At POWER-GEN International 1999 in New Orleans in December, Randy Zwirn, Siemens WestinghousePower Corp. president, stated that the “probability of frustration is 100 percent.” He was referring to the lessons learned from the merger of Siemens and Westinghouse, but the frustration comment is universally applicable to all areas of the power generation industry, particularly operations and maintenance. Whether you’re frustrated by rising O&M costs, reduced resources or changing dispatch schedules, frustration is omnipresent and unavoidable.
Panther Creek Energy Facility. Photo courtesy of Panther Creek.
And don’t expect it to get any better in the near future. Recent events are introducing new monkey wrenches. Prices are going up for equipment and services. Baltimore Aircoil Co., for example, announced a 2 percent price increase on certain equipment. Large gas turbine prices have risen about 20 percent in the past 18 months. Dresser-Rand terminated its 16-year OEM agreement with GE to supply gas turbines for compressor packages, citing “unacceptable price hikes.” Mergers, acquisitions and consolidation within both the utility industry and the supplier industry are also impacting O&M activities, as plant personnel are changed, operating strategies are modified, service packages are altered and parts supply chains are updated.
In the face of such frustration, two strategies are critical for plant owners and operators: take a long view and demand better service. Instead of focusing exclusively on up-front capital costs for parts and equipment, find the solution that will provide the lowest average costs over a length of time, thereby enhancing unit profitability. Demanding better service acknowledges that price increases may be necessary, but that the OEM or service provider must justify that increase with service that results in lower non-capital costs and lower frustration levels.
The Long View
Point Beach Nuclear Plant. Photo courtesy of Wisconsin Electric.
Taking a long view with respect to power generating assets involves establishing an asset’s current competitiveness, determining where it needs to be to remain competitive in the future, and then making the necessary operating and maintenance modifications to get there. The Panther Creek Energy Facility-a Bechtel-designed 83 MW Pyroflow circulating fluidized bed boiler plant in Nesquehoning, Pa., managed by Panther Creek Partners (a joint venture between Constellation Power and Ahlstrom Development) and maintained and operated by Constellation Operating Services-provides an excellent example.
When Panther Creek began commercial operation in October 1992, the typical annual outage schedule included a weekend boiler inspection and a standard one- to two-week boiler outage each year. Turbine stop and control valve inspections were on a two-year schedule. Capacity factors during the first three-plus years of operation met expectations. In 1996, however, four months after the scheduled annual outage, a failed generator stator bar forced a lengthy 35-day outage to make a temporary repair. Concerned that the plant could not financially sustain two consecutive years of 35-plus days without generating revenue, plant management decided to evaluate an extended outage interval.
Trend data from predictive maintenance tests-equipment vibration, lube oil condition, boiler tube wastage, subsystem thermography-revealed only minor downward trending at the 12-month point, indicating that an additional period of operation would not negatively impact performance. Consultations with the OEMs and the insurance carrier also did not uncover any technical barriers to implementing an 18-month outage interval. Furthermore, outage reports showed that annual outage costs were consistent for repetitive projects. After considering all these factors, Panther Creek decided to forego its planned 1997 outage and establish an 18-month schedule for the boilers and a 36-month schedule for the turbine valves.
Plant management was particularly concerned with refractory maintenance costs, which typically accounted for a large fraction of the annual outage budget. Variability in yearly refractory costs led the plant to negotiate a long-term, hard-dollar refractory maintenance contract, possibly an industry first, according to plant manager Richard Gawel. The contract places the burden of unanticipated refractory damage on the supplier. The supplier responded by proposing various upgrades-such as installing precast assemblies at the loop seal return combustor interface and support bracing to prevent deterioration of the crossover ducts-that would reduce their exposure to refractory failures.
Panther Creek also upgraded selected systems, recognizing that equipment would have to run longer between inspections. Gravimetric and volumetric feeder belts were fitted with abrasion-resistant covers, ceramics were installed in the fuel feed and ash removal systems, and a portable oil filtration unit was fabricated for use on the various fan and pump bearings. Additionally, Panther Creek began performing more maintenance work while on-line, such as condenser and circulating water pump inspections during cold-weather months when only either half the condenser or one pump is required for full-load operation.
Since establishing the extended cycle, capacity factors have been 98.9, 90.0 and 96.0 percent for 1997, 1998 and 1999, respectively, with most losses attributed to utility-issued minimum generation emergencies and a downed transmission line (Figure 1). The first two inspection outages revealed no adverse conditions likely to lead to an extended or unplanned outage. Significantly, forced outages due to boiler problems have actually gone down since instituting the extended outage cycle. Both boilers experienced only one forced outage apiece for tube leak repairs during the three-year period. Most importantly, one planned outage has been eliminated every three years without sacrificing availability.
At Your Service
While OEMs and service providers may be raising prices, on the plus side, they are increasingly shifting their mindset to focus on the customer. It has become apparent that it is in their best interest to work with the power plant instead of for the power plant, and that by striving toward common, mutually agreed goals, both sides can succeed.
Evidence of this increased level of service is the growing use of baselining and improvement processes, which peg contractor payment to the performance of a given project relative to a baseline project. TVA has been using this technique for several years, initially focusing on major maintenance and capital improvement projects, but subsequently extending the practice to smaller O&M activities.2 Similar O&M projects are grouped into families-such as condenser retube, coal nozzle replacement, pulverizer reversal, hydraulic governor replacement-and then equalized by reconciling scope differences and either adding or subtracting the associated hours and dollars.
The contractor receives a performance award if there is improvement from the previous best-to-date project. The more times comparable projects are repeated, the more cost-effective the contractor should become. Table 1 illustrates how the process works. If actual costs are at least 11 percent less than the equalized value of these projects relative to the baseline, the contractor receives the entire award fee. If, on the other hand, actual costs are only 1 percent less than equalized value, the contractor must provide a 50 percent credit to TVA. In fiscal year 1998, TVA applied the process to 31 projects with an actual cost of about $18.5 million. The estimated savings from the previous “best-to-date” project performance was approximately $2.2 million, a 10.7 percent improvement.
Parts is Parts
Installation of copper reducing filter at a power plant. Photo courtesy of Pall Corp.
Additional evidence of customer service is the growth of localized service centers. OEMs are making a concerted effort to bring resources closer to the customer, following the distributed generation model. Siemens Westinghouse is adopting a multiple “home-market” presence, according to Zwirn, where parts and engineers can be supplied quickly in a given region. ABB Alstom is also expanding its service network, particularly in the gas turbine market, according to Ralph Sherwin, Industrial Turbines President. Unit profitability is optimized by ensuring that the appropriate resources are brought to bear in a timely fashion, reducing downtime.
Spare parts represent a key consideration for all plant owners. The issues include inventory levels, parts availability, quality control, forecasting, traceability and automatic ordering. As economic pressures on individual plants mount, implementing an effective spare parts strategy becomes a critical decision. Outsourcing all or part of this function may be the ideal choice.
Nuclear power plants are particularly concerned about parts availability because of the need for safety-qualified components and because a single day off-line can mean $1 million or more in lost revenue. Two characteristics have hampered the parts availability strategies at many power plants, according to Brad Strella, director with Wesco Aircraft’s Engineering Products Division. First, plants have had to maintain a large inventory of on-hand components because manufacturers and/or distributors are unwilling or unable to handle it themselves. Second, expedite fees to speed component delivery have become commonplace, driving up O&M costs.
Several utilities, including Consumers Energy, Wisconsin Electric, Illinois Power, Energy Northwest, Vermont Yankee and Southern Nuclear, are investigating inventory management solutions from Wesco that may eliminate these problems. These include everything from complex inventory just-in-time programs to simple long-term contracts. In the latter case, components will be available at a fixed, firm price for a period of two to three years. Utilities will have on-line access to Wesco’s inventory, providing peace of mind (i.e., less frustration) that parts are indeed in-stock and available. Parts are purchased only from approved nuclear vendors and are dimensionally checked both coming in and going out.
A few nuclear utilities are pilot-testing this type of arrangement by identifying a “top 100” list of usages for various fasteners available from Wesco. When a part is needed, the utility can immediately confirm on-line that it’s in-stock; the part can then be ordered at a fixed price without incurring an expedite fee. There are also opportunities for implementing buyback and consignment plans that reduce utility risk.
With consolidation evident across the entire power generation industry, standardized maintenance programs should become more common, providing better service and economies of scale that reduce prices. If an Entergy or an AmerGen ends up owning 15 to 20 nuclear power plants in the next several years, by integrating inventory functions, they can then issue blanket orders for certain components, thereby enabling larger manufacturing runs and lower costs. If successful, this technique can be expanded to many other high-use power industry components, such as valves and bearings.
Component obsolescence is another important parts-related issue, particularly for nuclear utilities. Marietta Williams, senior analyst with Southern Co., has performed a non-scientific but representative survey of the nuclear utility industry that sheds some light on the criticality of obsolescence.3
Several factors have contributed to limit nuclear plants’ abilities to maintain original design configurations:
- Vendors are willing to refurbish some equipment and supply some parts, but are discontinuing or reducing their support of others.
- An estimated 50 to 70 percent of traditional nuclear suppliers have or will abandon the nuclear indusy within the next few years, partly due to the high cost of maintaining 10CFR50 Appendix B quality programs.
- Maintaining antiquated designs with outdated production equipment is not a priority with many OEMs.
- As suppliers exit the nuclear marketplace, obtaining historical design documents becomes more difficult.
- Utilities are stocking fewer spare and replacement parts to control inventory.
The result of these factors is that prices for certain items have risen dramatically. Williams, for example, recently received a $53,000 quote to replace a coupling that cost $1,500 in 1985, a 3,400 percent increase.
Williams identifies four points that characterize current industry practice. First, plant owners typically operate in a vacuum, responding to obsolescence on a case-by-case basis and believing each instance of equipment obsolescence is unique to their plant. Second, plants operate independently of each other in a “don’t ask, don’t tell” relationship. Once a solution is found, information is rarely shared with other plants. Third, plants often recreate the wheel, incurring unnecessary expenditures to solve a problem that has been encountered elsewhere. Fourth, by producing multiple unnecessary solutions when one good solution would suffice diminishes the already small degree of common specifications among plants.
Plant owners have taken various steps to deal with equipment obsolescence, such as stocking obsolete items, repairing or refurbishing existing equipment, and qualifying commercial-grade items for nuclear safety-related applications. The problem with all of these solutions, however, is that they have limited impact on addressing obsolescence over the long term. Tom Westbrook, president of value-added reseller Divesco, notes that although 70 percent of his inventory is obsolete equipment, his stock will only delay the need for utilities to develop a longer-term solution.
More innovative solutions are in use and under development that leverage knowledge across companies to reduce redundant effort and target long-term impacts that increase the industry’s options. EPRI’s Obsolete Items Database, managed by NUS Information Services, contain records that identify an obsolete part and identify potential replacement options. Use of this service, however, has been below expectations, with no clear reason for the underutilization. NUS also sponsors the on-line RAPID spare parts directory and discussion forum, which provides another avenue for addressing obsolescence concerns.
Many suppliers have recognized the global impact of obsolescence. Suppliers need reliable information only utilities can provide regarding operating requirements and data on the size of potential markets. They are forming alliances with utilities to share design data and reduce risk to all participants. Reverse engineering, widely recognized as the best long-term solution, significantly increases utilities’ options by widening the pool of suppliers. Its use, however, has also been restricted, partly due to real barriers such as legal issues, patents, trade secrets and proprietary data, but also due to utility concerns about damaging relationships with existing suppliers. Most of these barriers can be overcome, and the willingness of some utilities to reverse engineer products has led some OEMs to make design data available for a fee.
While obsolescence is most pronounced in the nuclear industry, the techniques developed to address this issue will benefit the entire power generation industry, in the form of increased competition and lower O&M costs. Greater use of reverse engineering, for example, could reduce gas turbine maintenance and repair costs and maximize equipment availability. In 1999, the National Materials Support Group, the users group for RAPID (representing every U.S. nuclear utility and about 200 fossil/hydro stations), established a subcommittee to address the issue of generating equipment obsolescence. Objectives include designing an industry-wide strategic solution, developing industry standards embracing the use of reverse engineering, resolving risk and proprietary data issues, and recommending industry standards for sharing design data. In November 1999, 16 utilities representing 30 nuclear plants met to discuss the obsolescence issue. As a result of that meeting, they formed the Nuclear Utilities Obsolescence Group, which will work in concert with NMSG to promote awareness and solutions to obsolescence problems. There is also discussion of a similar effort in the fossil arena.
While there is no guarantee of success, such approaches represent collective efforts to ease O&M headaches, and recognize that the key to lowering O&M costs in an open, competitive market is being open to unconventional approaches. The probability of frustration will remain 100 percent, but the magnitude of frustration will be reduced considerably.
- Gawel, R., “18-Month Planned Outage Cycle Supports Plant Financial & Reliability Goals,” POWER-GEN International 1999, New Orleans, La., Nov. 30 – Dec. 2, 1999.
- Simpson, T.E. and M.H. Gillstrap, “TVA’s Project Baselining and Improvement Process,” POWER-GEN International 1999, New Orleans, La., Nov. 30 – Dec. 2, 1999.
- Williams, M.C., “Replacing Obsolete Items – Industry Practice and Innovative Solutions When Upgrading Is Not the Answer,” POWER-GEN International 1999, New Orleans, La., Nov. 30 – Dec. 2, 1999.
Copper deposits on steam turbine blades reduce turbine output and often result in costly outages. As more and more copper dissolves from copper alloy tubing in heat exchangers and condensers and then deposits in the steam path, output is increasingly curtailed. At a certain point, plants must be shut down to perform a chemical cleaning that will restore the turbine to original capacity. Once the unit is brought back on-line, however, copper fouling begins again. Various on-line copper reducing technologies have been developed and evaluated in the past, but with little success.
Several utilities have recently tested a new copper reducing filter from Pall Corp. that reduces total copper and iron concentrations in feedwater systems. Virginia Power’s Chesterfield Station evaluated the filter for its ability to reduce copper concentrations during plant start-up; at Chesterfield Unit 6, copper contamination is found throughout the condensate/steam loop, with peak concentrations appearing during plant start-up. As shown in Figure 2, copper concentration downstream of the filter was maintained at
a low level throughout the first 14 hours of start-up, and stayed low even after a copper spike associated with a crud burst caused by a system blow-down, effectively protecting the turbine from copper deposition damage.
At Pacificorp’s 330 MW (net) Naughton Station, Unit 3 had historically been forced to derate because of copper plating. In 1995, Naughton placed condensate particulate removing filters into continuous service to remove solids, and extended the time period before derating occurred from about three weeks to about three months. Recent tests with the copper reducing filter indicated that it could further extend the time period before derating, leading Pacificorp to purchase a full set of copper reducing elements for installation in early 2000.
Field testing has demonstrated that the filter can reduce copper concentrations by more than 95 percent, from 600 ppb to less than 20 ppb. Two to three disposable filter changeouts are estimated each year to maintain effectiveness, but the cost should be offset by the increased interval between chemical cleanings and the recovered generation capacity. Filter changeouts will be dependent on the initial influent level of copper and iron. As deposits within the plant piping are filtered out, the service life of the elements will increase as a result of less residual contamination. Once an equilibrium contamination level is achieved, the filter maintains an excellent life and level of cleanliness since there is so little dirt left to foul the filters in the condensate loop. Some systems now operate for 18 months without need for a filter change.