Maintenance and reliability warnings

Turbomachinery Symposium

Klaus Brun, program director of the machinery program at Southwest Research Institute (SWRI), will present “Gas Turbine Degradation” with Rainer Kurz of Solar Turbines and Cyrus Meher-Homji of Bechtel at the 43rd Turbomachinery/30th Pump Symposia in Houston on Sept. 23 at 8:45 AM. The tutorial will explain deterioration mechanisms, including compressor and turbine fouling, erosion, increased clearances, and seal distress, along with their manifestations, rules of thumb, and mitigation approaches. The treatment will deal with simple-cycle gas turbines in power generation and mechanical drive applications and will also address the impact of performance deterioration on combined and cogeneration cycles. Brun and Timothy Allison, also with SWRI, will present “Acoustic Instability in Pilot-Operated Pressure Safety Valves” at the 43rd Turbomachinery/30th Pump Symposia in Houston on Sept. 25 at 10:30 AM. In this case study, pilot-operated pressure safety valves (PSVs) will be shown to be susceptible to a dynamic instability under certain conditions where valve dynamics couple with upstream piping acoustics. This self-exciting instability can cause severe oscillations of the PSV piston, damaging the valve seat, preventing resealing, and possibly causing damage to downstream piping. Test data will be presented, showing damaging unstable oscillations in a blow-down rig, and a methodology for modeling PSV acoustic instabilities will be explained. Modeling results will be compared with measured unstable operation in a test rig to show that the modeling approach accurately captures PSV behavior near unstable conditions. Learn more about the 43rd Turbomachinery/30th Pump Symposia at http://pumpturbo.tamu.edu.

Dr. Klaus Brun is the program director of the machinery program at Southwest Research Institute (SWRI) in San Antonio, Texas. His experience includes positions in engineering, project management, and management at Solar Turbines, General Electric, and Alstom. He holds six patents, has authored more than 150 papers, and has published two textbooks on gas turbines. Dr. Brun has won an R&D 100 award in 2007 for his semi-active valve invention and ASME Oil & Gas Committee Best Paper awards in 1998, 2000, 2005, 2009, 2010, and 2012. He was chosen to the "40 under 40" by the San Antonio Business Journal, and he’s the past chair of the ASME-IGTI board of directors and the past chairman of the ASME Oil & Gas Applications Committee. Dr. Brun also is a member of the API 616 and 692 task forces, the Middle East Turbomachinery Symposium, the Fan Conference Advisory Committee, and the Supercritical CO2 Conference Advisory Committee.

PS: How can plants use statistical and probabilistic tools to mitigate unpredictability in component performance?

KB: Basic statistical algorithms allow for the prediction of mean time between failure of individual components and can be combined to determine overall probability of plant failures. Thus, these tools can be valuable to determine maintenance or scheduled plant outage events and to avoid forced outages. However, one has to be careful since risk of failure increases with component operating hours. An over-reliance on time between failure statistics for maintenance, parts replacement, and overhaul intervals can lead to an increased risk of costly catastrophic failures and forced outages.

PS: Can you describe how Monte Carlo simulation and risk analysis more accurately define process uncertainty and its impact on machine performance?

KB: Monte Carlo simulation randomizes plant process input variables and performance parameters, usually — but not necessarily — based on a Gaussian distribution. This results in the ability to model the plant’s behavior not at a single operating point but over a realistic wide range of operating points. Other statistical methods exist that achieve similar simulation results, but Monte Carlo, based on numerical simplicity, is one of the easiest to implement into solver algorithms.

PS: Should plants be designed to perform under most-likely scenarios or worst-case scenarios?

KB: That depends on the operating philosophy of the plant. For example, a power plant that provides emergency power for a hospital should be designed to continue operating under worst-case scenarios. On the other hand, a plant with adequate system redundancy or backup power could be designed to less stringent and less costly average-based statistical operating scenarios.

PS: How can a gas turbine operator’s understanding of performance characteristics help with equipment sizing, as well as limiting degradation and deterioration?

KB: One of the most common problems in the industry is over-sizing of plants due to overly conservative estimates of component performance characteristics and over-estimating degradation or deterioration. By properly understanding and modeling the total plant performance based on accurate individual component performance curves, one can correctly size a plant to meet the application requirements without over-sizing it. This saves both capital and operating costs.

PS: How does proactive condition monitoring allow gas turbine operators to make intelligent service decisions based on actual equipment condition, rather than on time-based maintenance intervals?

KB: A well-implemented condition-monitoring system can allow operators to increase maintenance intervals and reduce plant operating costs. However, all condition-monitoring systems are only as good as the instrumentation accuracy and availability that are being utilized within the condition and failure prediction algorithms. Improper condition-based maintenance practices can lead to severely increased risks of plant failures.

PS: Can you describe water-washing and best practices?

KB: Modern gas turbines are usually supplied with on-line and off-line axial-compressor washing systems that consist of a nozzle spray manifold mounted at the bell-mouth of the gas turbine. Keeping the axial compressor clean is critically important for the total performance of the gas turbine since 55-65% of the power produced by the machine’s turbine section is utilized to drive its compressor. On-line and off-line washing procedures are very machine-, site-, and application-specific and an operator should spend some time and effort to optimize these procedures for each individual machine to limit degradation and failure rates. On-line washing should be performed frequently to be effective and only extends off-line washing intervals. Off-line washing is far more effective in removing deposits from the axial compressor but requires significant downtime since the machine needs to be shut down and allowed to cool before off-line washing can be started.