Transcription
Well, uh, good morning to those of you in the Americas. Uh, good afternoon if you're in Europe and Africa, and good evening to those of you further east. My name is Steve Lewis, and I'm a Senior Director of Risk, based in the UK. Thank you very much for joining us and welcome to this Risk Tech webinar, which is the fourth in our fourth series that we've run since May last year. The topic today is about Reliability Centered Maintenance, and what it can do for you. Hopefully, we can provide some useful and practical insights for you.
Before we get going, if you've attended any of our webinars, you'll notice we upgraded to a more automated registration process. Hopefully, you found that a lot easier to register for, certainly for more than one webinar at the same time. And we're also using Zoom rather than Skype for Business. Uh, I think because people do seem to prefer that.
A quick spot of housekeeping. We've muted everybody, so the sound won't be distorted by any background noise. Um, and if you'd like to ask any questions, then please use the Q&A function. So that's if you, if you move your cursor down the page, then the Zoom tab pops up at the bottom, and there are a couple of speech bubbles that say Q&A. Click on that, type your message in. I'll keep track of all those questions during the course of the, uh, the webinar, and then at the end, I'll cover as many of them as we can, and certainly within the 40, 45 minutes we've got available, uh, today.
Okay. Um, I'd now like to briefly introduce Risk Tech for those of you who don't know us. Uh, apologies if you've heard this 101 times before, I'll be pretty quick. Um, so everything we do is around risk and safety management, particularly in sectors where the impact of loss is high. We've well over 300 people across 16 offices worldwide. We are part of TÜV Rheinland, who are a two billion euro provider of testing, inspection, and certification services. We provide our services across five business lines: consulting, uh, where we look to deliver proportionate solutions, uh, to help reduce and manage risk. We have online and classroom training, postgraduate education. We provide associates who work at client locations to help fuel resource and skill shortages. We also provide industrial inspection and assessments as part of asset integrity management programs, and we also conduct research and development in the field of risk and safety management.
Now, I'd like to introduce our speaker today. Uh, speaker is Dave Fiddler. Dave's a Principal Consultant. He's based in our Warrington office here in the UK. He's over 20 years of varied risk management and technical safety engineering experience, mainly in the civil nuclear industry in the UK, but also Central Eastern Europe. So, its clients include EDF Energy, Slavinsky Electron, and Shares, with a focus on maintenance optimization and supporting safety case submissions. This experience includes the application of RCM, which is Reliability Centered Maintenance, and our topic today, identification of safety critical elements, development of performance standards, safety case production, HAZOPs, and hazard identification. So, I mean, Dave has a huge amount of experience. Uh, it's really delighted Dave is here today to be able to give this webinar. So, over to you, Dave.
Hello, ladies and gentlemen. Um, yeah, welcome. It's a pleasure to speak to you today, uh, regarding what RCM can do for you. And today, we're going to focus primarily upon how we can optimize the effort required to apply Reliability Centered Maintenance. But just to set, um, a few kind of, uh, ground rules, we'll look at a couple of the basics to begin with. First of all, why do we actually do maintenance and testing? Well, as we can see from the slide, in in its purest form, maintenance has three primary objectives: to reduce the probability of failure, to restore the inherent level of equipment reliability. Now, what do we mean by that? There's been a long-held common misperception that somehow, you know, a higher level of reliability can be maintained into equipment, or a higher level than that was originally designed in, and that is simply untrue. It doesn't matter how much maintenance you do, or how correct the maintenance is, the best you can hope for is that you will restore the original level of reliability that was inherent in the design in the first place. Uh, can you improve reliability? Yes, of course you can, by modification, but not by simply applying maintenance. And we attempt to ensure that performance is not degraded by neither time nor usage. So, for example, if you're going to overhaul a piece of large rotating plant, a pump, for example, a diesel generator, gas turbine, then the idea is that after the overhaul is complete, then the efficiency of the item should be fully restored.
Now, there are differences to that. Certainly, as plant gets very old, then yes, no matter how well you've maintained it, especially mechanical plant, rotating plants, there may be a fall-off in overall efficiency. And whether or not money is spent to actually correct that, for instance, by replacement or not, is really an operational decision. So, for example, for very old plant that is nearing the end of its life, it may be an operational decision whereby you simply live with the small degradation of efficiency because the cost of replacement would not really, uh, give you a return on the investment.
I think, uh, testing, and specifically testing of standby or redundant plant, um, no test, however successful it is, will all, will, you know, guarantee that the item will work the next time it's called upon to do so. However, if all works well every time we test something, you know, a backup system, for example, it does provide confidence that the redundant systems will operate effectively when called upon to do so. And for most cases, we sincerely hope that will not happen. For example, emergency shutdown systems. We always hope we will never have to operate an emergency shutdown system in anger, but of course, we have to, you know, test them periodically to, to give us some confidence that they will operate when called upon to do so. However, if we have a, you know, a significant number of, uh, routine tests that actually don't work or throw problems, then that in itself can prompt modification action to improve reliability, uh, based upon the, the operating experience gained.
So, the RCM process itself. Well, as you can see, it began in the U.S., U.S. airline industries, in, uh, the late 1960s. It's been developed across, uh, many industries over, over the, uh, the decades since. There is an international standard. Uh, for those of you who aren't aware, SAE is the United States Society of Automotive Engineers. So, as you can see there, although the RCM process was developed in the U.S. airline industry, the actual recognized international standard is actually from the automotive industry. Okay. And the RCM process is really very, very simple, and it simply starts with the seven questions as listed below, work through in the order that they are listed. So, really, the RCM process is no more than a formalization, or if you like, providing a structure of what good engineers have been doing for many, many, many years.
First question is: What is the item supposed to do and associate performance standards? Okay. So, not only what is its function, but how well does it have to function? So, for instance, you may say a pump is designed to pump water, whatever fluid it is, but then you would say, at this particular pressure, at that particular mass flow, at that particular temperature. Those performance criteria are really, really important to include in functionality. In what ways can it fail? Uh, what are the events that cause each failure? What are the mechanisms, if you like? And what happens, or what are the consequences of each failure? And in what way does each failure matter, if indeed it matters at all, in terms of safety or reliable operations? For those items whereby failure does matter, then we ask: What preventative maintenance task can be performed proactively to prevent or diminish to a satisfactory degree the consequences of the failure? And the last question, of course, is: What must be done if a suitable preventive maintenance task cannot be found? And that is normally correcting a design or modifying the plant.
So, applying RCM to optimize, uh, optimize preventive maintenance for existing facilities. As you said, the original RCM process was intended to inform the design of airliners, primarily the Boeing 747. Now, it's often an inherent part of designer facilities across many, many industries. And it is also often used across many industries with the purpose of reducing in-service failure which adversely affects the bottom line. And that is perfectly acceptable to do so. Because if we think about it, when facilities are brand new, much of the maintenance, or the preventive maintenance, is reliant upon the, um, manufacturers, you know, uh, maintenance plans, you know, that they're advised maintenance. However, as time goes on, and the facility generates operating experience, then it can be the case, and it often is the case, that the original, um, maintenance plans provided by the manufacturer can, in fact, be changed in terms of tasks and frequencies to, to optimize the effort required, uh, in the PM plan.
However, when we look at RCM as a function of the design, it's normally a top-down functional approach, as appropriate. I.e., you take the design, you break it down into its different kind of subsystems, into its various functions, and then you have a look at your, those functions in accordance with the seven questions we saw on the previous slide. However, it's rarely suitable for in-service equipment, and this kind of forces a bottom-up hardware approach. I.e., systems and subsystems that are effectively analyzed on a component-by-component basis. And by that, I actually mean a team of engineers is often around the table with a large print piping and instrumentation diagram, going through the system component by component, answering the seven questions we saw on the previous slide. Unfortunately, one of the primary outcomes of doing so is that during the FMEA, when you're answering those questions, the failure effects of many components are found to be negligible in terms of adversely affecting production. But as you can see on the slide here, by then, it's too late. The time, the cost, the effort of the analysis has already been incurred.
However, what about a solution? One of the activities we can undertake that effectively, uh, precludes that wasted effort is, before you begin the actual RCM analysis, undertake a process called critical component identification. Now, this, a few years ago, was, uh, developed by the Institute of Nuclear Power Operations (INPO) as part of their what they call AP 913 Equipment Reliability Process. And this can be readily applied to any given industry. It's not nuclear-specific, I assure you. And it begins with a relatively simple process of identifying and categorizing components into these four levels of criticality, which we'll look at in more detail in a moment. And they're actually categorized against this set of predetermined criteria. Of course, that criteria will be industry-specific, and it's usually, you know, performed by a single person as a desktop. Yes, there'll be a simple challenge process to check for accuracy, but it doesn't require a team of engineers. It requires a single trained person and can be done as a desktop exercise.
So, looking at the four categories: We have critical. And you can see there, the failure of the component is intolerable and is likely to directly affect reliable production. There's an important thing to note here, that a subset of critical components are those categorized what we call Single Point Vulnerabilities, or SPVs. Now, the subject of Single Point Vulnerabilities is probably adequate for, for a webinar all of its own. But just to give you an example, uh, something like control valves, especially air-operated control valves, um, often critical in terms of production. So, if you've got valves which are controlling the flow, uh, of three different chemicals into a reaction vessel, for example, it's obviously massively important to production that those flows are very accurately controlled. And of course, something like an air-operated control valve is often used because of the speed and efficiency with which it responds to changes in the required modulation of the valve. However, when we look at something like an air-operated control valve from the programmable logic controller that is actually modulating the valve from its actual actuator, its pneumatic or air supply, um, it's, uh, the actual valve components. There are a whole host of things that can go wrong that will effectively cause the entire control function of the valve to actually fail at that point. So, that's just an example of, of an SPV.
Non-critical. Okay. The failure of the component is tolerable, but it's simply more cost-effective to conduct a preventive maintenance task than to allow the component to fully fail. Uh, that could be simply because of the downtime involved in repair, or it could be the actual cost of the spares. Okay. Low significance. The failure is tolerable. It's not considered worthwhile performing PM tasks. This is often low significance. It's often referred to as run to failure. Okay. We have, we haven't got that an issue with this degradation will be detected by routine plant monitoring, and eventually, it'll end up under, on the plan, and it will be repaired. Not assessed. Usually passive components whose failure would obviously have no adverse effect on production. A good example of that is something like a drain valve, where the drain valve is only opened to drain a vessel, for instance, for inspection. If the valve fails to open, the normally manual valves, then all you've really got is a minor delay in, in getting the, the vessel drained and the inspection carried out.
So, when we look at the INPO AP 913 streamlined process, we can see that we have the critical criteria feeding into the top. Uh, we have the non-critical criteria feeding in, uh, later on. And effectively, most of, or many of those, those boxes you're looking at there are actually inputs of previously performed work, for instance, critical component identification into the process. So, it's not, it's, it's no longer an integral part of the team-based RCM. It's predetermined criteria we're now looking at, which makes the whole process of RCM analysis much slicker and much more cost-effective.
So, the actual critical component identification process itself, as you can see, is very, very simple and lends itself to a desktop exercise. It's listed components. Okay. What are the consequences of each component failure? Match that against the critical criteria, which we will come on to some examples in a short while, and just ascertain whether it falls into critical or non-critical, or indeed, it ends up being low significance and therefore to run to failure component. And the activities that you follow to identify whether or not, uh, if it is critical, it's a single point vulnerability.
So, I did say we'd come up with some critical criteria. Um, so I've picked some from UK civil nuclear, and I'm, I'm rather hoping that everyone will see just how simple a process, uh, this is. So, these are predetermined criteria. So, we said earlier that for critical, the failure of the component is intolerable to the facility and likely to directly affect reliable production. Well, there's the criteria against which the, the categorization of critical is measured. For example, an unplanned reactor trip, an unplanned control reactor shutdown, or a significant power transient or derating the reactor to greater than 100 megawatts electrical. Any one of those reactor trip, shutdown, power transient, or d-rate, none of them are going to have a major safety consideration because the nuclear systems are designed to allow these things to happen safely, but they will cost a great deal of money. And with electricity generation, we are now talking into millions of, of euros, dollars, etc.
If there's an unplanned entry into what we call a technical specification action condition requiring a shutdown equal to or less than 72 hours, or the immediate suspension of the hardliner movement of transfer radiated fuel, that again is not a safety consideration, but it's gonna take an awful lot of money off the bottom line to, to, uh, to recover from that. Any degradation of a critical safety function. Now, that doesn't mean that degradation of a critical safety function when it's called upon to operate. That's normally when you go and test the critical safety function, you've found it's in a failed state or there's an issue with it. Everything's fine, reactor's at power. Yeah, there's, there's no issues here, but the fact is, you're gonna have to actually spend quite a lot of money to make sure that that is corrected immediately. And any unplanned actuation of post-trip sequencing, effectively spurious actuation of safety systems that cause you to bring the reactor off unnecessarily. And it's the same thing with any unplanned reactor protection system.
For non-critical, as you can imagine, it's, uh, less than 100 megawatts now, but greater than five. So, it's not gonna cost quite as much money, but it's still going to be quite expensive. So, the choice is, let's have a look and see if we can provide a preventive maintenance task which will actually head off that failure, if you like, and bring you back to the three objectives of maintenance that we saw on the, on the very first slide. And again, we've got loss of redundancy, or a likely health and safety, environmental, or radiological reportable event. We're now obviously talking about near misses, I think. So, any components whose failure may actually bring about a near miss, then that would be categorized as non-critical, and yes, we would look to apply, uh, preventive maintenance to try and reduce that probability of failure. As we mentioned earlier, touched on, if the, the cost of refurbishing or replacing a component is, you know, kind of higher compared to performing the, the preventive maintenance task, then obviously it's the, you know, the kind of cheaper option to perform the PM task. And in a similar way of, uh, the previous one, a loss of function of emergency response or emergency preparedness equipment, because these things cannot be tolerated for a lengthy period of time, in case you need to bring them into use. Therefore, there would be, uh, classes non-critical, and they would be maintained accordingly.
So, some illustrative examples from other industries. Uh, I've chosen two here which kind of go hand in hand: train operators and rail infrastructure companies. For critical, a failure resulting in unexpected cancellation of a mainline train journey or a delay of more than three hours, because certainly in the UK, any delay more than three hours or cancellation, we have legislation which allows passengers to, uh, to request compensation claims from a train operator. They're normally given in, kind of, vouchers for, you know, future free train journeys rather than, uh, trying to send actual your money back to the client. For non-critical, a failure resulting in the unexpected cancellation of a local service. Now, that would incur the cost of providing alternative road transport. Again, in the UK, the train company would be expected to immediately hire a series of coaches and allow the passengers to complete their journey by road, an expensive exercise in itself. For the rail infrastructure company, a signaling failure, for example, that results in the unexpected unavailability of a main line for a period of greater than 24 hours. That in itself would incur penalties from the train operators, who will probably have, you know, already been asked for compensation by the passengers. So, that cost is going to be passed on. For non-critical, uh, a failure resulting in unexpected requirement to undertake corrective action during quiet hours, deployment of nighttime maintenance crews. In the UK, I, I live about 70 meters from, uh, a main, uh, the West Coast Mainline, and you may be surprised at how many times, uh, nighttime maintenance crews are deployed to actually, you know, uh, recover from, from problems that have happened either on the line or in the signals, and it all has a cost associated with it. So, maintenance is the better option.
For other industries, similar industries, specific criteria can have been developed for many industries: energy providers, retail, logistics, especially those providing just-in-time spares, manufacturing itself. Thing. So, similar criteria can be really provided for any given industry with, with not too much time, cost, and effort associated with it, thereby streamlining the, the RCM process and allowing that process to focus only on the plant which is really important in terms of safe, reliable operations.
So, in conclusion, RCM really has attracted an unfair reputation for being time-consuming, expensive. Unfair because there's only in the case when adequate preparation has not preceded the analysis. As with many processes, if you target your efforts, you will normally achieve an improved cost benefit. I'll now hand you back to Steve, who's going to administer the Q&A session.
Okay. Yeah, thanks very much, Dave. So, we're open now for any questions. So, if you'd like to type your questions, uh, into the Q&A function, uh, we'll, we'll, uh, see if some come through there. But just whilst we're waiting for the first, uh, first few to come through, um, yes, Dave, so let's think about, so how much time and effort does it take to do this properly? So, say that you've got a new, I mean, do you go about this for a whole, for a whole new facility? Um, which stage of the design would you do it? You know, how long would it take? How much effort, um, to do the critical component analysis and then perhaps move on to the, you know, the deeper RCM tasks for the critical components?
That's a little variable depending upon the, um, uh, the depth of the design. For a new facility, um, but that would normally be done really as part of the design reviews. So, that's not really any great additional effort, uh, in itself. In terms of actually, um, looking at it, whether it's design or whether it's for an existing system, normally when you've got someone who's quite experienced, has gained some experience in doing this, you can only do a fairly major system in, you know, a day or two, really. So, for instance, at a UK nuclear power station, I would expect to do the entire diesel generator, or the emergency diesel generator system, including all of its auxiliaries, fuel oil, lube oil, air cooling, air starts, uh, protection, electrical. I'd expect to do the entire thing in two days by one person as a desktop.
Okay. Okay, very good. Thank you. Um, we've got another question come through from, uh, Rajesh. Um, do we need to apportion for reliability in RCM? I'm not quite sure I understand the question. Can you think one that's getting that, Dave?
Yeah, I, I think so. And obviously, um, the, the gentleman can come back at us if I've, uh, misunderstood. Um, in terms of detailed reliability claims, for example, QRA or PSA claims, they're not strictly speaking part of the RCM process. Because the RCM process simply deals with what we call probable failure modes. So, for example, the FMEA supporting the RCM process would be significantly different from FMEA supporting a PSA or a QRA. The second case, yes, the, the reliability data is very much a component of that process. Whereas for the RCM, we only look at what we call probable failure modes. If you like, we look at the usual suspects. So, if we're looking at a horizontal pump, we know that the pump shaft will wear, we know the impeller will wear, we know the bearings will wear. So, we will simply look at the maintenance required to, uh, slow down that wear, or, you know, kind of reduce the probability of it failing, without really looking specifically at the actual, uh, reliability data as it were. I hope that's answered the question adequately.
Okay. Thanks, Dave. Oh, there's a follow-up from Rajesh here. He's referring to a railway metro, and specifically, how would we go about doing an RCM for the track infrastructure and the stations?
Right. Okay. What you would actually do is simply take each of those, those aspects of, of the metro, yeah, in, in, in isolation to an extent. Well, initially in isolation, they're going to come together at some point. And then you would actually simply, and that means simply, break it down into very subsystems. So, for the, the station, you could be looking at, uh, how the passengers move about, you could be looking at, uh, lighting, communications equipment, thing, um, you know, all of the kind of the physical aspects. And take each of those subsystems in turn, and then actually do the critical component identification for each system, and then apply the RCM. Likewise, for the track infrastructure, uh, because obviously we're dealing with a metro now, so we're also dealing with the, if you like, the, the civil or structural aspects, uh, thing, of, of the tunneling system, uh, and all of the signaling, etc. Uh, and included in that would be emergency escape routes, for example, that would be a subsystem all of its own. You would again do the critical components identification, identify your critical components, including your SPVs, your non-critical, and then push those through the RCM analysis to, to ascertain the appropriate, uh, preventive maintenance regime for those. Again, I hope I've answered that question adequately.
Okay. Thanks, Dave. Um, I think it's kind of a related question, really, from, uh, Mahanad. Does RCM cover all the fields, um, including different equipment? And he's saying, for example, if you take an oil and gas field, you know, a large infrastructure, and particularly offshore, then you've got lots of categories of, uh, plants. For example, you might have a semi-submersible, uh, drilling rig, a drill ship, you know, a jack-up, a fixed-leg jack-up platform. Does RCM cover all of those different types?
Absolutely. Um, and just to kind of, um, widen that response a little bit, uh, certainly in, in oil and gas, I think, uh, just like most industries, if we think about it, most of the components are not necessarily specifically designed for the industry. You know, because pumps are fairly generic, valves are generic. Yes, of course, in oil and gas, you do have some specific components in terms of well completions, in terms of your formation isolation valves, uh, packers, etc., which are indeed entry-specific. They're not used anywhere else other than in drilling. Now, I think, but for the vast majority of the plant, it's, it's made up of pretty generic type components, by all means. And as you break the systems down into subsystems, down into assemblies, then it's, it's no difference in that respect from any other industry, uh, once you're applying the critical component identification.
Okay. Thanks, Dave. A good question from Steve. Is condition-based monitoring the way ahead, as equipment failure can be anticipated and the defects that can be introduced by conducting, uh, planned maintenance eliminated?
Yes, is the short answer, with with a caveat that it depends upon the operating context of the component. For example, if Steve would like get in touch with me at some point, I'd, uh, draw his attention to a particular system, say, Hinkley Point B nuclear power station, whereby in the water treatment plant, every duty and standby pump that's in that treatment plant, um, can in fact be common, because the, the duty pump is nearly always running. As long as you routinely change over the standby, you can come on those as well. But there are, uh, there are two pumps there which are what you call the emergency, um, makeup pumps. To actually start those, the run, for instance, the vibration monitoring during normal operation would put such a slug of water forward into the system, it would trip out all of the feed heaters, it would flood them all out, and thereby you'd probably be pulling your reactor power down eventually by about 11 to 20 megawatts and lose a lot of money. So, because of the operating context of those pumps, there's no recirc line to, to start them back to their, their water supply, you can't really run them to come on them. So, on on that particular occasion, you revert back to traditional teardown inspection and overhaul. I hope that's answered Steve's question. But again, get in touch with me, you know, after this, and I'd be only too happy to discuss further with anyone who chooses to do so.
Okay. Thank you. Uh, question from Mahmoud. Can we consider failures happening because of power? Let me rephrase that. Are failures that happened due to power failure critical?
That rather depends upon the availability of backup systems, primarily. The answer to the question is normally yes, because you really do not want to have an over-reliance upon emergency power generation. So, for instance, at a power station, the, the, you know, the, the kind of the connections to the grid are vitally important, because, yeah, you can say, well, if you lose the grid, then you've got your emergency diesel generators. But of course, if you lose the grid, the first thing that happens is the reactor trips, which is a huge loss of production. Uh, so the emergency diesel generators are simply there to actually keep the post-trip cooling and keeping the fuel cool. I think you are no longer an ability to generate electricity until you've recovered. So, if those grid connections were inherently unreliable, uh, yes, you would indeed regard those as is critical, not just in terms of safety, but in terms of lost production.
Okay. Thank you. Another question from Muhammad. Okay, so what's the difference between RCM and JD Edwards? I don't know if you're familiar with JD Edwards, but I think that's a maintenance management software, like sort of Maximo.
Dave: Ah, right. Okay. So, what's the difference between RCM and a maintenance management system for completing the preventive maintenance for all equipment? It looks the same to me.
Yeah, the, the things, you know, like Maximo, uh, or any kind of asset management system, or so, for instance, uh, every RCM workbench. These are just tools for the recording of the RCM analysis. The RCM analysis is performed by human beings. It requires the human brain and it requires the knowledge of human beings on the functionality of the plant. Which, for instance, why any RCM analysis, it would be hugely important to have both operations and maintenance involved. Operators are very skilled, normally, at determining the consequence of a failure of any plant item, whereas maintainers are very good at identifying the causes of such failure. I think. So, it's really important. It's also really important to get work planning involved, because when you embark on any RCM study, the first question, when I must do it, the first question I always go and ask the facility is, how do you intend to implement this? How can I present this analysis in the most efficient manner possible, so you can enter it onto your Maximo, or your JD Edwards, um, you know, whatever your asset management system is. So, these are just systems for recording the results of the analysis. The analysis is still performed by a human being. RCM is not automated in that respect.
A question from John. Thank you, David. This is really interesting. I see significant alignment between the RCM process with its seven questions and a bow tie analysis taken to a very detailed level. And I think what he's referring to there, we have a barrier against a threat or a mitigation event against a consequence. We can sort of drill down to what we need to do to make sure that barrier is working. What I'm not sure of is how criticality is determined within the two processes. I'd appreciate any comments you may have about your view of the similarities.
Ah, right. Okay. Um, certainly in RCM, the criticality, as you can see, is, is determined against that, you know, specific list of criteria, which is, you know, unashamedly aimed at protecting reliable production, I think, rather than safety. In fact, if we think about single point vulnerabilities, certainly in the UK, because of the goal setting or ALARP-based UK legislation for determining, uh, safety cases, I think a single point vulnerability effectively cannot exist in a safety system, because all safety systems in the UK under the ALARP process must comply with what we call a single failure criterion, i.e., no single failure can take out the entire safety function. So, an SPV can't really exist in those circumstances. So, SPVs only really exist in production plant, if you like, uh, often called balance plant, I think. Um, and another thing to, to be very, very, uh, very much worth noting is that there is one major caveat with RCM, and that is RCM, when you're performing an FMEA, it is focused entirely upon equipment failure. And what that really means is, when you assign a preventive maintenance task, it's the appropriate frequency. The RCM process will effectively take the view that that task will be performed correctly. It does not take into account human error, because, you know, that type of process cannot legislate for human error. As you can in a bow tie, I think. Like I said, the RCM process will actually simply take the view that if a maintenance task is assigned, it is logistically and technically feasible, and it will be performed correctly. The maintenance teams will do exactly what they're expected to do. I hope that kind of helps. But again, come back to me, um, you know, you've got my email address, I'd be only too happy to, to, you know, to talk to you in greater detail.
Okay. Thanks, thanks, Dave. Um, there's no more questions coming through, so we're up to 40 minutes. So, I think we'll wrap up now. Um, what we will do is make a recording available to everybody, hopefully within a sort of 24 hours. Um, and when you leave the webinar, you should automatically receive a survey in your browser. Uh, it would literally take you, I don't know, 10, 15 seconds to complete that. We really do appreciate your feedback. You know, especially if you can identify any topics that you'd be interested to hear about for future webinars. I mean, what we want to do is, you know, provide topics that you find interesting. So, um, just let us, let us know what you'd like to hear. And as Dave said, if you've got any questions, any technical questions arising from the webinar today, get in contact with Dave directly. His email was at the at the start. Get it all, we can find him through the company. There's a couple of email addresses up there, or you can just go to our website, and there's a ton of forms on there, we can just sort of fill in, and we'll get back to you. So, thank you, Dave, once again. Uh, thank you everybody else for your attention. Really grateful you take your time out of your day to listen into, uh, our webinars. The next one is on Thursday, same time, same place. Um, and we're looking at risk-based inspection of hydrocarbon-containing equipment. So, look forward to that, hopefully, you can join us for that one too. So, in the meantime, please stay safe, stay secure, enjoy the rest of the day. Um, thank you. Goodbye.