary as well to keep things operational. The longer we run, the more we discover, and the less these failures interfere with production.
• Support systems—including hydrogen generators and hydrogen supply, plant cooling water systems, and steam generation—proved to be inadequate for multi-line operations and many of the these support elements had to be changed out for higher capacity. We are also dependent on utility systems of our larger federal sites, which supply our utilities. Some systems are more than 50 years old. In Piketon, a one-minute site electrical shortage caused a one-day shutdown of our plant. We have to engineer around that.
• The transition from construction to operations demonstrated that many of our business systems—payroll and accounting, inventory and warehousing, procurement, project controls, information technology—couldn’t support constant operating conditions. We have updated them all.
• The work culture you operate in is important, and we had several lessons to learn about DUF6 culture. First, we had three different competitive cultures, at three different locations. One was Paducah, one was Piketon, and one was Lexington (Kentucky), our home office. We introduced video-conferencing at all locations to increase collaboration. It is an active system, with morning meetings joining all three management groups, and a variety of functional conferences throughout the day. The sites may still be competitive, but now they are also collaborative. We have solved many technical problems, small and large, with this kind of collaboration.
• We hired most of the employees of the previous contractor, who were accustomed to a construction culture, where events are planned sequentially, generally happen once, and are concluded. Operating plants call for multiple points of planning and implementation; nothing (like housework!) is ever finished, and surprises are par for the course.
• Today, our operating and maintenance staffs are fixing things on the fly, keeping the plants operating while components are being repaired. These are significant cultural changes.
• The need for change also impacted the qualifications of our people. We needed to install new training systems, certifying our operators in their new roles. The Conduct of Operations programme needed revision and implementation. Procedures required updating or discarding. The DSA/TSR safety controls—established before actual operations—were overly conservative, impeded production, and were revised.
• It takes a series of moves to get cylinders through the entire conversion process, most of them inside the plant, including modification of the cylinders for refilling. We believe some of these moves are cumbersome, and unnecessary in the long run. We are presently working with experts to refine and ease this system and the cranes we use for it. This becomes increasingly important as we move more cylinders, faster . . . for obvious reasons.

There is more work to do, of course. Neither we, nor our customer, know entirely what lies ahead. We have developed an operating strategy for this year that we believe will support technical growth and move us rapidly towards faster throughput, and that’s where we plan to go.

George E. Dials is the president of BWCS and the DUF6 Project manager. A graduate of the U.S. Military Academy at West Point, his career includes positions as the chief executive officer of the Waste Isolation Pilot Plant, the Yucca Mountain Project, and the Y-12 National Security Complex. He calls himself the ‘nation’s nuclear garbageman.’

Robert C. Hogg served as a senior technical advisor to BWCS and its parent company, B&W, and was recently named Paducah’s deputy plant manager. He is author of the company’s revised start-up plan, and chairs the company’s collaborative task force on throughput improvement. He previously served as a project manager for the U.S. Nuclear Regulatory Commission.