RISK MITIGATION THROUGH STRATEGIC MAINTENANCE AND RELIABILITY

(1)

RISK MITIGATION THROUGH

STRATEGIC MAINTENANCE

AND RELIABILITY

Manufacturing plants today contain numerous equipment items that

require some form of maintenance to help deliver company output

and profit. This paper presents a practical case study detailing various

methods that can be applied to achieve a structured approach to

(2)

PAGE 2

Risk Mitigation Through Strategic Maintenance and Reliability

INTRODUCTION

Manufacturing plants today contain

numerous equipment items that

require some form of maintenance

to help deliver company output

and profit. This paper presents a

practical case study from the initial

Reliability Centered Maintenance

concept using a computerised

simulation approach, followed by

the detailed Root Cause Analysis

on some high consequence failure

modes.

The company had recognised through their internal data management processes that a particular critical equipment item was causing unexpected outages resulting in a loss of business profit. It was therefore decided by the business to embark on a continuous improvement path that would enable them to fully understand the equipment and how to prevent the unexpected losses from re-occurring. The approach taken for the critical equipment item was to initially perform a Reliability Centered Maintenance (RCM) study, in which all the likely and the dominant failure modes are identified. A Failure Mode Effect and Criticality Analysis (FMECA) model was then populated with the consequences of failure using the corporate risk matrix and financial cost of failure. The outputs of the study now gave the company the ability to optimise the choice of maintenance tasks from both a cost and safety perspective.

Once the optimum maintenance plan was developed, lifetime simulations using a Monte Carlo simulator were performed to provide a forward maintenance budget prediction and also to assess the effectiveness of the plan in reducing the impact of any critical failures. Where maintenance was not able to reduce the failure effects to acceptable levels, they were subjected to detailed root cause analysis in order for effective solutions to be identified.

In conclusion, this paper illustrates how a manufacturing plant developed a fully documented maintenance plan with associated work instructions for implementation through the site maintenance management system. The RCM study addressed a critical item of equipment within the manufacturing facility to develop a plan to ensure safe, reliable and economical use of the plant item throughout its life.

1.1 RELIABILITY CENTERED

MAINTENANCE

Reliability Centered Maintenance (RCM) is a systematic approach that focuses on preserving function rather than preserving the asset. It addresses only failures that matter using a logical process for making maintenance decisions.

During this process, the equipment is broken down to the lowest maintainable item. The failure causes for each maintainable item are then identified and analysed. If reliable failure/ maintenance history is available, then this can be used to establish the failure rate or expected life of each failure cause. However, more often than not this information is not readily available or its accuracy is not very reliable.

In such cases this information can be

estimated based on technical, operational and maintenance experience on similar machines, as well as general engineering knowledge. The maintenance strategy for each failure cause identified must also be derived. These strategies may be predictive maintenance, preventive maintenance or a run-to-fail strategy.

As part of the RCM process, the failure effect or failure consequence is assigned to each failure cause. It is important to identify the failure causes that have a potential impact on the availability of the equipment so that the equipment can be correctly modeled when the data is transferred into a Reliability Block Diagram.

(3)

PAGE 3

1.2 RELIABILITY

BLOCK DIAGRAM

Reliability Block Diagrams (RBD’s) are a tool used to carry out a system availability analysis and produce performance predictions. They are made up of series and parallel relationships that represent equipment redundancy levels. The RBD can predict downtime, number of interruptions and the Mean Time Between Failures (MTBF). The information collated during the RCM process can be utilised to create the Reliability Block Diagram (RBD) model.

1.3 ROOT CAUSE ANALYSIS

Root Cause Analysis is a problem solving methodology that is aimed at identifying the root causes of problems or events. If the root causes are identified and effective solutions implemented then it would be fair to say that we should not see the problem reoccur again in the future.

2. CASE STUDY

ZINC STRIPPING MACHINE

The equipment identified for

further improvement by the

business was a zinc stripping

machine. The stripping machine

is not something available “off

the shelf”; it was designed

specifically for the business and

is unique in its operation. The

stripping machine main function

is to strip two 20 kilogram

sheets of zinc every 4 seconds

with a daily target of 450 tonnes

a day.

2.1 THE ANALYSIS

The RCM study began by collating as much information as was available, including

equipment lists and current maintenance plans. Work order history, and performance data were analysed using Weibull Analysis to determine what failure patterns were occurring across the stripping machine. This information was then used to develop a “Desktop” Reliability Centered Maintenance (RCM) model using the simulation software package Availability Workbench™ (AWB)1_{. The asset hierarchy was built using}

AWB based on the current structure within the business SAP Computerised Maintenance Management System (CMMS). This alignment with SAP was required for future use of the AWB SAP portal and the uploading and downloading of maintenance plans.

With the asset hierarchy developed the next stage of the process was to define equipment functions, functional failures and failure modes. This was predominately completed using available information and equipment experts located within the business. The first model developed was the current maintenance strategy which used the plans that were being currently performed on the stripping machine. Once the current maintenance plans were implemented the next step was to assign the consequences of failure using the corporate risk matrix. The severity levels within the table were calibrated to the companies risk matrix. See Table 1.

With the current practice AWB model complete the business was now in a position to simulate using the Monte Carlo Simulator built within AWB to determine the impact of the maintenance tasks on the lifecycle costs and risk levels. The simulation results provided rapid analysis, and allowed for further optimisation of maintenance plans to take place. Where maintenance was not effective in either preventing or predicting a failure mode then Root Cause Analysis was performed to determine what effective solutions could be implemented to prevent reoccurrence.

(4)

PAGE 4

2.2 THE RESULTS

With the AWB desktop model now complete, we needed to perform a reality check and challenge the model results. This is carried out by first examining the Cost Criticality Pareto chart. The Cost Criticality Pareto ranks the failure modes based on total cost which includes any business costs associated with failure, labour cost, spares cost and equipment cost. See Figure 1.

ID Description Cost Per Hour OccurrenceCost Per SeveritySafety Operational Severity

S&H1 Report Only No treatment 0 0 24 0

S&H2 First Aid Treatment 0 0 730 0

S&H3 Medical treatment or lost time injury 0 0 8760 0

S&H4 Extensive or Permanent Part Disability 0 0 87600 0

S&H5 Fatality (s) or permanent serious disability 0 0 876000 0

ENV1 No impact or measurable impairment 0 2500 24 0

ENV2 Minor effects on biological or physical environment 0 25000 730 0

ENV3 Measurable impairment on biological or physical environment 0 250000 8760 0

ENV4 Serious environmental effects with impairment of eco system 0 2500000 87600 0

ENV5 Very serious environmental effects 0 5000000 876000 0

LEG1 Unlikely to result in adverse regulatory response 0 0 0 24

LEG2 Minor non compliances of breaches or regulations 0 0 0 730

LEG3 Serious breach of regulation 0 0 0 8760

LEG4 Major breach of regulation 0 0 0 87600

LEG5 May be considered wilful or negligent. 0 0 0 876000

QS1 Can easily be absorbed through normal activity 0 0 0 24

QS2 Consequences can be absorbed. Minor delivery delays 0 0 0 730

QS3 Significant event which can be managed under special

circumstances 0 0 0 8760

QS4 Major event with prioritised and focused management,

Customers lost 0 0 0 87600

QS5 Extreme event with failure to lead to failure of business. Key

customers lost 0 0 0 876000

STK1 Little or no stakeholder interest 0 0 0 24

STK2 Minor adverse local media attention 0 0 0 730

STK3 Attention from local media or heightened concern from local

community 0 0 0 8760

STK4 Significant adverse national media/public attention. Share

price may be affected 0 0 0 87600

STK5 Serious public outcry. International coverage. Share price

seriously affected 0 0 0 876000

DT1 Downtime Cost Smelter 19000 0 0 0

Table 1 Consequence Table

Prill Build Up Bench Chain Worn Busbars Dirty Prox Misaligned Broken Switches Brushes Worn Hangers Damaged Air Gap Damaged Flag damage Light Curtain Trip

$0 2m 4m 6m 8m 10m 12m

Total Cost Contribution

Failur

e Mode

(5)

PAGE 5

Emergency Stop Circuit Failure Emergency Stop Circuit Failure Emergency Stop Circuit Failure Emergency Stop Circuit Failure Emergency Stop Circuit Failure Emergency Stop Circuit Failure Emergency Stop Circuit Failure Edge Growth Not Removed Emergency Stop Panel Failure Emergency Stop Load Failure Emergency Stop Unload Failure Emergency Stop Switch Failure

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Safety Criticality - Target = ‹1

Figure 3 Safety Criticality Current Practice

Failur

e Mode

Emergency Stop Circuit Failure Emergency Stop Circuit Failure Emergency Stop Circuit Failure Emergency Stop Circuit Failure Emergency Stop Circuit Failure Emergency Stop Circuit Failure Emergency Stop Circuit Failure Edge Growth Not Removed Emergency Stop Panel Failure Emergency Stop Load Failure Emergency Stop Unload Failure Emergency Stop Switch Failure

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Safety Citicality - Target = ‹1

Failur

e Mode

Figure 4 Safety Criticality Optimised

Using the Cost Criticality Pareto Chart, the high cost failure modes were examined in detail by carrying out a validation of the failure mode information. Once validation was complete, the Cost Criticality Pareto identified areas for improvement. The type of recommendations made for improvement included root cause analysis (RCA), equipment redesign and maintenance strategy

optimisation. The second method used to challenge the model results was to examine the historical maintenance costs and compare to the budget prediction using AWB.

The goal of this was not primarily to ensure the budget prediction matched the historical spend but rather to clearly understand the reasons for any discrepancies in the results. In this case study, the predicted results were within 4% of actual maintenance spending. See Figure 2.

Figure 2 Actual Cost v Predicted Cost

Cos t $700k $600k $500k $400k $300k $200k $100k 0 SAP Actual

Year 1 SAP Actual Year 2 RCMCost Current OptimisedRCMCost

Spares $361,520 $404,287 $354,028 $353,121 Resources $214,507 $236,551 $241,344 $238,344

If any failure modes exceeded this target then a safety risk exposure existed. With the stripping machine, four failure modes exceeded the target and showed safety risk exposure. These failure modes were associated to emergency stop switches and the testing regime for those switches.

The comparison between current maintenance and optimised maintenance can be seen with a reduction in safety risk to below the target of 1 following maintenance optimisation. See Figure 3 and Figure 4.

Now failure modes had been addressed from a cost perspective, the next step was to determine if the safety risk associated to the stripping machine was less than the acceptable business exposure threshold. This threshold was calibrated with the business risk matrix to align with a target of 1.

(6)

PAGE 6

With the model details validated, it was then important to recognise the comparison between the alternative strategies on offer to the business. In this case, run to failure, current practice, optimised and optimised with redesign strategies were compared. See Figure 5. Once optimisation was finalised a maintenance budget prediction was developed to determine spares and resource requirements. Using AWB, the budget prediction could be interrogated at many levels including system, sub-system, individual asset and individual failure mode level. See Figure 6. We were now in a position to determine what impact the optimised maintenance strategy was having on the stripping machine overall availability. This would indicate single points of failure and bottlenecks within the machine and identify further opportunity for improvement. See Figure 7

Figure 5 Average Annual Cost Strategy Comparisons - Stripping Machine

Cos t $10m $8m $6m $4m $2m 0

Run to Fail Current Optimised Opt Redesign

Effects $22,056,284 $3,289,972 $2,334,326 $759,258

Resources $45,107 $241,344 $238,661 $238,344

Spares $320,654 $354,028 $353,121 $353,121

Figure 6 Average Annual Budget Predictions - Stripping Machine

$700k $600k $500k $400k $300k $200k $100k 0 1 2 3 4 5 6 7 8 9 10 Spares Labour

Figure 7 Predicted Availability Performances - Stripping Machine

99.0 97.0 95.0 93.0 91.0 89.0 87.0 85.0 1 2 3 4 5 6 7 8 9 10 Av ailabilty % Year Predicted Availability: 91.6% Total Downtime: 8.4% Outage due to Planned Maintenance: 4.8% Outage due to Unplanned

(7)

PAGE 7

2.3 THE IMPROVEMENT

WORKSHOPS

With a new maintenance strategy developed for the stripping machine, it was important to recognise that two failure modes that were occurring on the machine could not be prevented or predicted with any form of regular maintenance. See Figure 1 Cost Criticality Pareto. It was therefore decided to conduct a detailed RCA (Root Cause Analysis) workshop on these two failures.

The approach was taken to create a common reality and build a cause and effect chart that would address the conditions and the actions surrounding both problems using the Apollo Root Cause Analysis™ Method2_.

The goal of the Apollo Root Cause Analysis™ Method is to identify not only the cause and effects but identify the effective solutions that will prevent reoccurrence of the problem. The methodology is based on the four elements of cause and effect principle:

• Cause and effect are the same thing • Causes and effects are part of an infinite

continuum

• Every effect has at least two causes in the form of actions and conditions

• An effect exists only if its causes exist at the same point in time and space.

The ‘light curtain trip’ identified 63 cause and effects and 12 solutions.

The ‘flag damage’ identified 22 cause and effects and 3 solutions.

With potential solutions identified a cost benefit analysis was carried out to ensure payback on the cost of solutions.

2.4 FMECA SUMMARY

The Stripping machine FMECA analysed a total of 1798 failure modes (see Table 4). The new optimised maintenance plan produced 51 Maintenance Schedules with a total of 1466 maintenance tasks. The maintenance strategy was categorised into 10% Preventive Work and 90% Inspection Work.

2.5 CONTINUOUS

IMPROVEMENT

With the new maintenance strategies uploaded into the CMMS via the AWB SAP portal, it was recognised that a periodic review would be performed to validate the strategy and update according to new information and data becoming available. This will form part of the business continuous improvement process.

EQUIPMENT DESCRIPTION COST PER YEAR FREQUENCY

Monorail Light curtains

trip $1,107,070 Once/shift for 5 min

Side Shifter Flag damage $493,765 Once/week

for 30 min

Table 2 Failure Modes for RCA

OPTIMIZED FMECA SUMMARY Action Number of Tasks

Use As Is 929 Deleted 0 Extend Interval 54 New Tasks 471 Reduce Interval 0 Run to Failure 344

Table 4 FMECA Summary

PROBLEM COST/YEAR SOLUTIONCOST OF PAYBACK DAYS

Light Curtain

Trip $1,107,070 $34,800 11.47

Flag Damage $493,765 $2,880 2.17

(8)

PAGE 8

The paper has presented a Reliability Centered Maintenance approach to evaluating current maintenance regimes. Once complete, the models can be used to predict budgets, maintenance labour and resource requirements and spares usage. The approach also incorporated Root Cause Analysis on high cost failure modes that could not be prevented through the maintenance regime.

The paper shows through the breadth of results available, the benefit in applying an RCM and RCA approach to plant strategy optimisation. The outcomes can help ensure that the throughput goals are achieved at least cost and minimum risk exposure to the business.

About ARMS Reliability

Since 1995, ARMS Reliability has been at the forefront of proactive asset

management strategies for a range of blue chip companies throughout the world. These companies have entrusted us with delivering business goals through effective asset management and improvements in operational productivity.

ARMS Reliability is a service, software, and training organisation providing a “one stop shop” for Reliability Engineering, RAMS, and Maintenance Optimisation for both new and existing projects.

If you would like to discuss your situation and what are the essential elements to achieving best practice Asset

Management, then give ARMS Reliability a call or

Apollo Root Cause Analysis™

ARMS Reliability are global partners and providers of the Apollo Root Cause Analysis™ methodology. Please visit our website apollorootcause.com to or to view our

3. CONCLUSION

Enquire now

Learn more

worldwide training schedule

www.armsreliability.com [email protected] NORTH / CENTRAL / SOUTH AMERICA

P: +1 512 795 5291 AUSTRALIA / ASIA / NEW ZEALANDP: +61 3 5255 5357 EUROPEP: +44 122 445 9362 SOUTH AFRICAP: +27 10 500 8232