By Teresa Hansen, Senior Editor
Spring 2008 was one of the busiest nuclear refueling outage seasons the U.S. nuclear industry has witnessed in several years. According to the Nuclear Energy Institute (NEI), 41 reactors either completed or were undergoing refueling outages at the end April. The average duration of these outages was 36.7 days. But consider that as recently as 1990, NEI data showed the average refueling outage lasted 104 days. In 2007 that figure shrunk to 40 days (Figure 1).
“The bar has been raised” when it comes to outage duration and efficiency, said Chris Vlahoplus, partner with ScottMadden Inc., a general management consulting firm that specializes in the energy industry. In this past outage season, those units in the top quartile averaged 25 to 27 day refueling outages. Around 14 percent completed their refueling outages in less than 25 days, he said.
A few units have refueled in 20 or even 15 days, but not many. In 2002, the year when the industry achieved its shortest average outage duration ever, Tennessee Valley Authority’s Browns Ferry 3 (a boiling water reactor in Alabama), set an industry record for the shortest refueling outage of any U.S. plant, completing work in 15 days. The following year, Exelon Generation’s Braidwood 2 (a pressurized water reactor) in Illinois also completed its refueling outage in 15 days.
The industry has seen not only a reduction in average refueling outage duration, but also in steam generator and reactor head replacement outages, which have become common in the past 15 years.
The first steam generator replacements in the United States occurred at Dominion’s Surry plant in Virginia. The three steam generators in Unit 1 were replaced in 1979 and the three in Unit 2 the following year. The Unit 1 steam generator replacement outage took 303 days and, while the Unit 2 replacement outage was almost one-third shorter than Unit 1, it still took 209 days. A few more plants replaced steam generators in the 1980s, but by the mid 1990s, steam generator replacements had become commonplace.
In 1990, the average steam generator replacement outage was 160 days long. By 2007, the average was approaching 75 days. Photo courtesy URS.
According to Richard “Dick” Miller, senior project director with Bechtel Power Corp., in 1990, the average steam generator replacement outage was 160 days. By the mid-1990s, average duration was down to somewhere between 80 and 90 days. And, by 2007, the average was approaching 75 days. A few steam generator replacement outages have come in below that. In 2007, Luminant’s Comanche Peak’s steam generator replacement in Texas was completed in 55 dayseight days less than the previous record for steam generator replacement.
Several things have led to shorter outage duration, said Vlahoplus, but “basically, plant owners have learned to run more efficient outages that save both time and money.”
“Outage durations have decreased because contractors and plant personnel now plan and execute better than before,” said Art Lembo, President of SGT, a joint venture between AREVA and the Washington Division of URS Corp. “More risk assessment and better contingency planning have also been beneficial.” In addition, because steam generator outage durations are often affected by refueling activitybefore and after the generator changeoutthe improvement in refueling outage efficiency and duration has played an important role in decreasing steam generator outage duration.
Owners have made significant advancements in conducting refueling outages in ever decreasing time by limiting outage work to refueling activities and then focusing only on the work necessary to comply with regulations, commitments and selected other activities, said Miller. Owners now work to keep backlogged work to a minimum so that a scheduled refueling outage contains less work than was common in the past.
“Outages are also shorter because the NRC (Nuclear Regulatory Commission) mandated modifications are not as frequent as they once were,” said Peter Kopyscianski, URS Director of Nuclear Construction. “The industry has matured and the modifications are not as great as before. This has allowed utilities to reduce the outage work scope.”
In addition, the people working in the industry have also matured. Kopyscianski said the industry has more than 20 years of experience with the sorts of planned maintenance activities typically performed during outages. The Institute of Nuclear Power Operations (INPO) issued best practices focusing on many outage activities and the industry has simply improved due to more experience and information.
Kopyscianski said one big change has been in the outage preparation model. In earlier years, mandated scope changes were often incorporated into the outage schedule before the engineering was complete. It wasn’t unusual for engineering still to be underway even after the outage began. In addition, many of the work scopes where engineering had already been completed were only a week or two in advance of the outage start. These situations made it all but impossible for personnel to effectively plan and schedule work. Material, equipment and labor needs were not always fully determined before the outage began. What’s more, project scopes were not always fully defined, meaning they could not be effectively evaluated and assessed. Also, little time was available before the outage for a thorough risk analysis, so contingency plans were not established. This practice led to delays and sometimes inadequate levels of response during outages.
“Today, utilities do not include work in an outage schedule unless the engineering work is completed and it has been fully scoped,” Kopyscianski said. Early engineering allows the utility to identify materials, equipment and labor needs, determine how to stage equipment, evaluate risks and put contingency plans into place.
Vlahoplus agreed with Kopyscianski. “Today’s plants operate in a different culture than 20 or 30 years ago,” he said. The culture at nuclear power plants is much stricter now. If a job scope is not in the schedule, it likely will not be performed during that outage. In past years, it was common for additional work to be added after the outage began. “This is not done today unless the new job is absolutely critical,” he said.
Jeff Knisley, outage manager at Pacific Gas & Electric’s Diablo Canyon Power Plant, cited INPO as a driving force behind enhanced outage performance, particularly in the areas of safety and plant reliability. He said INPO is a valuable resource for peer support. “The industry is determined to work together and help each other succeed.”
Improved refueling equipment reliabilitysuch as the manipulator crane, the fuel transfer system and polar craneshas also led to shorter outage durations, Miller said. Many owners have also modified their plants to help decrease outage time. For example, they’ve improved the speed of their fuel handling equipment and installed improved reactor head packages to speed up head removal and reinstallation.
Knisley agreed and said “more advanced technology enhances efficiency and reduces the time it takes to get the job done.” He cited the vacuum degas process used to refill the reactor coolant system and more accurate machine welding techniques as key outage duration reducing technologies.
In addition, contractors specializing in refueling and/or inspection activities have further improved their capabilities by retaining experienced personnel and consistently assigning them to the same plants. Planning also has improved, along with abilities to recognize resource-related bottlenecks in the schedule and plan around them, Miller said.
Miller said most owners now have dedicated full-time outage management organizations whose sole functions are to ensure that outages are planned, organized, directed and controlled as efficiently as possible. These organizations have matured in the last decade. As a result, completing outages safely, on schedule and under budget has become the rule rather than the exception. Outages are managed like a business, with risk assessments performed and risk mitigation in place long before the outage starts. Performance indicators are established and tracked on a real-time basis.
“Industry leaders have set high standards for outage duration,” said Knisley. “Everyone in the industry is trying to meet or exceed these standards.” Economic considerations have also played a major role in bringing down outage durations. “Our outage goals are put together based on the amount of time it takes to safely get the job done. But like everything today, we are impacted by rising energy costs.” The longer a nuclear power plant is offline, the more costly it is to the company, Knisley said.
Earned value management has helped improve the owners’ abilities to forecast and move resources on a day-to-day basis. “In the past few years, owners view outage performance as a challenge and an opportunity to excel instead of dreading it,” Miller said.
Workforce shortages are common and labor is one area where shorter outages in general have caused headaches for nuclear power plant owners.
“Shorter outage durations have had a large impact on the availability of outage workers,” said Miller. That’s because most outage workers are travelers, meaning they have a home somewhere other than near a nuclear plant. Most were attracted to outage work because of the potential to work many weeks with lots of overtime hours. Shortened outages have made this kind of work less attractive, at least when compared to other jobs that offer these same workers longer stretches of employment and overtime.
People who make their living by traveling from plant to plant for outage work have always been a staple. “As outages become shorter, it may become more difficult to recruit skilled labor to travel all the way to the Central Coast of California for a 16-or 18-day job,” Knisley said.
In addition, a shortage of both union and nonunion craft labor exists, said Kopyscianski. “We’re finding that we must bargain with both to attract them and get them to stay on the job until it is completed.”
Some bargaining chips include higher wages, per diems, incentives for taking the job and coming to the site and various bonuses and incentives for staying on the job; these incentives often increase the longer the employee stays on the job. For example, if the employee stays two weeks, he or she might get one incentive. If he or she stays another two weeks, then a second, larger incentive might be awarded, and so on. If the employee stays until the job is over or until he or she is laid off, then another larger bonus is often awarded.
“So far, we aren’t paying nonunion labor quite as much as we are union labor,” Kopyscianski said, “But it is headed in that direction. Some of the companies/clients we work with are already paying nonunion workers as much as they’re paying union workers.”
The labor shortage affects all of North Americanot just the United States. In Canada, too, competition exists not only from others in the power business, but from other industries, especially the petro-chemical industry.
The labor shortage will only get worse as demand continues to increase and older workers retire. Just recently, the average age of our craft labor on a steam generator replacement project was 45 years old, said Miller. Out of the 40 crane operators, the average age was 57.
Miller said that in addition to using bonuses and incentives to attract labor, Bechtel works with companies that supply specialty contractors. Part of the company’s contingency planning involves packaging all of one type of work, such as welding, into a particular time span, for example, a two-week window, he said. The specialty contractor then schedules this window of work with similar work at other plants to try to guarantee individuals several months worth of continuous work at different plant sites. This is more attractive to workers than fragments of time during several months.
Lembo said he also expects the labor shortage to get worse as demand continues to increase and older workers retire. More than ever before, more labor resources are needed. Demand is rising and URS is trying to come up with ways to attract younger people to the industry, he said.
Are Shorter Outages Worth It?
The average duration of a refueling outage has been holding steady for several years. The experts interviewed for this article all agreed that further drastic reductions are not likely.
“Some improvements in duration could still be obtained,” said Vlahoplus. “But I don’t believe refueling outage records are going to continue to drop significantly. The rest of the industry does, however, have room to close the gap on the top performers.”
Miller agreed. “I see the average refueling outage duration coming down some over the coming years. But drastic improvements like 10 more days shorter, such as what has been accomplished in the last decade won’t happen,” he said. The physical work of removing a reactor head, shuffling and reloading the core and closing up/starting up limits how much improvement can be made.
“Some plants got there (to shorter outages) quickly and then flattened out; others took longer to get there,” Lembo said. “However, generally refueling outage durations have stabilized.”
As for steam generator replacement outage durations, the predictions are similar.
“As a steam-generator replacement company, we are constantly looking at and making investments in ways to shorten outages,” Lembo said. The industry has been able to shorten steam generator replacement outages and optimize resources through better planning, better welding and rigging techniques, better contingency planning and better risk assessment.
Lembo said he believes room probably exists for more improvement. But the question is whether it is worth the investment. If improvements can be made through advances in existing operations, such as rigging and construction sequencing, without having to increase manpower then “we will certainly explore it,” he said. “In most cases, however, to get additional reductions on a steam generator outage would require more manpower resources. Because manpower is already an issue, recruiting and paying to attract the additional manpower would be expensive.”
Speaking from an owner’s perspective, Knisley said that shorter outages must be looked at individually. “Our business plan calls for shorter outages in the future,” he said. “However, this should be looked at on a plant-by-plant basis, as some plants may not be able to adjust operations to reduce outage duration any further.”
The experts agreed that a point exists where shortening outages probably doesn’t make sense. “Sacrificing plant reliability for shorter outages not only doesn’t make economic sense, it would compromise the safety of the plant and the employees,” said Knisley.
“Every plant is striving to find the right balance,” Lembo said. Each individual plant must decide if the extra expense involved in shortening outages is warranted.”