1. A Predictive Maintenance Strategy based on Real-Time Systems
SPE Paper
ADIPEC 2012
Presented by Dileep Chandran
National Drilling Company, Abu Dhabi

2. Outline Introduction: Predictive model & Advantages
Aims and objectives
Methods of analysis: Manual & Real-time measurements
Data analysis
Performance trends analysis
Optimization requirements
Proposed real-time system
Recommendations
Outline

3. Maintenance plays an important role to realize full productive capability of industry
Preventive maintenance involving inspection, servicing, repairing or replacing physical components of machineries, plant, and equipment is general practice
Predictive maintenance represents a further improvement in preventive maintenance strategies
Predictive maintenance is based on actual condition of the machine and performance trends, rather than some pre-set schedule
Introduction

4. Predictive Maintenance
Uses Equipment Specifications
Compares past and current Performance Measurements
Enables optimized scheduling of maintenance activities
Predictive Maintenance

5. Advantages Increased Equipment efficiency and reduced operation losses
Minimize probability of unexpected failures
Reduce maintenance cost and secondary damage
Extend the life of plant items
Improve product quality
Advantages

6. Develop strategies for optimization of maintenance schedules based on performance measurements
Propose improvements by using modern real-time data collection and analysis systems
Recommendations for construction of predictive models to optimize maintenance schedules using real-time systems
Aims and Objectives

7. Manual Measurements Equipment identification information
Description of the measurements
Date of the maintenance
Lower operational window limit
Upper operational window limit
Measured value
Manual Measurements

8. Method of Analysis Data Collection Manual Measurements Real-time
Trend Display & Analysis
Trend Display & Analysis
Plan Maintenance
Plan Maintenance
Method of Analysis

9. Manual Data Collection

10. Performance is represented with the use of trend lines
Data Analysis

11. Fast changing performance measurements
Data Analysis

12. Variability in performance measurements
Data Analysis

13. Performance Trend Analysis
Stable performance within the tolerance levels
Performance Trend Analysis

14. Performance Trend Analysis
Equipment exhibiting degrading performance
Performance Trend Analysis

15. Performance Trend Analysis
Equipment with faulty performance
Performance Trend Analysis

16. Post-Intervention Graphs
Performed overhaul solves performance degradation
Post-Intervention Graphs

17. Post-Intervention Graphs
Performance degradation that merits a new intervention
Post-Intervention Graphs

18. Results for Manual Data
Positive Results
Predicted degraded performance and failure points
Repeated identified interventions
An optimized maintenance strategy
Results for Manual Data

19. Results for Manual Data
Shortcomings
Measurements of low resolution and density
Time consuming treatment of the data
Fast changes in performance cannot be detected
Lack of automation
Results for Manual Data

20. Real-time Measurements

21. Results for Real-Time Data
Positive Results
Detected smaller performance changes
Allowed higher frequency of measurements
Absence of transcription errors
Results for Real-Time Data

22. Results for Real-Time Data
Shortcomings
Not widely used
Requires additional data collection techniques
Results for Real-Time Data

23. Optimization Requirements
Reduce down time and eliminate failure points
Automate the work orders, and
Automate the performance measurements and analysis
Optimization Requirements

24. Proposed Real-Time System
General Characteristics
Automated and frequent data collection
Monitoring by exception
Overall view of all operations
Integration with the current systems
Proposed Real-Time System

25. Criteria for a Real-Time System
Platform considerations
Data management considerations
Modeling tools
System functionality
Criteria for a Real-Time System

26. Components of a Real-Time System

27. Real-Time Functionality
Data Collection
Real-time
Measurements
Monitoring by Exception
Integration with existing Systems
Trend Display & Analysis
Plan Maintenance
Real-Time Functionality

28. Recomendations Increase the maintenance frequency when required
Establish post-maintenance quality control reviews
Expand the classification of normal operational windows
Classify equipment according to their criticality
Monitor critical equipment on a continuous basis
Monitor non-critical equipment on a regularly scheduled basis
Recomendations