Latest / Upon Further Inspection / Episode 4 - Gotta Document Logic

Transcript

Episode 4 - Gotta Document Logic · Upon Further Inspection

• 0:00
  Upon further inspection, a Mechanical Integrity podcast goes beyond the data and dives into the people challenges and stories behind reliability and inspection.
• 0:11
  Whether you're in the field or in the office, this podcast is for you because mechanical integrity isn't about assets.
• 0:18
  It's about the people who keep them running.
• 0:20
  You are listening to episode four.
• 0:22
  Gotta document the logic.
• 0:23
  Discussion on RBI.
• 0:40
  Branden: So as usual, I'm here with Greg Alvarado.
• 0:42
  Branden: I'm Branden Stucky.
• 0:44
  Branden: How's it going, Greg?
• 0:45
  Greg: It's going great, Branden, how are you doing buddy?
• 0:47
  Branden: I am doing wonderful today.
• 0:50
  Branden: I'm doing just fantastic.
• 0:51
  Branden: So.
• 0:52
  Branden: Today I wanted to kind of pick your brain and, and hear a little bit from you and, and kind of discuss with you the concepts of qualitative RBI versus the concept of quantitative RBI.
• 1:04
  Branden: Obviously when you, you look at the word quantitative meaning number you know, we don't quite have that in, in the qualitative version.
• 1:11
  Branden: Why do we even have these types of, why do we even have this difference?
• 1:15
  Branden: Why, why is this a thing right now?
• 1:18
  Greg: Well, to put things in proper context, we go back to beginning of this project that is the API 5 81 RBI project.
• 1:26
  Greg: I should state we really didn't know exactly where we were gonna end up or how it was gonna go.
• 1:31
  Greg: We didn't, we actually thought RBI would all be done on spreadsheets.
• 1:36
  Greg: Graph paper or something.
• 1:38
  Greg: And of course at the time, DNV was the prime contractor and all the other members at API user group were user are, were users of, of the technology.
• 1:50
  Greg: So anyway.
• 1:51
  Greg: We began developing those risk management tools and we saw that the more quantitative we got, in other words, the more data hungry we got, the more actual go get the data the more confidence we felt in the calculations.
• 2:08
  Greg: But we also had to look at the models and how much accuracy we wanted to achieve in the answers.
• 2:14
  Greg: And initially, and, and again, I'm just using the API project as an example, we actually created what we call a qualitative RBI and what we called at the, at the time, a semi quantitative RBI.
• 2:27
  Greg: And just real quick, Branden, then I'll let you ask me some more questions, but.
• 2:33
  Greg: On the qualitative RBI, we ask things we wouldn't ask, like what temperature does it operate at?
• 2:39
  Greg: We'd ask them to pick from a list what of ranges of temperatures that a piece of equipment might operate at.
• 2:45
  Greg: Instead of asking for certain specific numbers and information we'd ask for.
• 2:50
  Greg: More qualitative or general types of information.
• 2:54
  Greg: And we felt like we were kind of feeling our way in the dark, and then we began developing a more quant, more quantitative mos.
• 3:04
  Greg: But in actuality, I'm glad we're having this discussion because I actually juxtaposed them.
• 3:08
  Greg: We started out with a more quantitative approach.
• 3:13
  Greg: Then the group wanted to back off on the quantitative approach because of perceived level of effort to do it, to go to a qualitative approach.
• 3:23
  Greg: What I think gave us an advantage at that time to develop a qualitative approach is that the quantitative approach in, in, in, in doing that, we came up with calculations where we felt like.
• 3:36
  Greg: There was very little uncertainty, so to speak.
• 3:39
  Greg: We understood a lot about what was going on.
• 3:41
  Greg: There wasn't a lot of guesswork, so that when we popped back to a qualitative way to do it, we said, okay, what are reasonable cutoff ranges of temperatures, pressures, and other types of things, metallurgies?
• 3:55
  Greg: How specific do we really need to be?
• 3:59
  Greg: And after a while we figured out, and as technology and computers continued to develop, that we could definitely become more quantitative and, and we felt like that was gonna bring a big, bigger payback, that it would also make it a more systematic and, and reproducible.
• 4:17
  Greg: Consistent process you, as we became more quantitative, but we realized we didn't, we become so quantitative and coming up with a method to prioritize inspections.
• 4:28
  Greg: With that, it became impractical to do so.
• 4:33
  Greg: That's kind of how we started.
• 4:35
  Branden: Yeah.
• 4:36
  Branden: What's the timeframe of this?
• 4:37
  Branden: Just for the listeners?
• 4:38
  Branden: What, what was the timeframe of when all this happened?
• 4:40
  Branden: Because it, it's been a little while and some of the things that you're talking about might be.
• 4:45
  Branden: Perceived differently.
• 4:46
  Branden: So what, what was the timeframe here to starting?
• 4:48
  Greg: Yeah, this started in the early nineties, like 91, 92.
• 4:52
  Greg: We were doing a lot of little pilot projects, proofs of concept along the way at, at different refineries.
• 4:58
  Greg: There was a user group that was made up of only owner operators.
• 5:02
  Greg: And then in about I don't know, 1993 we realized we couldn't do this on paper.
• 5:08
  Greg: It was gonna need to be done by computers.
• 5:10
  Greg: And then in about, I don't know, 94, or 95, we started getting a better handle.
• 5:16
  Greg: On the semi-quantitative versus qualitative thing.
• 5:20
  Greg: So, you know, we, we were, we were, we had a, we had now abandoned in 581 the old qualitative approach by the late eighties, or excuse me, to late 90s And we're moving more toward just what we call the level two approach, which then morphed in the late 90s into what people see now when they use the 581 software, or at least more toward that.
• 5:45
  Branden: So, so you talk about software, right?
• 5:47
  Branden: So the original qualitative method, the point was to try and make it, you could hand calc now, technology's a lot better than it, than it was.
• 5:56
  Branden: It's advanced a ton since then.
• 5:59
  Branden: Do.
• 6:00
  Branden: If you were to have done this again, if, if API were to have started this whole JIP now, how do you think it would be different, do you think?
• 6:07
  Branden: Do you think we would be further than we are now?
• 6:11
  Branden: In just that short period of time, do you think that we would be leveraging AI or other types of concepts now?
• 6:18
  Branden: Or do you think that the committee still would've started out with trying to do a hand calc type model?
• 6:25
  Greg: I, I don't think it's a either or answer.
• 6:28
  Greg: I think we'd be, if we started now, I mean, I don't know if we looked at where we were back in 1990 inspection programs and worked up until now, I think the industry would've been okay if the industry would've started doing damage mechanisms or views without even having RBI yet.
• 6:48
  Greg: Because that was one of the big benefits that RBI brought is it forces the practitioner to do a systematic damage mechanisms review and understand the damage better.
• 6:57
  Greg: Where is it gonna be?
• 6:58
  Greg: What's it gonna look like?
• 7:00
  Greg: What kind of NDE techniques should I use Prior to the early nineties, honest and in my impression, there were some people doing that, but not a lot of people.
• 7:10
  Greg: And there were other, there were other side benefits, right?
• 7:13
  Greg: It's kinda like, the NASA space program.
• 7:15
  Greg: If you look back, you know, to the sixties and all the work that was done there, there were some other benefits that came out of it than just landed on the moon, right?
• 7:24
  Greg: Landing on the moon, there were microwave ovens, there were a lot of other technologies that happened, good things that happened along the way.
• 7:32
  Greg: And so with RBI, it, it just, it causes to become more systematic.
• 7:36
  Greg: Quit taking a lot of shortcuts.
• 7:38
  Greg: Not as much of this, but the important thing is that it using risk, it showed us where we really needed to be more accurate and where we could zoom out a little bit.
• 7:51
  Branden: Yeah.
• 7:51
  Branden: And, and some of those side things, right?
• 7:54
  Branden: That the side benefits that came out of this, those are things like damage mechanism reviews, those weren't those, those specifically coming out of the Chevron failure.
• 8:03
  Branden: And more recently we're seeing documents around CC, corrosion, control documents, CCDs and, and iws integrity operating windows.
• 8:10
  Branden: So, yeah, I think if we look back, whether, whether it's the qualitative or the quantitative method of RBI, as you said, the systematic method.
• 8:19
  Branden: Has been put in place to be able to help progress the industry, not just in risk analysis, but in overall documentation and identification of potential failure modes.
• 8:29
  Branden: I think that's pretty easy to see,
• 8:32
  Greg: so, yes.
• 8:32
  Greg: Yes.
• 8:33
  Greg: And the other thing it's caused us to do in the process is to document things.
• 8:38
  Greg: Okay?
• 8:39
  Greg: If, if you measured five mils per year, corrosion rate on this piece of equipment.
• 8:44
  Greg: How come you've got a six mil per year corrosion rate in the RBI program?
• 8:49
  Greg: If you're gonna do something out of the way, document it so that people can understand your logic.
• 8:56
  Greg: Gotta document the logic.
• 8:57
  Greg: We, you know, and, and I think the industry's still working hard on that.
• 9:01
  Greg: I think we still have a long way to go.
• 9:04
  Greg: But yeah.
• 9:06
  Branden: Yeah.
• 9:06
  Branden: Right.
• 9:07
  Branden: And, and, and to documentation, right?
• 9:09
  Branden: There's plenty of softwares out there now that help support with that documentation, and in the past it wasn't a thing, right?
• 9:16
  Branden: There's a whole new industry, software industry that's been developed specifically to facilitate this, not just at the ERP level, but all the way down at the actual, in inspection level and being able to provide those audit trails.
• 9:28
  Branden: So, yeah, I mean, when you think about it.
• 9:30
  Branden: Qualitative versus quantitative.
• 9:32
  Branden: I don't think it really matters when you look big picture wise, the benefits that have come out of just this one, single JIP I mean they continue to keep bearing fruit.
• 9:41
  Greg: I agree.
• 9:42
  Greg: Yeah, and, and I think another big benefit of becoming more quantitative is we now have metrics to help us get better.
• 9:51
  Greg: So as we get smarter, as our technology improves, as far as computers and software and all these things.
• 9:59
  Greg: We're getting smarter, we're able to crunch more numbers.
• 10:03
  Greg: But to go back to your original question, what if we just started today?
• 10:07
  Greg: So we have AI today and I just started out with like BMRs.
• 10:10
  Greg: Of course, iws are part of that.
• 10:13
  Greg: And the other thing is piping circuit.
• 10:16
  Greg: Years ago, people did, they called it piping ization, but it was just p and ID line numbers.
• 10:22
  Greg: And then that grew into a ization as we began to appreciate the thermodynamics of damage mechanisms, you know, what causes them.
• 10:30
  Greg: And so then we started going into ization, and this was the RBI process was driving all this.
• 10:36
  Greg: The only other thing that could have driven this is with, in my opinion, is if an owner operator would've had one of these other systems, our complimentary systems, like you're talking about, a inspection database management system or an IDMS is if I don't have RBI.
• 10:52
  Greg: But I've done a DMR and now I've identified my corrosion circuits, which are supposed to be groups of piping of pretty much the same damage mechanisms, damage rates, so forth and so on.
• 11:03
  Greg: Materials I, if I do that, then let's say I don't have RBI and I've got my IDMS and my IDMS is saying.
• 11:12
  Greg: Oh, on this TML, on this circuit, or CML, it's six mils per year.
• 11:16
  Greg: On this one, it's five mils per year.
• 11:18
  Greg: On this one, it's six mils.
• 11:19
  Greg: On this one, it's nine mils per year on this, but yet they're all part of that same circuit.
• 11:24
  Greg: And so a couple of the challenges the industry had been facing is, number one, people were thinking they could become more efficient without losing any effectiveness by using circuit average corrosion rates.
• 11:37
  Greg: In, in these IDMS systems and then managing that circuit based on the circuit average rate instead of each little individual TML rate.
• 11:47
  Greg: Does this make any sense?
• 11:48
  Greg: And, and, and that exists regardless of RBI, you know, it's like what corrosion rate am I, and it's, and it's also about how do I manage the work?
• 11:59
  Greg: How do I manage the system, right?
• 12:00
  Greg: How do I manage the work?
• 12:03
  Branden: It's, it's, it's interesting when you start thinking about how everything's progressed all the way through.
• 12:09
  Branden: Even, even there, you, you, you called'em TMLs and, and switched up to calling'em CMLs.
• 12:14
  Branden: All of this has, seems like it has stemmed not just from the, the inspection standards, but from the, the added pieces that we've been trying to do to make ourselves better.
• 12:23
  Branden: So.
• 12:25
  Branden: Well, Greg, I wanna thank you for your time today.
• 12:27
  Branden: We'll catch up again.
• 12:28
  Branden: Probably ne we'll catch up again next time.
• 12:29
  Branden: All right,
• 12:30
  Greg: sounds great, Branden.
• 12:31
  Greg: Good talking, everybody.
• 12:32
  Greg: Alright, anytime.
• 12:33
  Greg: Thanks.
• 12:34
  Greg: Have a good one.
• 12:34
  Thank you for listening to Upon Further Inspection, a Mechanical Integrity podcast.
• 12:39
  This episode was co-created by inspection, hearing, and Corr solutions.
• 12:43
  Our producers are Nick Schmoyer, Jocelyn Christie and Jeremiah Wooten.
• 12:49
  This podcast is for informational purposes only and does not constitute legal or professional's advice.
• 12:56
  Listeners should seek their own qualified advisors for guidance.
• 13:00
  If you enjoyed this episode.
• 13:01
  Please join us next time wherever you listen to your podcasts.
• 13:05
  Until then, stay safe and stay informed.