*** Application of Reliability Centred Maintenance to Dynamic Plant Introduction

*** operate T&D, Power, E&P and LNG assets and have requested advice on the application of RCM to determine their maintenance needs.

Context
Reliability Centred maintenance has evolved from work carried out in the civil aviation industry from the 60s onwards through their maintenance steering group (MSG) and has been set out in both military and commercial standards (SAE JA1011) so that the fundamental principles are now well understood. In 1978 Nowlan and Heap of United Airlines produced a paper called Reliability Centred Maintenance for the US Department of Defense that became the basis for the US Air Transport Associations standard MSG-3, currently the accepted RCM standard. RCM is a process to determine the maintenance tasks and frequencies for a given item of plant in it operating context. The approach is rigorous, transparent and auditable. It forces a structured review of the consequences of failure and focuses attention on the maintenance activities which have the most effect on safety/environmental integrity and on equipment performance. Properly applied, RCM ensures that money spent on maintenance is spent where it will do the most good. To do this we need to understand the plant, the influences on its reliability, the performance criteria that the plant was designed to fulfil (what it can do) and our desired performance standard (what we want it to do). If we want the plant to do something that is beyond what it is capable of doing then we are not dealing with reliability but with design issues. When RCM is applied it determines the maintenance tasks and the natural task intervals; a natural task interval is the ideal interval at which the maintenance should be carried out. In practice, however, it will be necessary to rationalise the natural task intervals to create rational task intervals for each task the rationalisation process ensures that the eventual maintenance schedule is both workable and cost effective. For example, if we must go to some significant effort to perform a task it may be prudent to do others at the same time even though they may not be yet due. RCM is a micro technique that looks at the maintenance needs of individual plant items; there are alternatives, such as Kaizen, Total Productive Maintenance (similar in nature to TQM) and others but these operate at a macro level and seek to make continuous improvements through operator and maintainer feedback and benchmarking from other plants etc. By working at a micro (failure mode) level, RCM not only focuses on determining the maintenance requirements for plant and equipment (ie the technical side of maintenance) but is a culture changing process (ie it addresses the human side of maintenance). RCM is applied by those people who know the equipment best (those that have day-to-day contact with the equipment) which improves the general level of understanding about the equipment but also secures involvement and commitment. If it established as a living program RCM is an excellent vehicle for continuous improvement.

RCM
The SAE standard JA1011 sets out criteria for determining whether a maintenance analysis method complies with the accepted understanding of what is RCM by testing whether it satisfies 7 questions:
1. 2. 3. 4. 5. 6. 7. What are the functions and associated desired standards of performance of the asset in its present operating context? In what ways can it fail to fulfil its functions? What causes each functional failure? What happens when each failure occurs? In what way does each failure matter? What should be done to predict of prevent each failure? What should be done if a suitable proactive task cannot be found?

By following these seven questions, RCM looks at the context in which the plant operates. It identifies and then classifies each potential failure mode according to the consequences of failure. Once each failure mode has been categorised according to the consequences of failure, RCM uses a logical, structured algorithm to assess whether each failure can or cannot be predicted or prevented. By recognising that some maintenance tasks may not be justifiable (or may even be counter productive), RCM also optimises the costs by ensuring that neither too much nor too little maintenance is specified. The approach provides an audit trail and is often used to support a safety case. RCM is applied by the people who know the equipment best. Typically they are those who have day-to-day contact with the equipment (ie the maintenance craftsmen and supervisors, equipment users/operators, specialists and possibly the equipment suppliers). By drawing on knowledge and experience, RCM2 addresses the human side of maintenance - improving the general level of understanding about the equipment and securing involvement and commitment. In this way RCM results will endure. The RCM approach employed by Ltd is RCM2; this form of RCM was developed by John Moubray whose company Aladon Ltd provides RCM training through its worldwide network of licence holders (currently covering 44 countries and 10 languages). RCM2 is the most widely applied form of RCM in the world and is fully compliant with SAE JA1011 (most other forms of RCM are not, including MSG-3). John Moubray is recognised as leading the development of RCM and he is involved in the evolution of MSG-3 to remove some of its recognised flaws.

Streamlining
The application of RCM is time consuming (if applied properly) but the investment is more than rewarded by the benefits that it brings (again, if applied properly). As a result, there have been many attempts to streamline both the application of RCM (by reduce the overhead of carrying out the analysis) and the process itself (by developing less rigorous derivatives). Some of these streamlined approaches are described in the following sub-sections. None of the streamlined RCM processes complies fully with the requirements of the SAE standard and in most cases they are streamlined because something is being omitted. Leaving things out inevitably increases risk, so from this viewpoint and from the lack of defensibility of output, adopting a streamlined RCM approach could leave an organisation (and its responsible managers) vulnerable in court should an incident occur. The regulators have also woken up to the existence of streamlining and are aware of the risks involved. For example, an independent report was commissioned in the mid 90s by the HSE in the UK to investigate the application (and misapplication) of RCM processes as there was some concern over the way in which RCM had been applied by some companies in the offshore sector.

Criticality Analysis A shortcut used in the application RCM includes the approach of analysing critical functions only, or to subject only critical failure modes to detailed analysis; this approach is seriously flawed. The process of dismissing functions and/or failure modes as being non-critical entails making assumptions about what a more detailed analysis might reveal; such assumptions are frequently wrong since apparently innocuous functions or failure modes are found on closer examination to embody elements that are highly critical in terms of safety and/or environmental integrity. For example, an innocuous looking functional failure may be caused by a failure mode that has secondary damage which results in turn in serious safety/environmental or operational consequences. The practice of prematurely dismissing functions or failure modes results in much riskier analyses, but because the analysis is incomplete, no-one knows where or what these risks are. Another way of shortcutting the RCM process is to carry out a criticality analysis; that is, applying RCM only to plant and equipment considered to be of more importance than the rest. This implies that there will be a partial adoption of RCM, which in turn suggests a two-tier maintenance management system and the plant not considered by RCM analysis will still require maintenance, but to what standard? There may be some debate over what is considered critical and non-critical plant and there may be no correlation between the importance of a section of plant and the effort that will be saved by not analysing its maintenance needs. When considering a complex installation there may be many differing techniques employed to determine criticality so the first task is to decide which of these to employ. One may consider how deeply safety is affected by failure of the plant, how production is affected, how much time or money is spent maintaining it and so on. As a result, criticality is likely to be viewed differently as each of these criteria is applied. Whichever criterion is used, there will be a significant effort to analyse and document criticality, effort that may be better spent on actual maintenance analysis. Ultimately there is no guarantee that the time spent assessing criticality will be paid back in reduced RCM effort. It is valid, however, to apply some form of criticality analysis in order to determine where to start (rather than as a means of deciding where not to apply RCM); such prioritisation should also include economic factors rather than just the notion of criticality. Generic Analysis A fairly widely-used shortcut in the application of RCM entails pasting (often blindly) an analysis performed on one system to another identical or similar system. In fact, some organisations even sell such generic analyses, on the grounds that it is cheaper to buy an analysis that has already been performed by someone else than it is to perform your own. Great caution is needed when considering the adoption of complete packs of analysis data generated outside the organisation. This is because it is unlikely that the imported analysis has been carried out on technically similar equipment that is operating under the same conditions and that is configured in the same way as the plant to which it is going to be applied. For example a generic analysis of the maintenance needs of a pump may seem attractive to an organisation that has a lot of pumps on the asset register. For the resulting pump maintenance to be meaningful, however, the imported analysis must be of a pump which undertakes the same duty, in a similar configuration, is carrying the same fluids under the same operating conditions (pressure, temperature, flow rate etc) and the consequences of failure must also be the same. If this is not the case then the effort required to adjust the generic model to each additional pump may be close to that of carrying out individual analyses on each pump. Furthermore, a generic analysis must also provide a maintenance strategy that is consistent with your organisation (for example it may assume that maintenance resources are available close at hand when in fact this is not the case).

It is reasonable, however, to use an existing analysis as a template (ie starting point) for another analysis but only if due care is taken. On complex plant the savings can be significant but it is likely that this generic model will only be available in-house. Reverse Engineering of RCM One of the most popular methods of streamlining the RCM process starts with the existing maintenance tasks and attempts to identify the failure mode that each task is supposed to be preventing. Subsequently the failure mode is worked forwards again through the last three steps of the RCM decision process to re-examine the consequences of each failure and (hopefully) to identify a more cost-effective failure management policy. This approach is also known as backfit RCM or RCM in reverse. This form of streamlining is amongst the most dangerous of the streamlined methodologies, for the following reasons:

The approach assumes that existing maintenance programs cover all the failure modes that are reasonably likely to affect the assets under consideration. This assumption is simply not valid It is a common misconception that whether or not a failure is hidden or evident refers primarily to the failure mode; in fact the correct question to ask is whether or not the the loss of function caused by the failure mode will become evident to the operating crew under normal circumstances. This question can only be answered by establishing what function is actually lost when the failure occurs. If a failure is not correctly identified as hidden or evident (which is common in the retrofit approaches) the resulting maintenance can be fatally flawed (possibly literally so) Retrofit approaches are particularly weak on specifying appropriate maintenance for protective devices since many existing maintenance programs do not ensure that protective devices receive the appropriate attention (if at all). If a retrofit approach is used there is a very significant risk that many protective devices will continue to be maintained inappropriately (if at all) - this weakness of retroactive RCM alone makes it completely indefensible Retrofit RCM focuses on maintenance workload reduction rather than plant performance improvement (which is the primary goal of function-oriented RCM). Since the returns generated by using RCM purely as a tool to reduce maintenance costs are usually (much) lower than the returns generated by using it to improve reliability, the use of the retroactive approach becomes self defeating on economic grounds.

Streamlining the RCM approach Another common way in which the RCM process is streamlined is by skipping various elements of the process altogether. The step most often omitted is the definition of functions. Such an approach starts by listing the failure modes that might affect each asset, rather than by defining the functions of the asset under consideration. This particular form of streamlining has two main shortcomings:

As described above, it is a common misconception that whether or not a failure is hidden or evident refers primarily to the failure mode; in fact the correct question to ask is whether or not the the loss of function caused by the failure mode will become evident to the operating crew under normal circumstances. This question can only be answered by establishing what function is actually lost when the failure occurs. If a failure is not correctly identified as hidden or evident (which is common in the retrofit approaches) the resulting maintenance can be fatally flawed (possibly literally so)

From a technical point of view, the identification of functions and associated desired of performance also makes it easier to identify the failure modes where the asset is simply incapable of doing what the user wants it to do, and therefore fails too soon or too often. For this reason, eliminating the function definition step reduces the power of the process.

Purpose
General Organisations apply RCM for many reasons and the approach may well be worth considering if one or more of the following apply:

There is a high dependency on plant for output and/or quality Maintenance costs are spiralling and/or there is severe disruption when failures occur The organisation has high spares inventories and/or a shortage of skilled labour There are stringent public and legislative demands concerning safety and environmental integrity.

Increasing levels of automation and high throughputs mean that organisations are susceptible to quite small failures even where equipment has been designed with high levels of redundancy; these can have a profound effect on output, customer service or safety/environmental integrity. In many industries, therefore, both profitability and being a good employer/neighbour is increasingly becoming a maintenance issue. Over the years companies have sought to achieve higher output whilst simultaneously cutting costs; maintenance has been a frequent target for cost cutting. There is evidence that companies and the regulators (in some industrial sectors) are waking up to the fact that many of the cuts have been too deep as there are well documented cases of serious downstream consequences (both in terms of falling equipment output and rising numbers of breaches in safety or environmental integrity). Whilst some companies still apply RCM with the intention being to cut maintenance costs still further (in order to improve productivity), others are realising that productivity can often be more readily improved by focusing on increasing output. One solution to increasing output has been to focus on doing the maintenance work correctly (ie, the technicians are doing the job right); in practice, serious improvements in output are rarely achieved unless steps are taken to ensure that the maintenance tasks that are being carried out are those that should be carried out (ie, the technicians are doing the right job). This is where RCM can lead to substantial improvements, as it is an approach that focuses specifically on determining the correct/optimum maintenance tasks for the plant and equipment in its current operating context (ie, it determines the right job). Applied correctly, RCM forces a structured review of the consequences of failure; as a result it focuses attention on the maintenance activities which have most effect on safety/environmental integrity and equipment performance. This in turn means that money spent on maintenance is spent where it will do the most good. RCM is a zero-based approach in that each analysis derives the final maintenance tasks from scratch (without assuming that the current maintenance is either correct or incorrect). This means that tradition and practice are bypassed and excess baggage in the current maintenance regime is avoided. As the name implies, RCMs focus is on reliability and maintenance can only be selected if it is both technically feasible to do and worth doing. From a technical point of view the approach helps to eliminate unnecessary and counterproductive maintenance and from a human point of view it can significantly shorten maintenance fault-finding and improve ownership of the maintenance tasks.

In the early days of the application of RCM in the process industries, the emphasis was on cost reduction (rather than performance improvement). There were successes in this respect but not great as many were led to believe (this is discussed later) In summary, organisations apply RCM for many reasons since it is an approach that can help an organisation to increase output and contain maintenance costs without compromising plant safety or environmental integrity. *** In the context of the ***, the purpose for applying RCM may well differ from one area of the business to another; at this stage it is unlikely that RCM would be sold to all areas of the world-wide business on equal terms since the rationale for applying the approach may differ considerably according to local circumstances. For example, for Transport and Distribution the purpose of applying RCM may well be to contain or reduce maintenance costs (assuming that the design of the gas network/system is robust) whereas the focus may be on maximising equipment reliability if the integrity of the gas network is not particularly robust. For Power applications the focus of an RCM application will almost certainly be on achieving high levels of reliability and availability as it is likely that power supply contracts are heavily performance biased. The focus of the RCM initiative may, however, change depending on whether or not *** is the owner/maintainer or just the maintainer of the equipment; if *** is just the maintainer on behalf of a separate owner the focus may well have to include careful consideration of life-cycle cost measured over the contract period. Exploration & Production is typically characterised by large operating costs and a robust design - in which case the focus of RCM may well be on reducing direct maintenance costs whilst maintaining safety/environmental integrity. If, however, the equipment is operated by relatively cheap local labour the emphasis may change from cost reduction to ensuring compliance with local regulations in addition to maintaining safety/environmental integrity. The focus for RCM within the LNG side of the business would depend on whether the site was storage, liquefaction or gasification. In all instances safety/environmental integrity will probably be the main reason for adopting RCM but equipment performance may also be very important considerations.

Quality (maybe better titled Benefits of RCM?)

As stated above, organisations apply RCM for many reasons but these reasons tend to focus on one or more of the following: increasing output; contain/reducing maintenance cost; maintaining/improving plant safety or environmental integrity. Each of these is discussed in the following sections. Using RCM to improve plant reliability The Aladon Network of companies reports that the process/petrochemical sector continues to be a major user of RCM worldwide and that there is some evidence that the emphasis is shifting from cost reduction to improving reliability within the sector. The companies in the Aladon Network are bound by confidentiality agreements not to discuss specific organisations or applications of RCM but the following are some general observations. Cost reduction (discussed later) is a recurrent theme within the sector but amongst some organisations there is evidence that companies are suffering from the effects of cuts that have gone too deep. This sector is characterised by robust design (compared with other sectors such as, for example, the pharmaceutical industry) with much built-in redundancy in terms of stand-by plant. Some companies have taken this built-in redundancy as an opportunity to reduce maintenance by relying heavily on standby equipment this over-reliance is, in some

cases, a direct result of companies applying some form of streamlined RCM approach where the temptation was to go for the quick and easy wins at the time (ie reducing routine maintenance altogether) with little regard to the future problems that equipment neglect can bring. A number of companies in the process sector started out with the RCM2 approach promoted by the Aladon Network but were lured away by competition offering streamline approaches. Over the past 2-4 years a significant number of these companies have returned as clients to the Aladon Network because of their increasing concerns that production processes are becoming susceptible to poor reliability as a direct result of excessive maintenance cost cutting. In some instances companies have suffered significant losses in production throughput whilst others have been saved by the robustness of the original design and only suffered a near miss. A further problem that this sector has experienced was a lack of appreciation (at the time) that the abandonment of all maintenance to equipment would eventually lead to failures that could have significant corrective maintenance costs. Once again, the temptation was to opt for doing no maintenance rather than to consider more carefully the option of carrying out condition-based maintenance. Whilst this may be a function of the fact that condition-based maintenance techniques were not particularly well established when the RCM process was applied initially, it is also a sad reflection that RCM has not become a living program within most organisations and that they are having to try to re-start RCM initiatives that have, in many cases, fallen by the wayside. Selling RCM within this industry on the grounds of improved output or reliability is notoriously difficult (it is much easier to sell cost reduction). This is because there is often sufficient spare stand-by plant available (ie there is little or no measurable increase in production throughput) and the reliability of individual items of plant is rarely monitored (so any benefits in terms of improved equipment MTBFs remain hidden). Where companies have suffered production problems because of equipment neglect they can be receptive to (re-) consider RCM (applied properly) as the potential for lost output will usually more than justify some significant expenditure (both in terms of money and time) in its application. It is interesting to note that the FMCG sector (where margins are notoriously small) continues to be a significant user of RCM worldwide. Like the process sector, there is huge pressure to increase productivity but unlike the process sector, FMGC companies are often very susceptible to equipment failure as their designs are comparatively less robust. In the FMGC sector it is relatively easy to measure improvements in production throughput and so they buy RCM because the resulting increased output is visible; in general, the FMCG companies do not buy RCM as a means to cut maintenance costs. Using RCM to contain/reduce maintenance costs Traditionally the process sector has applied RCM to reduce maintenance costs and this remains the easiest way to sell the approach to the industry. It is easy to explain how RCM can remove unnecessary and counter-productive tasks and if applied properly, it will bypass tradition and practice and any excess baggage in the pre-RCM schedules. For many years the process sector has relied upon manufacturers recommendations as the basis for the routine maintenance carried out. Whilst this is less prevalent than it once was, the easiest way for a turnkey equipment design-and-build company to provide the maintenance schedules for the project is to source them straight from the equipment manufacturer/supplier. Where an organisations maintenance schedule is based upon manufacturers recommendations, it is not uncommon for RCM to reduce the maintenance workload by some 20-70% with little difficulty. Unfortunately this reduction is not reflected directly in a proportional reduction in headcount because it is unlikely that the maintenance actually recommended by the manufacturers is actually being carried out by the maintainers. This has lead to some disappointment in organisations as they were oversold RCM on the basis of a

20-70% reduction in workload. These apparent 20-70% savings have also lead to problems as some companies reduced manpower levels by these percentages after applying RCM only to suffer the short/medium term consequences of cutting maintenance costs too deeply. When RCM was applied in the E&P Morecambe Field in the early/mid 1990s the approach was adopted primarily to reduce the maintenance workload so that an unacceptably large maintenance backlog could be eliminated. The application of RCM was specifically targeted on systems that had high maintenance workloads and was successful in reducing the workloads to more acceptable levels. Whilst the maintenance workforce was not reduced, the backlog was brought under control and the scope of the assets maintained was increased when the North Morecambe field came on line. In other sectors (particularly transport), RCM is applied in order to gain better control over asset life-cycle costs. The railway industry is a sector where there are high equipment populations and which has followed manufacturers maintenance recommendations for many years. There has been a significant change in the UK railway industry (and more recently in other countries) in that the train builders are increasingly accepting long term (up to 25 year) contracts to maintain the trains on behalf of the train user in accordance with contracts that offer significant merit rewards if performance is met and swingeing penalties if performance is lacking. This sector is also closely regulated and so there is limited scope for adopting a cavalier attitude to maintenance. The major train builders have adopted RCM for three reasons: the process provides a complete audit trail which supports the vehicles safety case; performance improvements have been demonstrated which help to secure performance merit payments under the maintenance contract; the approach gives them reduced life-cycle costs measured over the contract. The experience of the railway industry is important in that they are interested in maintenance costs measured over the complete life-cycle whereas the process industry has been characterised as being primarily interested in short-term savings. The need to consider the life-cycle costs of the rail vehicle is accentuated by the number of multiples that the railway industry has if the total costs of the maintenance over the whole life-cycle are not taken into account, there is a significant risk that the maintainer will be made bankrupt in the long term. To date, none of the major train builders are using the streamlined versions of RCM but are using the rigorous forms of RCM (primarily RCM2 with the Aladon Network). As the Aladon Network has reinforced many times, RCM stands for Reliability-centred Maintenance and not Cost-centred Maintenance. If it is applied properly, RCM will usually save cost where current maintenance schedules are based on manufactures recommendations but it does not always result in maintenance cost savings - it can lead to more maintenance being required. Unfortunately many of the streamlined approaches have focused only on cost reduction in the process sector and delivered apparently impressive savings. The actual savings achieved are rarely as large as they appear (for the reason described above) and there may be safety/environmental dangers if a streamlined process is adopted (discussed elsewhere). If RCM is adopted primarily for cost reasons it is strongly recommended that these costs are analysed over the life-cycle of the asset rather than for quick/short term cost reduction. Applied properly, RCM will take the life-cycle viewpoint whereas most of the streamlined approaches focus entirely on short term savings with scant regard for the medium/long term. In summary, an organisation embarking on an RCM program must be aware that cost savings are only likely where an existing maintenance schedule is based heavily on manufacturers recommendations. Even in these cases cost savings are not likely to be as large as they first appear and are not guaranteed; furthermore, any maintenance cost savings should be considered over the life-cycle of the asset and not just the short-term. Using RCM to maintain/improve safety/environmental integrity The process industry is becoming increasingly regulated and monitored to ensure safety/environmental compliance. It is the experience of the Aladon Network that this remains

a significant reason why organisations chose to adopt RCM either because they have experienced a safety or environmental incident or are worried that they might be vulnerable to such an event. Increasingly RCM is becoming a stated requirement of some sectors of the process industry (the Norwegian oil sector, for instance) and this is forcing some companies to adopt the approach. As discussed elsewhere the regulators are becoming increasingly aware of the shortcomings of some applications of RCM in the industry and so it is advisable that the form of RCM adopted by an organisation is well recognised, applied and implemented correctly and (preferably) complies with an internationally recognised standard (such as the SAE JA1011 standard). The difficulty with applying RCM primarily for safety/environmental integrity reasons is that the costs are measurable whereas the benefits are less tangible. In these circumstances an organisation has to rely on a qualitative judgement as to whether or not RCM is or would be beneficial. RCM cannot provide total elimination of the risk of a loss of safety/environmental integrity but where companies have applied RCM there is no doubt that they are better prepared than they would have been otherwise. If it is applied correctly, RCM provides a complete audit trail of the decisions made which may be used to support a safety case. Furthermore, if such an event were to happen, an organisation that applied RCM correctly would be able to claim that it exercised prudent, responsible custodianship by applying a rigorous process that complies with an internationally recognised standard, and would, therefore, be in a highly defensible position. The dangers of adopting a streamlined RCM approach have been discussed elsewhere. An organisation would be ill advised to adopt a streamlined RCM if safety/environmental integrity were considered to be an issue, since the effect of a safety or environmental breach could be catastrophic in the process industry. The user of a streamlined approach would also find it harder to mount a credible defence should a safety/environmental incident occur. It is the experience of the Aladon Network that many organisations continue to adopt RCM specifically for safety/environmental integrity reason examples include companies in the transport, pharmaceutical, metals, food/drink, water/sewerage sectors as well as the chemical/process industries. It is also interesting to note that maintenance outsourcing is also cited as a reason for adopting RCM and it is likely that this is safety/environmentally driven. The approach to and extent of outsourcing maintenance varies from one industry sector to another and from one country to another. Many organisations who outsource maintenance try to transfer the risk by imposing performance requirements and safety/environmental integrity guarantees on the maintenance supplier. It is becoming increasingly clear that the outsourcing organisation does not remain immune in the event of an incident occurring. Railtrack is a good example of this and most of the train operators suffer the wrath of the public/media despite the fact that maintenance related delays are often attributable to the outsourced maintenance company. As a result, the Aladon Network has been approached by a number of organisations who are planning to outsource maintenance and want to use RCM as an approach to specify the minimum maintenance that must be carried out by the outsourcing company. In such instances the main driver has been the desire of the outsourcing company to protect itself (as far as is possible) from the risk of a safety/environmental incident occurring. experience benefits Odorisation study and benefits have carried out a maintenance review of the local gas treatment systems used to odorise gas in . Prior to this review, maintenance was carried out to the specification compiled by the equipment manufacturers and suppliers. This was a calendar based regime of inspection and strip-down replacement of soft parts. Post RCM, the maintenance regime has become a

4 weekly routine check of pressures and levels and running of redundant equipment streams, a 6 monthly functional test of certain items and an annual maintenance consisting of a calibration check of the control system, functional check of a number of valves and reliefs and the scheduled replacement of a solenoid valve. This cost savings by implementing RCM is shown by comparing two 12 month periods, 1/10/99 to 30/9/00 and 1/10/01 to 30/9/02 as follows: History - All 126 units were commissioned between March and September 1998 - In 1999 and 2000 the equipment received an annual maintenance as recommended by the manufacturers. This maintenance included the replacement of solenoid valves, pump seals and bulk filters regardless of their condition. - Maintenance in 2001 and 2002 was carried out as per the RCM review. The main difference being that the review recognised no requirement for scheduled maintenance on the pump seals and bulk filters. It should be noted therefore that the oldest part is 2 years old and that eventually all parts will fail at some point in the future. 1999/2000 costs () 276,000 300,000 66,517 0 642,517 Total Annual Savings 307,806 Activity Annual maint Labour Annual maint Materials Pump breakdowns Filter breakdowns Total 2001/2002 costs () 72,576 100,000 151,326 10,809 334,711

Annual maintenance labour. The 1999/2000 costs is the system suppliers maintenance charge. The 2001/2002 costs are based on s own maintenance sections chargeable rate for 2 people for 8 hours to carry out the reduced maintenance. have the ability to carry out all of the RCM maintenance schedules. Currently only the system suppliers have the required competencies to carry out replacement of bulk filters and pump seals. Pump breakdowns. Number of breakdowns are taken from the odour fault database of which there were 36 for 1999/2000 and 126 for 2001/2002. It should be noted that for 2001/2002 the 126 breakdowns are a 100% capture. Whereas the 36 for 1999/2000 is estimated to be a 65% capture. Therefore the 1999/2000 figure has been projected to estimate a 100% capture. Pump breakdown costs Materials 471 Labour 730 (based on an average suppliers call out cost of 550 and a charge of 180 (36 x 5 hours). Filter breakdown costs Materials 491 Labour 730 (based on an average suppliers call out cost of 550 and a charge of 180 (36 x 5 hours). The reason for the increase in pump failures is a mixture of improved reporting (100%), increased gas flow therefore they are doing more work and the fact we are not replacing soft parts each year. The next RCM review should investigate whether those pumps that have failed are consistent with respect to the number of pump strokes done, and if so it should be possible to predict failure using this parameter. A review of the strategy on replacing solenoids every year should also be carried out.

Impact on *** company Impact on spares management Effect on maintenance Managing maintenance post RCM Fault data collection & analysis The Maintenance cycle set policy, analyse, do, record, feedback, review Process: Verify maintenance management tools currently in place Verify work feedback information arrangements Verify or create fault data collection arrangements Verify sources of reliability data (OREDA data currently used by and Optagon. Plus David Smith data.) Verify or create asset register Where to start applying RCM For a large organisation it may not be clear where an RCM initiative should start (regardless of the chosen method for implementation - ie whether local on-site teams are used to conduct the analyses or whether the super team approach is adopted). The prioritisation of the assets for study will be heavily dependent on the reason why RCM is being applied at a site and it is likely that the reasons will be different at another site. In other words, it is essential to understand what the objectives are of applying RCM at a particular site before it is possible to decide where the RCM analyses should start. For instance it would not be correct to stipulate that the, say, fire and wash system is the number one priority at all sites it may be at some but not at others. An approach for deciding where to star has been adopted by companies in the Aladon Network in other industries and it may well be appropriate to ***. The following is an outline description of the process; the approach described would have to be tailored to suit the *** and may require customisation according to the individual business sector ie E&P, T&D etc:

On a typical process plant the extent of the equipment installed can vary significantly. Possibly the most complex installation for the *** would be a manned exploration or production/processing platform in such installations the equipment is very varied as it covers everything from hotel services to fire fighting and heliport to the provision of seawater cooling. In order to make the process of prioritisation manageable it is necessary to write a series of high level functions for the installation. The physical assets/systems can be then be slotted into these functions. The next stage is to identify the criteria against which these functions will be judged as being important (or otherwise). For each criterion, a series of statements will have to be prepared so that the degree to which each criterion was applicable could be judged. Each of these statements is allocated a raw score. Clearly and example criterion would be Health, Safety, Fire and Environment; Impact on Production would be another etc. In a recent application of this approach to an offshore installation there were 8 criteria identified as being important to the business and each criterion had between 3 and 5 levels of importance identified. It is possible that a common list of criteria can be established for the *** (each with its levels of importance defined) Once the criteria have been established, weights are assigned to each criterion according to its relative importance to the particular site being considered; this ensures that the specific operating context/requirements of the site are taken into consideration. The weights are determined (as far as is possible) using a structured approach

The final stage is to apply the weighted decision criteria to each function in the site functional breakdown. A raw score is generated according to the level of importance allocated within each criterion; this raw score is multiplied by the weight assigned to the criterion to give a weighted score for that criterion. The weighted scores for each criterion are then totalled and recorded to give an overall weighted score for each function. In this way, the relative priority of each function is identified.

The approach outlined above is a weighted ranking assessment. It is not possible within the scope of this study to define in detail the approach that might be applicable to *** as a whole or to its individual businesses; this would have to be considered as a separate piece of work. Decide who will need to join RCM study team for given assets, thus producing list of people who will require 3 day RCM training (Train sufficient staff to gain sign-on to the process. On Morcombe, there was a 2 week shift system. Staff took on 3 days RCM training then on day 4 looked at Functions etc. Use super team later.) Decide on RCM facilitators Organise RCM 3 day and Facilitator courses Types of Asset: 1. Distributed assets, simple in nature, many similar installations T&D 2. Large complex one offs Compressors, Terminals, LNG 3. Complex systems, between 5 and 10 others, all slightly different sub-systems to complex installations 4. Simple sub-systems, more than 10 others, context differs in many ESD, firewater, pumping, heating, chilling Applying RCM To be completed How will application of RCM differ in these assets? 1. Apply the Model Model reviews, Templates, 100% & 90% copies; on Transmission sites see 4 below 2. Classical RCM one off reviews but with a view to seeking commonalities where evident eg fuel systems, cooling systems etc for re-use elsewhere 3. Create template for common functions then build reviews on templates eg gas treatment, flow control 4. Create more complex template containing some failure modes, build models based on templates then use 90% copy process to address contextual differences It is probable that there will be limited scope for templating on BG E&P, LNG and Power assets due to the limited commonality. However there will be considerable benefits from applying this to T&D assets. Where large, complex items are concerned it is likely that the manufacturers maintenance instructions will be adequate to the extent that, when targeting RCM, it is likely that more benefit will be found by applying it to peripheral systems. At review time Select plant for review Gather information drawings, maintenance instructions, history, fault logs, operations instructions etc Gather the experts RCM trained operators and maintainers, perhaps (in the early days especially) a manager or two.

Carry out review and record results and decisions output = tasks and frequencies, skills & competencies, resources & tools. Optimisation determination of actual work orders based on output of RCM. Schedule work, do work, capture work feedback Collect fault data Analyse all sources of reliability data built up since adoption of RCM based maintenance to determine effect on maintenance activity and availability (have non routine work items increased, decreased or stayed the same is availability better or worse)

Quantity
Scope of equipment and plant to which RCM can be applied Assessment of likely benefits in terms of maintenance costs and availability

Resources
RCM Review groups and facilitators 2 options. 1. Create teams locally from expert operators and maintainers. Team members will be required to have undergone 3 day RCM training. Team will carry out review of local equipment only, however, Facilitator may be peripatetic. 2. Create super-team of experts in the generic equipment under review (eg Compressors, LNG etc). Team will fly in to carry out reviews. This team will spend longer initially as they familiarise themselves with local plant and operating regime, however, they will become so well versed in application of RCM that the review will be more complete, consistent and correct than would be produced by the local team. There is a danger of common mode failure where an incorrect interpretation by the super team may spread throughout all reviews. Aladon consultants input Structure required, skills and facilities Tools RCM data recording, maintenance management moving information between them Fault data collection and analysis feedback into RCM

Timescale
Training Appreciation, application, facilitation Prioritising work Complex reviews Simple reviews Templates and models Review copying