It's common for all of us who had ever collected PdM data (vibration, lube oil parameters, temperature, etc.), to follow a "trend to trend" pattern with every parameter of our machines. But the point of a successful trend analysis is the correct management of this raw data.
A bullet-proof procedure includes the following aspects:
- Repeatability: First things first. You could not compare potatoes with apples. Define specific and stationary points a data collection in the equipment, according to standards, best practices or procedures that ensure also its reliability.
- Optimal Monitoring: Standardize your monitoring procedures so it could be understand by everybody in the maintenance department. It also ensures repeatability and data quality.
- Data management: There're many tools that could help us at this point. Some data could be manage by an specific software that could only trend parameter values. But what you need to analyze is not only an overall or absolute value but their criticality (how important are to the final diagnosis), type (time-dependent or independent), and limits definition (RoCrate of change or absolute).
Remember: almost everybody could collect data... but only the "chosen ones" could correctly manage it. Rely on your common sense and experience in combination with statistical software tools assistance. Don't' be a "data collector"... be a "data manager".
Tip provided by Jorge Ortega, PdM Engineer, PETROAMAZONAS ECUADOR S.A. Quito, Ecuador
Thank you for the tip Jorge - your Reliability Elements T Shirt is on the way!
Would you like your own Reliability Elements T Shirt? Send us a Maintenance-Tip here