Navigation
Home
News
Newsletter
Knowledge Bases
Tutorials
Directory
Maintenance-Tips
Conferences
Forums
Photo Albums
Reliability Radio
Distance Learning
Benchmark
Calendar
Gadgets
Jobs
Network Links
Archives
en Español
XML/RSS Feeds Advertise

Considerations for Planning and Scheduling Part 2

Reactive Maintenance Planning

by 

Howard W Penrose, Ph.D., CMRP

President, SUCCESS by DESIGN

Introduction 

In the first paper we discussed the Workflow Concept (WFC) and Design for Maintenance (DFM) processes for improved accuracy in planning and scheduling planned maintenance tasks.  The other type of maintenance that occurs, regardless of the type of maintenance performed, is reactive maintenance as the result of random failure.  All systems have the chance of failing unexpectedly, so methods must be in place to handle these situations in order to have the least impact on the planning and scheduling process. 

Where programs are advancing through towards higher levels of maintenance, random equipment failure can be an opportunity for maintenance.  However, the reaction is often to over-react and over-provide resources to the problem at hand.  The challenge is that too few resources, or too many resources, will both have the same negative impact on solving the reactive issue.  The opportunity can be outlined in a reactive maintenance plan for specific equipment in which there is a method of fault identification, fault rectification, root-cause-analysis at whatever level is appropriate, and planned maintenance to be performed when the machine or system is idle.

In this paper, we will provide an overview of how the reactive maintenance plan can be developed.  In Part 3, we will discuss how to blend the Reactive Maintenance Plan with the Planned Maintenance program.

The Reactive Maintenance Process

Once a system or component ceases to perform the function required by the owner, the equipment is considered failed.  At this point, the random fault has occurred with an urgency based upon the criticality of the equipment.  The correct process to address the failure is as follows: 

1. Fault Identification: At this point, discovery of the fault occurs, the failure is controlled and troubleshooting is performed;

2. Fault Rectification: This is the repair or replacement of the failure;

3. Root-Cause-Analysis: Using the evidence and findings of the fault and fault rectification, an RCA should be performed.  The depth of the RCA should directly relate to the criticality.  For instance, for a minor failure that is not repetitive or does not meet a pre-set value, a simple 5-Why process may be followed.  For a critical failure, or one that exceeds a pre-set value, a more rigorous process should be followed.

4. Additional planned maintenance that can be performed on the faulted equipment should be considered a possibility.  This may include additional testing to detect latent problems.

In order to ensure that these steps are performed as effectively as possible, a written process must be developed.

Fault Identification

Troubleshooting equipment or system failure can be time consuming and dependant upon the skills and knowledge the maintenance first responders.  Knowing the available skills, as outlined in Part 1, the maintenance planner can select the appropriate first-responders.  The challenge is then left to troubleshooting and the correct selection of inspection and technology techniques.

The tools that can be used to determine the appropriate troubleshooting and inspection techniques include the results from Reliability-Centered Maintenance (the Failure Modes and Effects Analysis), a Failure Modes, Effects and Criticality Analysis (FMECA), the manufacturers’ manuals, historical data, knowledge capture and/or other processes such as Root-Cause-Analysis studies.  The results of each of these opportunities should be put in the form of a logic analysis or troubleshooting chart, as shown in Attachment 1. 

The development of such a chart involves, first, a combination of the above information as well as the instrumentation available and the abilities of the maintenance personnel.  Such a chart provides direct troubleshooting abilities as well as provides confirmation tests, inspections and pass/fail values.  The long-term benefit of such charts, in particular for critical machines, is greater control over the time, effort and selection of skills through the understanding of the length of time such tasks should take.  This information can come from the time studies performed for preventive maintenance. 

It should be noted that some failures will require efforts well beyond the ability of these charts, which would instead provide a guideline.  However, they will reduce the time to troubleshoot and bring a system back online very quickly, controlling the impact of random failures as well as providing information on the number, type and capability of personnel required. 

Fault Rectification 

The fault rectification process requires that repair specifications are developed, for outsourced repair, internal best practices/procedures for common fault repair or replacement.  The development of an overall spares identification program combined with agreements with vendors will also provide a level of stability and control over the random failure. 

Fault rectification information can be an extension to the troubleshooting charts mentioned in the previous section.  The development of the process/procedures will assist in the development of training, the selection of vendors and an early indication when outsourcing is required. 

Root-Cause-Analysis 

Root-Cause-Analysis (RCA) is a critical process for Reactive Planning.  The criticality of the random failure and how repetitive the failure is, will determine how rigorous the RCA process should be.  Attachment 2 is an example of a simple RCA 5-Why process whose information can be used to improve both planned maintenance systems as well as improving the reactive maintenance process. 

The selection of trigger points should be selected based upon the severity of the failure.  For instance, if the failure is not repetitive and the impact does not exceed $100,000 or impact delivery more than 4 hours, then a 5-Why analysis is performed.  If it exceeds those requirements, a more rigorous program requiring greater experience is justified, such as the application of the PROACT system of RCA. 

Additional Planned Maintenance 

As part of the Reactive Maintenance Plan, any additional planned maintenance should be added.  For instance, if a fan system motor fails, inspection of belts, sheaves, fan bearings, cleanliness, and other inspections can be performed.  These should be planned around the availability of the personnel assigned to the random fault as much of the time personnel are on location, they are idle.  Random faults should be considered an opportunity to inspect and improve availability of the system once it comes back online. 

Time Planning of Reactive Maintenance 

Random faults should be considered a ‘job shop’ style process and qualifies for the application of a Critical Path Method (CPM) for determining how much time is required to perform the maintenance, especially because there can be a minimum and maximum time for each sub-task.  The times for the CPM should be obtained from the time studies performed for preventive maintenance plus any historical times.  The three times selected for each branch of the CPM are the fastest, average and slowest. 

In this example, we will discuss a 500 horsepower electric motor and pump application.  The motor fails to start and trips immediately.  The trouble chart is reviewed and a technician with an MCA (Motor Circuit Analysis) device is sent out as well as a second technician to check the pump seal packing and the alignment if the motor winding is good.  The motor is checked from the starter, following appropriate safety rules, a problem is found, so the motor connection box is open, the connection split and both the cable and motor are tested.  In the meantime, the second technician is checking the packing.  The cable is found bad and new cable is obtained.  Once the material is provided, both technicians are used to install the new cable and the machine is energized.  The 5-Why analysis is performed and it is determined that a previous FMEA did not identify cable testing as a requirement on this machine.  It is determined, however, that the test is not cost effective on its own, and it is determined that MCA will be performed on a quarterly basis. 

In a PERT chart, three numbers are shown associated with each task.  These are: The minimum time, the average time and the maximum time.  The CPM is then presented as shown in Figure 1.  The advantage of this type of chart is that it can be hand-sketched if one does not already exist.

 (Click Chart for a larger version)

While this example is very simple, it does demonstrate the process.  In fact, it now gives us three numbers associated with the reactive fault: Minimum – 95 minutes; Average – 180 Minutes; and, Maximum - 290 Minutes.  We also know that the number of personnel required for this project is two. 

The numbers for the PERT chart can be obtained from time studies performed for predictive maintenance.  Additional times, such as disconnecting both sides of the cable and cable installation will require separate time studies performed in the same manner as Part 1. 

Conclusion

While not as precise as the planned maintenance portion of planning and scheduling, the reactive maintenance process can be brought under some level of control.  This is done through the development of a reactive maintenance plan which encompasses fault identification, fault rectification, root-cause-analysis and the performance of other planned maintenance practices on the faulted equipment.  Such a plan allows the planner/scheduler and management to estimate the time on task, assign the correct and right number of personnel, confirm availability of parts and determine if steps can be performed to avoid future serious or repetitive problems.  In the next paper, we will discuss how to combine the planned and reactive planning components followed by a paper on developing your maintenance budget around this process. 

About the Author

Howard W Penrose, Ph.D., CMRP, is the President of SUCCESS by DESIGN Reliability Services.  SUCCESS by DESIGN specializes in corporate maintenance program development, motor management programs and maintenance and motor diagnostics training.  For more information, or questions, see http://www.motordoc.net, contact info@motordoc.net or call 800 392-9025 (USA) or 860 577-8537 (World-Wide).

Attachment 1: Sample Troubleshooting Chart 3-Phase Induction Motor (20K PDF)

Attachment 2: 5-Why Analysis Best Practice (50K PDF)