Understanding Reliability Assurance Methd in Mature Operatins Authr: Mr Steve Turner BEng MBA Directr and Principal Cnsultant OMCS Internatinal ABSTRACT Mst rganisatins try t eliminate defects and investigate reliability incidents. Observatins acrss many industries and cmpanies lead the authr t the cnclusin, that this prcess, while very imprtant, is ne f the mst fragmented and prly executed f all. Over the past few years, the authr has researched the ptential benefits, the methds used and difficulties experienced by a variety f industries and cmpanies. Research indicates that a strng apprach t incident management carries twice the benefit f implementing a CMMS and perfrming a maintenance strategy review. Hwever, in practice, it is fund that incident management wrks far better after a maintenance strategy review has been perfrmed and the preventive maintenance prgram is well executed. Research als shws that the better incident management systems are supprted by fcused plant dwntime recrding systems significantly advanced frm what is nrmally deplyed. The cmbinatin f these fur elements is named Reliability Assurance by the authr. This paper als discusses hw any rganisatin can embark n a Reliability Assurance prgram and mve quickly t a psitin where it can better capitalise n their defect eliminatin and reliability incident management prgrams. AIM The aim f this paper is t develp an understanding f, and sme tips abut, the fllwing: The Reliability Assurance prgram, The prcess f Incident Management, and Setting up and managing strategic data cllectin systems. The paper als cvers the relatinships f RCM / PMO2000 and the deplyment f a CMMS as part f a hlistic Reliability Assurance prgram. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 1
Sme Definitins Preamble The fllwing acrnyms are used thrughut the paper. CMMS OEE RCA Cmputerised Maintenance Management System Overall Equipment Effectiveness Rt Cause Analysis RIMS Reliability Incident Management System based n RIMSys RCM Reliability Centered Maintenance accrding t SAE standard JA 1011 PM Preventive Maintenance which includes fixed time cmpnent replacement prgrams and cnditin based replacement prgrams PMO Planned Maintenance Optimisatin - In this presentatin when we refer t PMO we refer t ur wn versin which is PMO2000. This versin is nt the same as PMO prgrams develped and used in the US Nuclear pwer industry. The difference being that PMO2000 prduces the same results as RCM whereas the latter may nt. Ntes t Readers The cntent f this paper assumes a basic knwledge f the cncepts f the RCM task selectin criteria accrding t the SAE standard SAE JA1011 titled Evaluatin Criteria fr Reliability Centered Maintenance prcesses. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 2
Prgress in the wrld f Industrial Maintenance and Reliability The Past Maintenance is a practice cnducted acrss the span f the animal kingdm. As humans have built mre sphisticated and dangerus machinery, and the wrld has becme increasingly reliant n machines t deliver wealth, maintenance and reliability practices have cntinued t evlve. Changes in philsphy seem t be accelerating as is the thirst fr better methds amngst practitiners. In the industrial arena, many bservers see the fllwing as being significant changes in appraches. Early 1900 s - Traditinal methds based n verhaul. 1978 Nwlan and Heap Reprt n RCM written fr United Airlines. In this reprt, Nwlan and Heap cined the name Reliability Centered Maintenance. This was, and still is, a prcess f maintenance analysis develped fr use in the design phase f the asset life cycle. (Mubray 1997) Mid 1980 s Prmtin f RCM as a maintenance analysis tl in industries utside f airlines. In the absence f any tl better suited t develping asset maintenance strategy, RCM was used widely in plants that were already perating. Mid 1990 s Develpment and refinement f Planned Maintenance Optimisatin (PMO) techniques. Thrugh the inability f RCM t adequately cater fr rganisatins that required a maintenance strategy review and ratinalizatin prcess, the prcess f PMO became ppular and was cnsidered by their regulatrs, a majr strength in sme Nuclear Pwer Statins. (Jhnsn 1995) Recent years Further refinement f PMO methds that prvide the same maintenance prgram as RCM but significantly faster. PMO2000 being ne f thse prcesses. The Future RCM has definitely taken a grasp n the maintenance cmmunity as a tl cnsidered t be best practice. Hwever, the failure f RCM t becme cmmn place in industry and the high number f incmplete analyses is testament t the prblems with its applicatin in industry. The majr prblem with RCM applied t existing plant is that it is a very time cnsuming prcess. The rise in ppularity f PMO prcesses is because they are cnsiderably faster than RCM. PMO2000 is a prcess that retains the speed while achieving the same maintenance prgram as RCM (Turner 2001). Fr these reasns the spread f PMO2000 has been rapid. Users f PMO2000 nw d nt get tied up with analysis; they quickly mve int implementatin and begin t lk fr greater pprtunity. In the authr s pinin, the future fr asset intensive cmpanies nw lies in the integratin f the fur independent systems depicted in Figure 1. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 3
CMMS PMO/RCM Dwntime Reliability Data Incident Management Management Figure 1 - The fur quadrants f Reliability Assurance Defining Reliability Assurance Reliability Assurance is a term used by a variety f industries in many different ways. Web searches reveal uses in the fllwing applicatins: Sftware develpment, Electrnic cmpnent design and manufacturing prcess design, and Hardware designs bth in cmputer and general applicatins. All the abve applicatins seem t be launched in the design phase f the prduct life cycle. Few, if any, prcesses f Reliability Assurance have been develped fr existing plants. The authr s mdel f Reliability Assurance is designed specifically fr assets in use r fr new assets where there are similar assets perating elsewhere and / r the vendr has prvided a maintenance schedule. It is shwn in Figure 1. My definitin f Reliability Assurance is A prcess that determines the inherent reliability and perfrmance f an asset in its peratinal cntext and serves first t increase the level f perfrmance t the inherent level by the applicatin f preventive and predictive maintenance, and then, by data cllectin and prblem slving, increase the inherent perfrmance level thrugh the intrductin f mdificatins t the machine design and the perating cnditins r methds. The Fur Quadrants f Reliability Assurance Frm a prcess perspective, a Reliability Assurance (RA) prgram cntains the fur quadrants, r elements, shwn in the mdel at Figure 1 and described belw: A Maintenance Management System A maintenance management system is the cre f the maintenance administratin functin. It allws effective planning and scheduling, cst management and ther administrative and executin capabilities. Such systems are ften cmputerised and knwn as Cmputerised Maintenance Management Systems (CMMS). Almst all mdern CMMS prducts are very gd at administratin and executin f maintenance and perfrm sme Reliability Assurance functins. Hwever, in reality, these are very limited and unsuitable fr efficient deplyment f the remaining three elements f an effective and efficient Reliability Assurance prgram. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 4
Strategy Review Systems A strategy develpment and review system is a system that cntains the list f all the plant failure characteristics and the strategy develped t eliminate r best-manage the cnsequence f unexpected failure. Typical methds f strategy develpment include Reliability Centered Maintenance (RCM) and Planned Maintenance Optimisatin (PMO). Perfrmance Lss Data Cllectin System A perfrmance lss data cllectin system is a system that recrds a variety f dwntime categries, rate lss and quality lss at cmpnent and failure mde level. Such a system ften cllects data electrnically thrugh DCS r SCADA systems, but then relies n links t strategy develpment t cde and turn the data int valuable infrmatin. Reliability Incident Management system A reliability and incident management system is a system which cntrls the Rt Cause Analysis and the Plan D Check Act cycle f cntinuus imprvement and failure eliminatin. Relating Management Systems t the Reliability Assurance Mdel T apprach maintenance perfrmance imprvement, it is useful t understand the relatinships between the fur quadrants within the mdel and what returns each prvides in terms f imprved machine perfrmance. The best independent infrmatin knwn t the authr cmes frm the DuPnt mdel f Up-Time (Ledet. 1994) featured in the Manufacturing Game 1. This mdel illustrates the relatinships between the varius elements f the Reliability Assurance fur quadrant mdel. The table belw illustrates hw Ledet has mdelled the relative effect f varius strategies n plant uptime. Tactic U p t i m e % C h a n g e U p t i m e Dwntime Reductin Reactive 83.5% 0% Planning Only + 0.5 % Scheduling Only + 0.8 % Preventive / Predictive Only - 2.4 % All three tactics + 5.1 % 88.6 % 30.9% Plus Defect Eliminatin + 14.8 % 98.3% 89.7% Figure 2 - Table shwing the effect f different reliability engineering activities n plant availability taken frm the Manufacturing Game and subsequent wrk by Ledet (Ledet 1994) http://www.manufacturinggame.cm Ledet s analysis suggests that if cmpanies fcus n planning nly they will imprve their uptime by 0.5%. If they fcus nly n maintenance scheduling, uptime will imprve by 0.8%. If they fcus n preventive and predictive maintenance nly, uptime will actually get wrse by 2.4%. If rganisatins fcus n all f these three aspects, they will gain a 5.1% imprvement in availability. These results may well sund appealing in their wn right, but subsequent t the reprt, Ledet fund that by adding defect eliminatin t the initiatives undertaken, a further 9.7% (taking the ttal 1 The manufacturing game is a practical learning prcess where participants learn in an interactive envirnment, the strategies which will best enhance plant uptime. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 5
imprvement t14.8%) imprvement in availability may be achieved in their plants. This infrmatin is prvided in the table at Figure 2. Tactic U p t i m e % Change Uptime Dwntime Reductin Reactive 83.5% 0% Planning Only + 0.5 % Scheduling Only + 0.8 % CMMS Preventive / Predictive Only - 2.4 % PMO/RCM All three tactics + 5.1 % 88.6 % 30.9% Plus Defect Dwntime Incident + 14.8 % 98.3% 89.7% Eliminatin Data Management Figure 3 - Relatinship between the fur quadrants f Reliability Assurance and the prcess elements studied by Ledet The relatinships between the fur quadrants f Reliability Assurance and the prcess elements studied by Ledet are shwn in Figure 3. This relatinship suggests that imprving planning and scheduling by implementing a CMMS withut a having a fcussed PM prgram will nt generate significant returns. Similarly, wrking hard at develping a fcussed PM prgram withut a gd planning and scheduling system will nt generate significant returns either. The suggestin is that rganisatins shuld wrk n their CMMS planning and scheduling systems and their maintenance strategy develpment as well. The ther imprtant factr is that the defect eliminatin prcess is the prcess that prvides the mst imprvement pprtunity. If this is the case, then the intuitive apprach t secure imprvement wuld be t fcus n defect eliminatin first and then wrk n the ther elements. This apprach will nt wrk withut a strng fundatin f preventive maintenance. This is because withut gd preventive maintenance, reactive maintenance will prevail. In reactive mde, a high percentage f the failures will be cause by a lack f maintenance, nt inherent prblems with machinery design r perating prblems. In this situatin, any prgram t wrk n defects will, in all likelihd, be unable t determine if the failure was due t lack f maintenance r design. In additin, the vlume f defects t analyse will prbably be t high t cver, making the defect eliminatin prgram exhaustive and ineffective. The prpsed starting pint therefre must be t get the fundamentals f effective PM in place which means that the first step in launching an Reliability Assurance prgram is t review the maintenance strategy. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 6
Describing the Elements in Detail Maintenance Management Systems The functins f a Maintenance Management System include the fllwing: Administratin and Executin Wrk rders, Wrk histry, Spare parts, Cst cntrl, Cntractr management, and Planning and Scheduling. There is cnsiderable literature available that prvides infrmatin n setting up and using a maintenance management system. Because f this, the fcus f this paper is nt in this area. There are, hwever, sme factrs abut the implementatin f maintenance management systems that make effective Reliability Assurance difficult t accmplish. These factrs are discussed in the fllwing paragraphs. Perceptin Issues One f the prblems in getting a Reliability Assurance prgram established is that many peple in rganisatins believe that the CMMS can d mre than it is capable. T put things bluntly, few, if any, CMMS systems have the data fields and data relatinships necessary t prvide infrmatin in the manner required fr maintenance strategy develpment, plant perfrmance management 2, and incident management. The realisatin that all f these systems are distinctly different yet need t be integrated is very imprtant. The purchase and use f a CMMS alne will leave a large vid in Reliability Assurance infrastructure. T be effective at Reliability Assurance, an rganisatin needs systems, sftware and wrk prcesses that deal with the ther thee quadrants. Use f the system Many CMMS s are prly used and nt set up prperly. This may be fr many reasns. One f them is that in setting up a CMMS, the rganisatin has seen it as a tl t manage maintenance administratin alne. They have nt understd the cncepts f Reliability Assurance. In cmputerising their systems they have failed t realise that setting up a system t administer maintenance is nt ging t wrk well if the underlying prgrams are nt well defined r d nt add value. The startling result f mst PMO2000 analyses cmpleted ver an eight year perid is that barely fifty percent f the preventive maintenance strategies cntained in CMMS remain unchanged after review. It is cmmn t delete 15% f the preventive maintenance prgram tasks as these add 2 CMMS can be used t manage dwntime, hwever in many industries there are ther perfrmance criteria t be cnsidered. These include rate lss and quality lss. CMMS d nt manage these well if at all. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 7
n value. It is cmmn t add the same amunt t manage failures that are preventable but have n preventive maintenance. The remainder f the changes becme interval extensins r reductins r mves frm time based verhaul maintenance t cnditin based maintenance r vice versa. Trying t plan and schedule a preventive maintenance prgram that is nly 50% effective can nt be gd management. Incident Management Reliability incidents can be defined as failures f plant and equipment that lead t any kind f lss, r increased risk t the business. In capital-intensive industries, the categries vary in terms f expsure and likelihd. Typically hwever, they include the fllwing categries: Threat t safe peratin, Threat t the envirnment, Threat t the cmmercial viability f the cmpany, Lss f custmer satisfactin, Lss f prductin r failure t cmplete the missin, Breach f security, and High repair cst. The prcess f incident management is t identify and reslve plant r human failures that result in greater expsure r the lss f any f the abve. Ideally, rganisatins wuld take steps t remve such risks befre they ccur, hwever in practice, predicting every risk and reducing each f them t an acceptable level is a very difficult thing t d. Many rganisatins have fund that they can create a fcussed maintenance strategy fr all equipment (critical and nn critical) within 12 mnths by taking a review and ratinalisatin apprach. The prblem with mst maintenance strategy develpment activities is that infrmatin is never perfect and assumptins are made. This means that the maintenance strategy is a living prgram which needs incident management t imprve it as better infrmatin cmes t hand. In additin t incrrect assumptins, there are ther factrs that culd cause unexpected equipment failure. Sme f these factrs are as fllws: Temprary repairs installed and nt remved, Maintenance errr caused by pr training r lack f adherence t prcedures, Maintenance nt being dne n time, Incrrect peratin f equipment, and Installatin f damaged r faulty parts. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 8
The prcess undertaken t review reliability incidents is relatively simple and quite cmmn place. It fllws a typical investigatin cycle fund in many prblem slving techniques. At a high level, the generic prcess that we prefer t use has seven steps which are listed belw: Originate Allcate Analysis Respnsibility Analysis and Recmmendatins Apprve Implement Review, and Clse As this apprach is cmmn, it is nt cnsidered necessary t discuss each step. Hwever, it is wrthwhile expanding n ne unique aspect that pertains t Reliability Assurance. When reviewing equipment failure, there is a specific prcess flwchart that we recmmend shuld be fllwed. The prcess is shwn belw in Figure 4. The starting pint [F] is any unexpected failure that has ccurred in the plant. The first step is t define the failure mde r mechanism f failure. Fllwing this [Failure Analysed?], it needs t be determined if this failure mde has been analysed previusly using RCM / PMO lgic. If it has nt [N], then it shuld be put thrugh an RCM / PMO2000 analysis [Apply RCM / PMO2000 ]. If it has been reviewed [Y], then the validity f the previus review needs t be assessed against the fact that the failure has nw ccurred unexpectedly [Failure Prevented?]. The previus analysis may have recmmended a "N Scheduled Maintenance" plicy in which case, the utcme was expected and n further actin need be taken except if the failure has nw becme mre f a prblem than riginally thught [Increasing prblem?]. Then mdificatins and a revisin f the RCM / PMO2000 shuld be undertaken based n the decreased reliability. If, hwever, the recmmendatin was fr preventive maintenance and the preventive maintenance F F Failure Establish Failure Mde Failure been Analysed? Y N Apply PMO Failure Prevented? N Y E End E Failure Prevented? Y System Dwnfall Remedial Actin N Scheduled Maintenance Develp Task N Increasing Prblem Y N N Md / PM Wrth Ding? Design and Incrprate Md Apply PMO Schedule Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 9 Figure 4 - Defect Eliminatin flw chart Y E
has failed [System Dwnfall], then the surce f the prblem needs t be identified and rectificatin actin taken. Clearly, t undertake this wrk, the rganisatin needs t have an efficient means f retrieving the maintenance strategy fr any given failure mde. Once again, the need t cnduct either RCM r PMO2000 befre deplying an incident management system is shwn. Strategic Data Cllectin Data Types Data cllectin in the maintenance envirnment can take many frms. It is imprtant t cllect data abut plant cnditin and what maintenance wrk has been dne. Hwever, that data is nt the data mst needed fr Reliability Assurance. The data required fr Reliability Assurance is data relating t equipment failure and the circumstances surrunding that failure. There needs t be a clear distinctin between these different types f data when setting up a strategic data cllectin system. Data Elements Machines exist 24 hurs f every day they are n the cmpany register. During this time, they may be in a number f states. Sme f these states are listed belw: Having upgrades r mdificatin, Nt required fr prductin, In transit r being changed t different prducts, In prductin, In planned maintenance, and / r In breakdwn maintenance after having suffered a failure and being repaired r running at a reduced rate. Many cmpanies track these states and establish a figure that cmpares prductin time t ttal time. This figure, when de-rated with quality and thrughput lsses is ften called Asset Utilisatin r Ttal Effective Equipment Prductivity. This paper is primarily cncerned with machine reliability and is therefre cncerned nly with the latter tw pints n the abve list. It shuld be nted that this paper is restricted t analysis f evident failures 3 as hidden failures by definitin d nt f themselves cause peratinal lss. These tw reliability elements can be expanded as fllws: Planned Maintenance Preventive Maintenance, r Crrective Maintenance 3 Evident failures are failures that can be detected by site persnnel under nrmal circumstances. Hidden failures can nt be detected by site persnnel under nrmal circumstances. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 10
Breakdwn Maintenance Expected Failure - Equipment breakdwns that have been assessed as N Scheduled Maintenance, r Unexpected Failure - Equipment breakdwns that shuld have been predicted r prevented. Inherent in these elements are sme cncepts that need t be understd clearly in rder that the Reliability Assurance apprach makes sense. These cncepts are explained with the assistance f the mdels shwn in Figures 5 and 6. Inherent Capability Lss The reliability and perfrmance f any machine is determined by tw factrs. These are as fllws: The way the machine was designed, and The way it is perated. Expected Failures Failure characteristics and ecnmics are such that fr sme failures, the defined maintenance strategy is N Scheduled Maintenance (NSM) 4. This may be because f the tw scenaris described belw: the failure is randm, and the PF interval is t shrt t be f any use, r because the cst f preventin is mre than the csts f the failure. This reality means that there will be a certain level f unavailability inherent in the design and perating cnditins. Failure mdes which have NSM strategies will inevitably becme breakdwns and result in capability lss. We call such failure mdes Expected Failures. This is because ver the life f the asset, it is expected that such failures will ccur and result in lss f prductin. Expected Failure Planned Maintenance Unexpected Failure Available Time Ttal Time Inherent Capability Lss Inherent Perfrmance Level Figure 5 Mdel shwing the definitin f Inherent Capability Lss (ICL). 4 It is wrth nting that if the failure cnsequences are hazardus, a NSM strategy can nt be made. In such cases, a mdificatin is required t reduce the risk f failure t a tlerable level. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 11
Planned Maintenance While sme failures r breakdwns will be accepted as being inevitable, thers will be prevented either thrugh cnditin mnitring r fixed time cmpnent replacement. Where these preventive actins require that the plant is taken ff line, then the preventive maintenance is anther lss that is inherent. In additin, cnditin mnitring may detect the nset f failure. The rectificatin actin taken in such cases may require the plant t be taken ff line. All f these lsses cmbine t frm the Inherent Capability Lss shwn in Figure 5. The Inherent Perfrmance Level is therefre the ttal time less the Inherent Capability Lss. Maintenance Capability Lss If the PMO2000 /RCM maintenance analysis was dne crrectly, and the machine is maintained and perated accrding t the apprved prcess, then it shuld suffer n unexpected failures. This is nt t say that the plant will nt fail. What it means is that all the failures the plant experiences will be expected. The unfrtunate reality in mst rganisatins is that sme failure mdes that receive preventive maintenance will fail unexpectedly. This means that even thugh sme failure mdes that have preventive maintenance activity t prevent them frm ccurring during prductin they still happen unexpectedly. These unexpected failures are knwn as Maintenance Capability Lsses (MCL s) and are shwn in Figure 6. Data Cllectin Systems Expected Failure Planned Maintenance Unexpected Failure Available Time The steps suggested are discussed in the fllwing paragraphs. Step 1 Setting up a Generic Data Cllectin System Ttal Time Maintenance Capability Lss Inherent Perfrmance Level Figure 6 Mdel shwing the definitin f Maintenance Capability Lss (MCL). Setting up a generic data cllectin system is nt difficult in thery. In practice, data cllectin usually invlves peple in the cllectin f the data, input int cmputers, and its use. The fllwing is a list f imprtant factrs that shuld be cnsidered early in the develpment f the data cllectin system. Data cllectin systems ften have a large number f interested parties r stakehlders. Fr this reasn, data cllectin strategies shuld nt be created by a single persn with nly ne Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 12
agenda in mind. The data cllectin shuld be a result f a systematic and wide ranging needs analysis. Data is ften cllected by peple wh perate machines. The degree f literacy and numeracy f these peple shuld be assessed and cnsidered. Aviding written ntes by using cdes is a gd idea. The training and caching prvided t the peple cllecting the data shuld be dne frmally. If manual lg bks are used, the prcesses and calculatins required need t be shwn in psters near the cllectin pints. Cnsider what types f cdes are applicable as the reprting lss cdes determine the reprts that can be generated. In mst cases, smene has t enter the data int a database. It is imprtant t minimise the effrt invlved in keying in infrmatin. In sme cases it is imprtant t cllect data abut rate variance and quality lss. If these are imprtant, then there shuld be plans put in place t accunt fr these lsses. Smetimes this data can be difficult t btain accurately hwever, it is ften wrth making sme assumptins and implementing smething rather than waiting fr the perfect slutin t be fund. The system shuld be recncilable. This means that the actual utput plus the lsses shuld amunt t the standard rate multiplied by the time in prductin. Infrmatin abut plant failure is best input after the fault has been crrected. The system shuld be such that cdes are nt entered at the time the fault ccurs. They shuld be entered after the rectificatin is cmplete. Step 2 Establish Current and Inherent Perfrmance Establishing the current perfrmance f an asset can be dne when the data cllectin system is implemented. It may take sme time t generate sufficient data t understand the average perfrmance levels as there culd be quite a bit f variatin ver time. The establishment f the inherent perfrmance level assumes that RCM r PMO2000 has been undertaken. The prcess cmmnly used t determine inherent perfrmance level is t cllect all the lss data frm a recent perid lng enugh t be valid and cnsider what failure mdes are treated with preventive maintenance and which nes will be left t repair when they fail. It is nw pssible t predict what the perfrmance wuld have been under the new maintenance strategy. This requires the assumptin that the failure mdes that nw have the preventive maintenance dne at the crrect interval will have been planned maintenance activities rather than breakdwns. This methd is a gd estimate f the inherent perfrmance level f the equipment in the prevailing perating cntext. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 13
s Set up Generic Data Cllectin System Revise data gathering system Establish current and inherent perfrmance Determine defects and quantify Track success f plan and revise plan if necessary Archive data Develp and implement next imprvement plan Y Develp and implement imprvement plan Start Revise inherent perfrmance Are there any mre prblems t slve? Finish F N Figure 7 - Flw Chart fr setting up strategic data cllectin Step 3 Determine Defects and Quantify Them Frm the data gathered, r frm discussins with peple clse t the plant, determine what the main causes f dwntime are and attempt t quantify them. During this step the perceived difficulty in slving and implementing the slutins shuld be cnsidered. Frm this pint, select a small number f manageable defects t wrk n. It may be a gd idea nt t start with difficult nes first. Step 4 Develp and Implement an Imprvement Plan Cnduct wrkshps t establish causal relatinships and create an imprvement plan. It is highly recmmended that these wrkshps invlve the peple wh perate the plant and cllect the data as this will build a sense f wnership in the imprvement plan. In this step, frmal RCA wrkshps can be used. Step 5 Revise Data Gathering System Once the imprvement plan is created, the cause cdes fr imprvement need t be reviewed t ensure that the cllectin systems can be used t determine whether r nt the imprvement plan is wrking. Step 6 Track the Success f the Plan and Revise if Necessary The data cllected shuld indicate if the imprvement strategy is wrking. If the strategy is nt wrking then a new ne needs t be created r the prblem needs t be listed as inherent in the system. Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 14
Cmmn prblems with Incident Management Systems The mst cmmn prblems rganisatins have with incident management systems are as fllws: The fundatin f PMO2000 wrk has nt been dne. There is n-ne in the rganisatin that is respnsible fr administering the system. The system is fragmented and cumbersme. Reprts get lst and nbdy is keeping track f things. Getting infrmatin is time cnsuming. CMMS, OEE r PMO systems are nt integrated. T many incidents are being investigated at nce. Data cllectin strategy is t generic and lacks definitin. There is little knwledge f what prblems are being lked at and fr what reasns. The systems are t cumbersme t get data int and get data ut. T much data is being cllected. There is mre than ne system cllecting the same data. This Frustrates the peple wh cllect the data, and Leads t arguments abut the data rather than a fcus n slutins. N-ne tracks whether the imprvements wrked r nt they are nt integrated with the OEE system The peple that cllect the data are nt invlved in using it t slve prblems. This results in pr data quality. The data is cllected at the wrng level and can nt be interrgated accrding t the necessary parameters. Fr example in a manufacturing plant, data may be cllected against the line and s investigatins cmparing perfrmance f different prducts can nt be easily dne. Manager s belief that their CMMS systems can be cnfigured t perfrm Reliability Assurance. References: Jhnsn L.P (1995) "Imprving Equipment Reliability and Plant Efficiency thrugh PM Optimisatin at Kewaunee Nuclear Pwer Plant" SMRP 3 rd Annual Cnference, Chicag Illinis. Mubray J M (1997) Reliability centred Maintenance. Butterwrth - Heinemann, Oxfrd Nwlan F S and Heap H (1978) Reliability centred Maintenance. Natinal Technical Infrmatin Service, US Department f Cmmerce, Springfield, Virginia. Ledet W (1994) Ratinal Cnsideratins Systems Dynamics Mdel (The Manufacturing Game) Gal/QPC Cnference Bstn, MA, USA Nvember 1994. Turner S J (2001) PM Optimisatin Maintenance Analysis f the Future ICOMS Annual Cnference Melburne 2001 Cpyright OMCS Internatinal www.reliabilityassurance.cm Page 15