1. STLE 2012 CBM and Reliability Engineering Conference Achieving Reliability from Data at Cerrejón A Living Reliability Centered Maintenance (LRCM) project Gerardo Vargas, Carbones del Cerrejón Ltda. Juan Carlos Consuegra, Carbones del Cerrejón Ltda. Oscar Hoyos Living Reliability (presenter) Murray Wiseman, OMDEC Inc. Dr. Daming Lin, OMDEC Inc.

2. Introduction – Main Actors Carbones del Cerrejón – World´s largest export coal mining operation OMDEC – EXAKT CBM Optimizing Software Living Reliability – Consultants in Living RCM (LRCM)

3. Commonly used terms LRCM: Living Reliability Centered Maintenance CMMS: Computerized Maintenance Management System Event type: How a failure modes life cycle ends? By: Potential Failure (PF) Functional Failure (FF) Suspension (S) PHM: Proportional Hazard Model. Extends Weibull to include CBM data. CBM: Condition Based Maintenance

4. RCM 1.What are the items functions to be conserved? (The performance requirement)requirement 2.In what ways can these functions be compromised? (The failure or failed state) 3.What causes the loss of function? (The failure mode)loss 4.What happens? (The effects) 5.How does it matter? (The consequences (H, S, O, N)? 6.What maintenance task should be done to avoid or lessen the consequences? 7.What if no appropriate maintenance task can be found?

5. Introduction - Assertions 1.Without an adequate data sample there can be no Reliability Analysis (RA) 2.Without analysis there can be no systematic verifiable improvement in reliability or in operational economy.

6. Introduction - What is a sample? Sample Right (Temporary) Suspensions: EF, ES: endings by failure, suspension B: Beginnings Life cycles: Left Suspensions: Calendar Time Work ord. 1, FF RCMREF15 Work ord. 2, FF RCMREF16 Work ord. 3, FF RCMREF16 Work ord. 4, S RCMREF15 Work ord. 5, PF RCMREF15 CMMS Work orders EF15 B15 EF16 B16 EF16 B16 ES15 B15 EF15 B15 Events table

7. Objective To describe a method wherein completed maintenance work orders capture RA enabling information

8. Agenda Introduction Objective CBM decisions The CBM model The obstacles The Living RCM solution Results Summary Questions and discussion

9. CBM Decisions Three decisions whether to: 1.Stop the equipment as soon as possible and perform a specific preventative action as indicated by the monitored data, or 2.Schedule an indicated preventative maintenance action within a specific and safe time period, or 3.Carry on with the normal operation of the equipment until the next CBM inspection and evaluation.

10. Cerrejon´s requirements for CBM CBM Method: Oil Analysis Failure mode: General Engine Wear RULE: 2090 hours StdDev: 1445 hours The three criteria 1.Optimal 2.Automated 3.Verifiable

11. CBM optimal model Predictive Model + Hazard model RULE and Confidence interval Decision based on: Probability RULE Scatter Cost model EXAKT Decision based on: Cost and Probability

12. The obstacles There are two possible reasons for the unsatisfactory performance of CBM decision model. 1.The condition monitoring variables that are available to the CBM program intrinsically bear little or no relationship to the actual failure modes that occur in the fleet. Or, 2.The data sample used to build the predictive model does not distinguish between Failure and Suspension.

13. Obstacle 1 the CBM variable have no relationship to actual failure modes Low predictability Working age t PDF f(t) Weibull Analysis FE ppm 100 Working age t PDF f(t) High predictability PHM Analysis

14. Non (low) influential indicators Obstacle 1 the CBM variable have no relationship to actual failure modes

15. Misreporting suspensions as failures (or potential failures) will weaken the model in two ways: 1.It will inflate the shape parameter causing decisions to be predominantly age based, regardless of intrinsically good (predictive) CBM condition indicators. And, 2.…. Obstacle 2 Mistaking suspensions for failures

16. 2.It will increase the scatter, and consequently confidence in prediction. This point raises a subject that RCM stresses as one of prime importance. What shall be the standard used to declare failure? Obstacle 2 Mistaking suspensions for failures

17. The Living RCM solution 1.Capturing the right information in the work orders system (CMMS) 2.Generating automatically a sample for RA 3.Motivation, leadership, and training 4.Low and high level performance metrics

18. 1. Capturing the right information in the work orders system (CMMS) Work Order RCM concepts Ellipse - Baseman Free text (updates) What I did? What I found? Event type (FF, FP, S)

19. Living Reliability Selecting the Event Type determines the entire sample point System Component Function Failure Failure mode Efects RCM as the main language of maintenance. Updates to the RCM Knowledge base 1. Capturing the right information in the work orders system (CMMS)

20. 2. Generating automatically a sample for RA

21. An LRCM project implementation succeeds based on a realization that personnel respond to the intangibles: 1.Recognition, 2.Empowerment, 3.Interest by management in their activities, and 4.Training. 3.Motivation, leadership, and training

22. 4. High and low KPI´s Performance metrics should point us precisely to what we need to improve currently in our maintenance process. That is, they should trigger a control action. Subsequently they should confirm and measure the extent to which the control action had the desired effect.

23. 4. High level KPI´s High level (lagging) KPIs : provide, at various levels of granularity, such measures as: 1.MTTF, MTTR, Availability 2.Costs, and 3.Yield

24. Low level KPI´s Low level (leading) KPIs : should measure such indicators as: 1.RCM knowledge added, The number of links between RCM knowledge and work orders, The number of RA performed CBM performance: a.Standard deviation in remaining useful life estimation b.The influence of current CBM variables as reported by the PHM shape parameter

25. It is the managers job to set low level objectives that: 1.Employees can influence by the way they perform their duties, and that 2.Support the high level organizational targets. The managers job

26. The results achieved 1.Better analysis (lower shape factor, lower standard deviation ) More confidence in making decisions 2.The maintenance personnel have now a method to register in a precise way the right information inside the W.O. system. 3.More Reliability Analysis 4.Develops, verifies, and continually improves optimal maintenance policies 5.Updates to the knowledge base

27. The results achieved Improvement in the quality of the information required for RA: Carbones del Cerrejón Fleet 789C Today 90% % of Satisfactory W.O May 40% July 70%

28. Carbones del Cerrejón Updates in the RCM knowledge base: The results achieved

29. Summary 1.Use the language of RCM to guarantee the right information from the work order system (CMMS) 2.Add the condition monitoring (CBM) data 3.Apply Reliability Analysis to generate optimal DMs. Age data (CMMS) CBM data Cost data Reliability Analysis (RA) Optimal decision models High availability Low cost

30. More information www.livingreliability.com Managing LRCM LRCM KPIs LRCM and HSE and Other related topics

31. www.livingreliability.com oscar.hoyos@livingreliability.com Thank you for your attention Do you have any questions? Questions ?