Advanced Asset Performance Management

March 5, 2025 lamrconsultinginc@gmail.com No Comment 8 Views Share on

In a quest to enhance asset performance and reliability, our client embarked on a journey toward advanced predictive analytics. This initiative created a comprehensive solution for monitoring and predicting compressor performance. This blog post delves into the challenges, solutions, implementation, results, lessons learned, and next steps of this project.

Challenges Faced

This client faced several challenges in managing its compressor assets, including:

• Data Silos: Data was scattered across multiple systems, making it difficult to get a holistic view of compressor performance.
• Lack of Real-Time Insights: Delayed data availability hindered timely decision-making and proactive maintenance.
• Inefficient Monitoring: Manual data analysis and limited visualization capabilities made it challenging to identify potential issues early on.
• Need for Predictive Capabilities: A reactive maintenance approach led to unplanned downtime and increased costs.

The Solution: A Comprehensive Compressor Performance Dashboard

To address these challenges, we implemented a solution centered around the AVEVA PI System, enhanced by Process Innovations’ expertise:

• Compressor Performance Dashboard: This dashboard integrates data from multiple sources, providing real-time insights and alerts. It offers a unified view of compressor health and performance.
• Customizable PI System Platform: Utilizing PI AF template plug-ins and detailed compressor maps, the platform analyzes and visualizes data, enabling in-depth performance monitoring.
• Process Plugins: Process Plugins offer specialized calculations like “Data Quality,” which checks the validity and accuracy of data; “Compressor Map,” which plots the operating point of a compressor on a map based on its specifications; “Efficiency,” which calculates the efficiency of a compressor based on its operating conditions; and “Load Step,” which measures the load step response of a compressor.
• Expertise and Support: Process Innovations provided expertise and support for dashboard implementation and maintenance, ensuring the solution met the client’s specific needs.
• Compressor Maps: Provided by “The Compressor Experts” which have extensive knowledge and experience in creating and validating compressor maps. They also provide custom compressor maps to suit different types of compressors and applications

Implementation: Integrating Data and Expertise

The implementation process involved several key steps:

1. Data Integration: Connecting various data sources to the AVEVA PI System to create a centralized data repository. A centralized PI data system is preferable to multiple disparate data sources because it simplifies data integration, quality, and security, and helps ensure data consistency and completeness
2. Dashboard Configuration: Configuring the compressor performance dashboard with relevant KPIs and visualizations.
3. PI AF Template Plug-ins: Implementing customizable PI AF template plug-ins for detailed data analysis.
4. Compressor Map Integration: Incorporating detailed compressor maps to analyze and visualize data based on compressor specifications.
5. Custom Compressor Maps: Custom compressor maps were created to suit different types of compressors and applications.
6. Process Plugin Integration: Implementing “ProcessInnovations ProcessPlugins” to ensure data quality.
7. Testing and Validation: Rigorous testing and validation to ensure data accuracy and reliability.
8. Training: Training client on how to use the new system and interpret the data.

Dashboard Features

The dashboard provides high-level metrics such as:

• Operating Status
• Series/Parallel Flow
• KPIs:
  • Isentropic & Polytropic efficiency
  • Availability
  • Health Score

The AVEVA PI Vision dashboard also allows for site-level and asset-level monitoring. The site-level overview provides facility visibility with detailed driver and compressor information. Asset-level views, available for both centrifugal and reciprocating compressors, offer detailed metrics and analysis, including:

• Displaying the current operating point of the compressor on its map
• Historical trend of the compressor’s efficiency temperature, pressure.
• Indications of Surging or Choking of Turbines
• Load Step analysis
• Pressure Differential
• Loading Efficiency
• Indication if data is within Range of ACl Services Model limits

Results: Improved Performance and Reliability

The implementation of the compressor predictive analytics solution yielded significant results:

• Enhanced Data Reporting: Improved data reporting through KPIs (Performance, Runtime, Reliability, Cost Based), monthly/quarterly detailed and high-level reports, and daily updated PowerBI dashboards for visualization of Trends.
• Improved Decision-Making: Real-time insights enabled better decision-making and faster response to potential issues.
• Increased Efficiency: Optimized compressor performance led to increased efficiency and reduced energy consumption.
• Reduced Downtime: Predictive maintenance capabilities helped minimize unplanned downtime and improve asset availability.

Lessons Learned: Key Takeaways

Valuable lessons from this project:

• Centralized Data System: A centralized PI data system simplifies data integration, improves data quality, and enhances security.
• Detailed Failure History: Maintaining a detailed history of compressor failures, causes, and solutions is essential for failure prediction and prevention.
• Machine Learning Training: Failure history provides valuable data for training machine learning models, improving the accuracy and reliability of predictive analytics.

Next Steps: Building on Success

Building on the success of this project, the following next steps are planned:

• Outage Management System: Utilizing PI information to build an Outage Management System to track compressor issues and tie them to anomaly detection. This system will automatically generate outages based on Event Frames, allow manual event creation, and use a template database for quick asset repeatability.
• Compressor Specific Predictive Analytics: Developing compressor-specific predictive analytics to identify anomalies and predict potential failures. This involves analyzing data to detect abnormal conditions, such as those caused by a severed anti-rotation pin or loss of compression due to a piston rod break.

Conclusion

By integrating data, implementing advanced analytics, and focusing on continuous improvement, we were able to significantly enhance asset performance and reliability. This project serves as a valuable example for other organizations looking to leverage data-driven insights to optimize their operations and reduce costs.