Lessons Learned in Fleet-Wide Asset Monitoring of Gas Turbines and Supporting Equipment in Power Generation Applications

Preston Johnson
Submission Type: 
Full Paper
AttachmentSizeTimestamp
phmce_14_032.pdf764.72 KBMay 27, 2014 - 7:28pm

Condition monitoring remains an important technology for equipment life cycle management. Historically, on-line condition monitoring systems are installed only on the most critical assets within a power plant, process plant, or manufacturing facility. Less critical equipment, while vital to operation of the plant, are only monitored or tested periodically using manual route based technologies. This historical practice leaves equipment specialists with a small amount of time for analysis of collected sensory data (vibration, temperature, oil, power, etc.) as they spend the vast majority of working hours collecting equipment sensory data. Fortunately, data acquisition technology has evolved, making it possible to transform standard and advanced machinery measurements from manual collections to on-line collections, increasing time for specialists to analyze, and yielding opportunities for automated diagnostics and prognostics. Taking advantage of automation, improves the ability of equipment owners and operators to lower life cycle costs and increase reliability of plant equipment.
The transition from route based measurements to a Fleet-Wide Surveillance program touches many elements from sensors to networked data acquisition nodes to servers to historians and predictive technologies. At Duke Energy, installation costs, information technology strategies, and long term vision comes together to create higher machine reliability at lower operational cost and new automation in performance monitoring, diagnostics, and advisory generation. With automation, comes increased sensory data from pumps and turbines that require new tools for data management, mining, and transformation into actionable information. The case study reviews the open and extensible data architecture of the system deployed, the ongoing efforts, and current benefits delivered to Duke Energy.

Publication Year: 
2014
Publication Volume: 
5
Publication Control Number: 
032
Page Count: 
9
Submission Keywords: 
Predictive Health Monitoring
pump
vibration monitoring
signal processing
Data Acquisition
Data Driven
Submission Topic Areas: 
Industrial applications
Submitted by: 
  
 
 
 

follow us

PHM Society on Facebook Follow PHM Society on Twitter PHM Society on LinkedIn PHM Society RSS News Feed