For many hydro facilities, the use of reliability-centered or condition-based maintenance practices extends the duration between maintenance routines, improves unit availability, and even eliminates some forced outages. Five experts share why they use these types of maintenance practices.
Reliability-centered maintenance (RCM) and condition-based maintenance practices can extend the duration between maintenance routines, improve unit availability, and eliminate some forced outages. In spite of these benefits, justifying the required initial investment to change from traditional maintenance practices, and the higher budgets required for these preventative-type efforts, can pose challenges.
Hydro Review conferred with key staff of five hydro project owners – individuals with experience in implementing RCM and condition-based maintenance programs – and these individuals shared their insights on this important topic. We consulted with:
- – Thomas W. Brittain, P.E., mechanical engineer with Grant County Public Utility District (PUD). Tom is one of the chief RCM analysts and is the lubricants manager for the condition-based maintenance program for Grant County PUD’s hydro generation facilities.
– Dan K. Golden, manager of the 28.6-MW Bishop Creek project for Southern California Edison. Dan coordinates all RCM activities, as well as handles budgetary issues and manpower allocation during outages.
– Viggo Lundhild, formerly general manager, Sault Hydro Operations, with Brookfield Renewable Power Inc., now vice president of the company’s system control center. Viggo was the planning supervisor responsible for implementing the company’s condition-based maintenance management system, as well as reporting on the condition of assets to the general managers.
– John MacPhee, section manager for performance and benchmarking with Ontario Power Generation. John co-managed a leading-edge maintenance program to implement streamlined RCM at the utility’s 30 large hydro stations and continues to support this initiative on request.
– Thomas J. McDermott, director of asset management with the New York Power Authority (NYPA). Tom is responsible for developing NYPA’s reliability centric asset management program, overseeing its implementation, and providing the analyses associated with the decision support components.
Hydro Review: How long have you used an RCM or condition-based maintenance system?
Tom Brittain: We have performed RCM analysis of our maintenance processes and procedures for about six years. Much of this time has been spent working with our computerized maintenance management system (CMMS) and in the implementation of the results of the RCM analysis. So we have not been working on analyzing the maintenance procedures the entire time.
We have always done some level of condition-based maintenance. This ranged from simple operator inspections of equipment condition leading to corrective maintenance to the performance of detailed, precise electrical tests to determine equipment condition. The RCM analysis of maintenance practices has led to the increased use of condition-based maintenance where we use predictive maintenance technologies like oil analysis, vibration analysis, and Doble testing.
Dan Golden: We have used this type of system for at least 11 years.
Viggo Lundhild: We have used components of an RCM system since the late 1990s.
John MacPhee: In 1999, Ontario Power Generation initiated a project to revise the preventive maintenance program at its large hydro facilities. A streamlined RCM tool was chosen to facilitate this initiative. By the end of 2004, the updated maintenance program was integrated into the corporate work management system and was fully functional across the entire hydro fleet.
Tom McDermott: NYPA has been using the concepts behind RCM and condition-based maintenance for the past several years. Over the past five to seven years, NYPA began using condition-based monitors on certain critical equipment. NYPA is now expanding its methods to capture data from a greater number of equipment and instrument sources that have been identified as critical to operations. NYPA has broadened the concept of RCM and uses the more inclusive term reliability centric asset management (RCAM), covering equipment and business operations.
Hydro Review: Please describe the system.
Brittain: To perform the RCM analysis, we used the generally accepted method of performing criticality analyses and failure modes and effects analyses to determine what maintenance should be done. We did not use complex statistical analyses because we did not have the time or the accurate data to justify use of more detailed analysis techniques. Thus, we took a heuristic approach using what maintenance history data we had, and now we observe the results of the new maintenance processes.
RCM should be a “living program.” Modification of procedures after they have been analyzed is considered normal and desirable when new field information comes forward. This information can be an observation, a bad or good trend in predictive maintenance data, an equipment failure, or the perfect operation of a system that can show we are doing the right work, but do not need to do the maintenance as often.
When it works right, condition-based maintenance is an integral part of a web of efforts that is a bit like a well-orchestrated piece of music. A single piece of equipment may have frequent operator inspections performed, predictive testing such as oil analysis or vibration analysis, preventive maintenance performed based on the need to keep the equipment clean and lubricated, and electrical testing and inspection of a driving motor and its controls. The key is that the RCM analyst has looked at all of the various maintenance tasks, takes credit for performance of each task, and directs each task toward preventing failures, not just doing something because it is a good thing to do.
Golden: Division orders dictate when maintenance needs to be performed on various components of the hydro generation assets. We hold an annual meeting of the key players to set maintenance priorities, based on available resources (manpower and budget). Division orders are reviewed at this meeting and adjusted to account for experience, generic industry issues, and maintenance history. Monthly planning meetings also are held to review the RCM system and to make any appropriate adjustments.
Lundhild: We do not have a formal RCM system. Our system is based on a traditional preventive maintenance system. We have overlaid a risk component analysis of each piece of equipment, which has identified a priority level for the maintenance routines for each piece of equipment. The risk evaluation was completed using the criteria of length of duration to repair a failure of a piece of equipment, as well as cost associated with the failure (lost production, direct costs).
Specific maintenance work procedures have been developed for each piece of equipment, and then matched in scope to the criticality of the equipment. High-risk equipment (high probability of failure and high cost to repair) is given a large amount of attention. These procedures are heavily focused on determining the condition of the equipment. Engineering review of the completed inspections allows a condition to be determined for each piece of equipment.
Additionally, we use oil analysis on all power transformers and annually perform an infrared scan of all electrical components. To complement this information, we have on-line, real-time performance indicators for all the critical components of our units. This includes vibration monitoring, oil temperature, bearing temperature, cooling water flow/temperature, and partial discharge analysis.
MacPhee: The streamlined reliability centered maintenance (SRCM) process involves breaking down a facility into a number of systems and sub-systems. The functions of the systems are described. Then the pieces of equipment within those systems are determined to be “critical” or “non-critical,” based on the consequence of their failure in relation to the system function. A set of criteria including production as well as safety and environmental consequences is used to determine criticality. Failure modes and effects are analyzed, and appropriate maintenance tasks are selected from maintenance templates according to facility class and criticality to address the failure modes. Wherever possible, condition-based tasks are chosen over time-based tasks.
McDermott: The RCAM program starts with identifying measurable performance parameters that are indicative of the equipment’s well-being. This data is then integrated using software devices (a state estimator) and performance models to determine if the equipment (or system) is in the normal, close-to-normal, or not normal range. A decision support system (DSS) evaluates these conditions (using equipment and system reliability models that have been developed) to assist in the analysis of when to intervene and perform maintenance in order to return the equipment or system to its normal operating state. This entire infrastructure serves as a resource for maintenance engineers, equipment specialists, and craft workers, who will review the results and ultimately make a decision when maintenance is needed.
The overriding principle is to reduce invasive recurring maintenance and depend on knowledge-based performance analysis, helping to ensure high standards for serviceability of NYPA’s assets.
Vibration monitoring is just one of many predictive maintenance practices at hydro plants that can be prioritized through use of a condition-based maintenance program.
Hydro Review: Why did you choose to implement this system?
Brittain: The maintenance engineering supervisor pursued the use of RCM to analyze our maintenance practices based on what he saw the rest of industry doing. He was involved in the Institute of Electrical and Electronics Engineers (IEEE) Rotating Machinery Committee and saw a cross-section of industry in North America moving toward this process, and the process made sense for us. We performed a work process review of all functions at our hydro plants, including how we decide what maintenance to do. This review formally recommended that RCM be used. An economic justification that used assumptions based on typical results expected from performing an RCM review was then approved.
Golden: SCE implemented RCM in order to base maintenance on historical experience and to reduce forced outages resulting from equipment failure.
Lundhild: We determined that there was value to be found in an increased availability factor (market opportunities required high reliability), the ability to extend the duration between maintenance routines, and the opportunity to eliminate some of our forced outages. Additionally, we determined that we could reduce our overall operations and maintenance costs if we diverted some of our focus away from reactive maintenance and unnecessary preventive maintenance. The system was not implemented “all in one shot” but rather has grown from the original plant management system as affordable technology became available.
MacPhee: Historically, although preventive maintenance standards existed at our hydro facilities, they were primarily time-based tasks applied at predetermined frequencies, and they tended to be conservative. The standards were the same for all facilities, regardless of the value of that facility to the company. A portfolio approach to investment had been instituted for capital investment, and the company desired to adopt a similar approach for maintenance. To achieve this, the plant managers needed a tool to help apply their limited maintenance resources more effectively.
McDermott: From a maintenance and engineering point of view, it was critical to begin capturing the knowledge of our experienced staff, many of whom are either close to retiring or will retire within a few years. By capturing this knowledge – through the development of performance- and decision-support models fundamental to the RCAM system – we will have access to and benefit from years of operating experience with NYPA’s plant equipment.
Efficiency in operations can be achieved by performing maintenance when it is needed rather than on a strict time-based system. RCAM allows you to look at critical equipment, which will have a negative effect on your operations if it fails, rather than taking valuable time to review the status of non-critical equipment. Non-critical equipment is allowed to fail and then be repaired or replaced in the most economical and least time-intensive ways.
Hydro Review: How did you justify the cost of implementing the system?
Brittain: Justification was based on saving money through more precise use of maintenance personnel and more proactive maintenance techniques. While reducing downtime due to forced outages (improved reliability) and for maintenance processes (improved availability) were recognized as goals, the costs/benefits of these reductions were not factored into the decision.
Golden: Forced outages are costly, both in lost generation and in customer satisfaction. A systematic approach to maintenance, that seeks to reduce forced outages, has a significant cost advantage.
MacPhee: The cost was justified by both financial and non-financial benefits. On the financial side, implementing SRCM was expected to improve unit availability by at least 0.5 percent annually. Additionally, a 5 percent reduction in maintenance effort was expected. Benefits that were not quantified included the application of consistent maintenance effort and tactics across the business based on the maintenance strategy appropriate for the facility class (portfolio approach). As well, the maintenance program and its basis would be properly documented, which is particularly important with an aging workforce.
McDermott: NYPA, like many utilities, is experiencing a shrinking pool of talent for these types of time-based maintenance jobs. The need for human input is critical in RCAM decisions. To the degree that we can develop an infrastructure that takes in the information – both from staff and from instruments – then systematize it to make decisions, NYPA can maintain its high reliability standards while reducing production cost.
Hydro Review: Was it funded as a new capital expenditure or through your maintenance budget?
Brittain: We funded the process as a capital expenditure because its cost was high enough, and the results would be used for much longer than five years.
Golden: We funded it through the maintenance budget.
Lundhild: Some of the project was funded through a capital investment, justified through our ability to redirect our internal work force availability to regular preventive maintenance, as well as to reclaim some capital work that was being contracted out. Much of the monitoring equipment was justified by adding those costs to other capital work on the equipment and piggybacking on other operational investments, such as the supervisory control and data acquisition (SCADA) system.
MacPhee: The SRCM initiative was funded from the maintenance budget.
McDermott: It is mostly a capital expenditure, with some amount of operations and maintenance expenditures.
Hydro Review: What has been the biggest implementation challenge?
Brittain: Our challenges have involved keeping engineering manpower on task to complete the RCM process and getting engineering manpower to analyze predictive maintenance results. We are still doing the RCM analysis on some systems.
While there was initial resistance to the RCM process from some of the work force, that has gone away with time. Presentations explaining the process and goals helped to dispel the rumors and dark humor. Craftsmen also have seen that we have too much work to do, nobody is going to get laid off, and the work that comes out of RCM makes more sense than a lot of the work that was performed in the past. They also have been a part of the process through discussion of the work that they perform and through validation of new procedures with the engineers who performed the RCM analysis. Most craftsmen now are pleased that their input to improving job plans is welcomed and valued.
Golden: Our biggest challenge has been supporting the RCM effort with appropriate staffing levels.
Lundhild: Buy-in from the maintenance staff was originally a hurdle, but it has been accepted over time.
MacPhee: Getting staff acceptance was a challenge, as the move to RCM required a culture change in maintenance philosophy from the traditional technical approach to a risk management approach. There also were concerns that this process was being implemented to reduce staff levels. To obtain buy-in, the program development team included representation from each major production organization in the hydro business. At least one plant from each of the major organizational divisions was analyzed by the development team. The team members from each plant group were charged with implementing the program to the remaining plants in their group.
McDermott: The biggest challenge is managing the transition from a time-based maintenance system to one that that is more reliant on condition data and performance models. Accomplishing this objective requires developing the confidence that the resulting analysis is accurate, robust, and sufficient to sustain the standard of service that we have all come to expect.
Hydro Review: What has been the return on your investment, financially and in improved maintenance?
Brittain: We have not looked at the cost to date and the results of that effort. Our accounting system makes this difficult. In addition, management believes that the RCM effort is worthwhile, so they are not interested in spending time in further justification.
One immediate improvement has been in the area of equipment lubrication.1 The RCM process pushed us to use oil analysis, and oil analysis has stopped the practice of changing oil based on elapsed time, such as annually. We now change or filter oil based on the results of an oil sample. This saves money by not changing oil that is still good, and by filtering oil that is good but dirty. We have not quantified the return on investment, but we feel this process pays for the cost of performing the oil sampling and analysis. The largest benefit, which is hard to quantify, is the improvement in equipment availability and reliability and the long-term improvement in equipment life because it has better lubrication.
Golden: Much better planning for maintenance reduces costs and assures adequate staffing levels. Although it is difficult to place a monetary value, clearly the savings have exceeded the initial implementation costs.
Lundhild: It is difficult to quantify the return. Our forced outage rate has consistently improved, but this also may have been affected by other capital improvements. Our overtime and call-out costs and occurrences have progressively dropped.
MacPhee: The results are difficult to measure because there are many other contributing factors, including significant capital investment in the facilities. Before instituting the revised program, the five-year average availability was 90.7 percent. Over the past five years, availability has improved to 92.9 percent, well over the expected 0.5 percent.
The change in maintenance effort varied from facility to facility, ranging from little or no change at some sites to a 45 percent reduction in some cases. However, even in cases where there was minimal net change in effort, the focus has been changed to ensure the right things are done.
An important side benefit is that we have an up-to-date documented inventory of equipment condition and maintenance standards for all plants.
McDermott: The transition to a fully functional RCAM system is anticipated to take five to eight years. The first years are devoted to infrastructure and model development. The pace of implementation is coincident with the projected retirements of NYPA staff vital to the preparation of the knowledge capture paradigms upon which the program is based. n
Mr. Brittain may be reached at Grant County Public Utility District, 15655 Wanapum Village Lane S.W., Beverly, WA 99321; (1) 509-754-5088, extension 2508; E-mail: email@example.com. Mr. Golden may be reached at Southern California Edison, Bishop Plant 4, 4000 Bishop Creek Road, Bishop, CA 93514; (1) 760-873-0725; E-mail: dan.golden@ sce.com. Mr. Lundhild may be reached at Brookfield Renewable Power, 480 Boul. De la Cite, Gaineau, Quebec J8T 8R3 Canada; (1) 819-561-2722; E-mail: firstname.lastname@example.org.Mr. MacPhee may be reached at Ontario Power Generation, 14000 Niagara Parkway, RR 1, Niagara on the Lake, Ontario L0S 1J0 Canada; (1) 905-357-0322, extension 7386; E-mail: john.macphee@ opg.com. Mr. McDermott may be reached at New York Power Authority, 123 Main Street, White Plains, NY 10601; (1) 914-681-6534; E-mail: thomas.mcdermott@ nypa.gov.
- Brittain, Thomas W., “Experiences with Establishing a Lubricant Maintenance Program,” Hydro Review, Volume 26, No. 7, November 2007, pages 20-29.