EXAMPLE 1 – Dealing Proactively With Gas Turbine Performance Degradation

The two basic types of performance degradation mechanisms in gas turbines are non-recoverable degradation and recoverable degradation. Non recoverable degradation occurs from physical wear and damage to internal components and only can be recovered with major overhaul and repair maintenance. Most recoverable performance degradation in gas turbines is due to contamination and can be recovered by online/offline washing and air filter replacement/cleaning. Due to varying operating conditions at each installation site, the best practice is to initiate the mitigation action based on its condition instead of time based. Monitoring compressor discharge pressure is the most effective way to keep track of fouling. Utilizing real-time process data from a distributed control system (DSC) and data historian, a simple automated tool can be developed that effectively monitors engine compressor performance and triggers mitigation actions.

The automated tool calculates the actual pressure, compressor, discharge (PCD) variation percentage and trends it to monitor performance degradation. Mitigation action can be initiated at the right time, when the threshold of the PCD variation percentage of -5 percent is reached as per original equipment manufacturer (OEM) recommendation, thereby securing recoverable degradation.

Figure 1: Snapshot of gas turbine compressor PCD variation percentage trend showing pre- and post-performance improvement after conditionbased maintenance activity

EXAMPLE 2 – Pump Performance Monitoring to Maximize Reliability Uptime % Out of BEP Zone

Centrifugal pumps account for the majority of rotating machinery in process plants. It is often observed that the centrifugal pumps are selected oversized to accommodate higher capacities. Fixed speed centrifugal pumps have a fixed best efficiency point (BEP) that coincides with delivering highest reliability. Continuous operation of a centrifugal pump far away from BEP affects pump performance and shortens pump life. Running multistage, high energy pumps <3000 kW at 30 percent BEP could reduce pump life by 50 percent.

Figure 2: Pump curve vs. reliability curve

Utilizing process data from the data historian, a simple, automated tool can be developed that could effectively monitor the pump’s uptime percentage out of the BEP zone (i.e., 80 to 110 percent of BEP). If the plot shows the pump operated out of BEP zone most of the time, that would reduce pump life. Accordingly, the reliability engineer can take necessary mitigation action to bring the pump within the BEP zone for long-term reliable operation.

Figure 3: Snapshot of automated pump uptime percentage vs. uptime percentage outside BEP zone

EXAMPLE 3 – Variable Speed Pump Operating Envelope

The real-time information from the data historian can be utilized to monitor the operating performance of variable speed pumps to take timely mitigation action.

Figure 4: Snapshot of real-time operating envelope for a variable speed pump

EXAMPLE 4 – Monthly Uptime % and Uptime % Rolling Trend

Automating the monthly uptime percentage graphically shows the monthly run hours performance of the assets. A lower uptime percentage can be investigated for critical assets for reasons of downtime (e.g., planned, unplanned, standby, etc.) to take necessary actions. Long duration standby machines (more than three months) can be easily picked to follow up on periodic test runs to identify and fix hidden failures.

Figure 5: Snapshot of automated monthly uptime percentage for a compressor

An uptime percentage rolling trend graph demonstrates the uptime performance trend of the asset, indicating whether it is improving or declining.

Figure 6: Snapshot of automated uptime percentage trend for a compressor

EXAMPLE 5 – Monthly Trip Rate (Number of Restarts/Month) and MTBT Trending

For machines with higher uptime and running within BEP zone most of the time, if they have frequent trips followed by restarts, this also would be a threat to reliable performance. Examples include the risk of increased failure rate of mechanical seals in centrifugal pumps and thermal shock degradation of gas turbines. An increasing trip rate is an early warning sign of deteriorating performance. It’s an important metric to be monitored for ensuring long-term reliable operation of critical machines. If equipment stoppages are the main cost drivers for the operation, then mean time between trips (MTBT) trending is better than absolute readings.

Figure 7: Snapshot of automated monthly trips for a compressor

Figure 8: Snapshot of automated MTBT trend for a compressor

Benefits of automated real-time performance metrics

• Automated performance metrics enhance monitoring capability for early detection and mitigation of potential failure modes, allowing efforts to proactively focus on problem areas.
• Real-time performance monitoring helps in optimizing maintenance activities by reducing or eliminating unnecessary maintenance intervention for equipment with good performance history.
• Real-time reporting supports condition-based maintenance by ensuring that maintenance is carried out only when necessary and convenient.
• Analytical visualization of real-time data gives necessary information to take needed mitigation action quickly and at the right time.
• Automated reporting saves manual data collection time to do more productive work and also eliminates the chance of human error in manual data recording.

Conclusion

The nice thing about automating performance metrics is that it doesn’t require spending a lot of money to implement it. It just requires utilizing data that already exists and intellectual resources available in your organization to explore many hidden opportunities that drive productivity.

References:

Beebe, Raymond S. Predictive Maintenance of Pumps Using Condition Monitoring. Waltham: Elsevier Science, 2004.