Episode Transcript
Transcript
01:30
on this episode of pipeline things 666 eschatology and the end of all things to come as we talk
01:37
about pipeline dents if you don't know what TMS means then you're not in the know and you're gonna
01:42
need to listen to this episode thanks for joining us all
02:00
right welcome to this episode of pipeline things when we are continuing our conversation I'm guys
02:06
I'm enjoying the series that we're heading into because I love talking about the stories that
02:14
define us as an industry and i like getting into some of these questions where people are always
02:18
like where did that come from and sometimes people kind of halfway know sometimes there's the rumor
02:23
mill out there but we don't always know this is a missed opportunity i'm gonna jump in here like if
02:30
if we are careful with how we do this podcast we could just make stuff up like sarah's told us
02:35
we're starting to have enough expert evidence on google's seo that we could just like redefine what
02:41
history was. When do you think the first time we'll get referenced in a publication will be?
02:46
According to Pipeline Things, June 5th of 2026. That'll be funny,
02:52
but I can't self-reference, that's for sure. We can't reference, that's too much. Guaranteed, AI
02:57
is already paying attention. Did our guest just jump in? Did you already let him in?
03:02
Yeah, I did. AI is already paying attention. They're going to remember me. Wow. That is also a
03:08
first. I want you to know. We always introduce the guest. The guest is surprise. It's like self
03:13
-excavating pipe. We have a self-introducing guest. So you know when I was thinking about today's
03:19
topic, Mike, now you're a part of this. Yeah, now you're part of this. Welcome. This is a first. I
03:22
have to fly a lot. People always ask me how much I travel. Probably like, what, Chris, probably 20,
03:28
at least 20 trips a year I'm flying. I think I could do tours through IAH, the airport in Houston
03:33
now. You know the one rule that I've never understood? I mean, it sounds like it's in theory is you
03:39
have to have your phones in airplane mode during takeoff and landing. Do you know why that is?
03:46
Somebody out there is going to show me. I'm sure somebody has read. Oh, the Miss Producer is
03:50
shaking her head. You want to fill us in, Miss Producer? Go ahead. Come on. You can't hear me. I
03:55
know, but I can relay what they're saying. Come on. This is your moment to be on the show.
04:00
Heard is a key word.
04:05
Oh,
04:09
so just so the audience. caught the myth of why this exists. It's because as the plane takes off
04:16
and you get further from the tower, the tower signal gets stronger, which interferes with the aim
04:22
playing. I think it's psychological. It's just getting everybody prepared to say, we're about to
04:26
take off, pay attention to the flight attendant and hear the safety instructions. Mike, you want to
04:31
take a stab at why, why they make you put your plane in airplane mode during tag off and landing.
04:36
Do you have a thought on this? Because they can.
04:40
Wow. I think the audience got a glimpse of the bike. I like that. That's called a flex.
04:47
So I will say, Mr. Producer, I think categorically reject your answer.
04:53
I don't buy that at all.
04:55
I've heard something else, which was, yeah, interference, but not related to that. But I think they
05:00
disproved all of that. Anyway, my whole point is nobody knows where this rule came from.
05:06
It's a lot like today's topic. We haven't talked to the right people. That's what this podcast
05:09
series is for, is the untold stories of why we do things,
05:14
bro. This is why we bring people on the show, like Mike and all the other great people that we've
05:19
had on. So just before we get into this, and we're going to introduce our guests in a moment. So
05:24
we're not going to make something up as what you just said? I think we already, about why the
05:28
regulations are? We just make something up. I mean, you can, because you know what we're going to
05:31
talk about. So audience, I'm going to ask you to think about. The dent regulations we currently
05:36
have. What have you heard about where 6% deep comes from?
05:43
What have you heard about where the original strain limits in B31.8 came from?
05:50
You've probably heard a bit of myth and a bit of fact. Today,
05:55
we're going to be talking to our esteemed guest. That's the word I've learned you have to use. You
05:59
want to make people forget about themselves, you say esteemed. Esteemed. Our esteemed guest, Mike
06:04
Rosenfeld. So, Mike, I think you've introduced yourself before on the podcast. You're a return
06:09
guest, but you got any words you want to introduce yourself to the audience before we jump right
06:14
into this topic?
06:18
Sure. Mike Rosenfeld, chief engineer with RSI Pipeline Solutions, RSI. It's right there.
06:25
I saw you came, Brandon, on the show. I liked it. It's a nice shirt, by the way. It is. I favor
06:29
white. I don't know if you've noticed. I really like white shirts. I like your shirt. I like the
06:33
color. At the last minute, I decided to put this shirt on. Before that,
06:38
I was wearing a white T-shirt with the Surf Ohio logo on it.
06:46
Surf Ohio is... colossal joke because I was going to say I have a feeling like Ohio versus
06:51
Michigan. I was like, this has to be a rivalry. No, no. Surfing. Yeah, I know. Ohio's a landlocked
06:58
state. That's what I said. This has to be a rivalry joke. No, no. No, it's just a joke joke.
07:04
Yeah. Ohio does have the North Coast.
07:11
Can you surf? Wait, what?
07:16
I was like, wait, hold on. Do the waves actually get big enough there? I didn't pause and think
07:20
around. I was like, am I making a stupid comment?
07:23
Actually, they can get some pretty big waves during a storm, like in the winter or something. I
07:28
don't think you'd want to be there. So, all right. Well, Mike, today you're on to help us sort
07:35
through the... The history of dent-based criteria in the United States,
07:41
hoping to set the record straight. I know some of this, some of it, though, you shed some good
07:46
light on in our pre-meetings. And so what I'd like you to do, if you don't mind, is for our
07:52
audience that is out there, you guys are familiar with it. There is a depth-based limit of 6%
07:57
that's been around for a long time. and then two percent on seam welds um can you shed some light
08:06
mike on on on your opinion where where where is that come from how did we get there into that
08:12
portion those those criteria in the code yeah so uh the depth-based criteria do have a long
08:20
history in the asme pipeline standards b31.8 for natural gas and a a different history arc for b31
08:32
.4 for hazardous liquids if we go back to the first b31.8 standard in 1955 the depth criterion was
08:44
for new construction uh one quarter of an inch in pipe operating at 50 percent of smys or more and
08:53
that's a that's a pretty tight criterion so in 1963 that was revised to a quarter inch or two
09:04
percent of the pipe diameter in pipe larger than a 12 inch nominal pipe size and pipe operating at
09:12
stress levels of 40 percent of smys but there were no maintenance specific criteria at that time
09:25
Wait, let me make sure I understand that correct. Hold on. So 1956, this is well before OPS.
09:31
This is well before 1955, well before what we call, what do you call it?
09:36
IMP 1.0. IMP 1.0, right. So I've heard, I actually, I might be, I'm going to say something
09:41
ignorant. I think a lot of people think 6% came about in the code, but it was clearly predated in
09:48
B31.8, 1956. And you said something that I had to stop. Because in my mind,
09:53
I was already wondering, which is you said there was no maintenance portion. So this was just the
09:59
design portion of the code? Well, the construction portion. Construction portion. You were supposed
10:05
to visually examine the condition of the pipe and the condition of the coating.
10:12
Examine the condition of the pipe before you coat it, because a lot of coating was done in the
10:16
field. Then examine the condition of the coating and the pipe. before you lower it into the ditch
10:22
and bury it so it's a construction standard and the also you would want to find this damage before
10:30
the construction contractor folds his tents and disappears so um so in 63 it became two percent of
10:40
the pipe diameter in 12 inch in pipe larger than 12 inch nominal and the same criterion was
10:49
introduced in B31.4 in 1971,
10:54
their first editions didn't talk about that. So that was introduced in 71 for hazardous liquid
11:02
pipelines. So B31.4 had no construction-related dent depth criteria until 1971 when they took it
11:09
from B31.8. Well, we say they took it. Maybe they did it on their own. No, I think they, well, who
11:15
knows? They came up with exactly the same criterion. Although their limit was pipe operating at
11:20
hoop stress of 20% SMYS instead of 40.
11:26
And then for the gas code, for maintenance purposes,
11:33
the code stated that injurious gouges, grooves, and dents would have to be removed if you discover
11:41
them. But they didn't define what was injurious. And that persisted with...
11:48
the 1992 standard but then in 1995 the the code said smooth dents do not require repair unless they
12:02
contain stress concentrators or affect girth girth or seam welds which makes them not smooth dense
12:08
or the depth exceeds six percent of the diameter so that was in 1995.
12:14
wow and um so
12:21
My recollection, because I had just started getting involved with B31.8,
12:28
maybe around 91 or 92, I distinctly recall a discussion at a committee meeting,
12:37
so if it came out in the 95 code, this might have been around 93 or 94, about doing a survey of in
12:47
-line inspection. vendors at that time to determine what size dent their tools could travel past
12:59
without basically coming out in pieces such that you couldn't get a reliable assessment downstream.
13:07
So these are caliper and MFL tools like we think of today. This is like the age of tubascope? Yeah.
13:12
Yeah. I mean, there were, look, there was back in that day, there were. Some of the same players,
13:19
you know, there was Enduro, there was TDW with their caliper tool. There was Tubiscope and Vetco.
13:28
At one time, they were separate companies, but then they merged. And I'm not quite sure when that
13:33
occurred. And I believe that, and PII was also in the mix then.
13:39
And another company called Nausco, which became BJ Services.
13:46
And Nausco, actually, they had an interesting deformation tool that used sonar to ping the inside
13:56
wall and get an echo for essentially geolocation of the deformation.
14:03
And then on the liquid side, there was also Pipetronics,
14:09
which I think was UT-based.
14:12
But some of those companies aren't around anymore. Bought, merged,
14:18
absorbed. Different ILI companies. And I think it was an informal survey because,
14:27
look, there are maybe half a dozen companies to call up. So somebody knew somebody and talked to
14:33
them and got a variety of answers. And the lowest or most limited case was a dent 6% of the pipe
14:43
diameter. That's what went into the code. It makes us think, Mike,
14:49
that that was like a constraint for the pig, right? So that the pig could traverse. So then that
14:54
makes me think that the logic was then, well, then the pig described it. Well, if the pig can
14:58
describe it, then at what point do you say it needs a repair? So I get the logic for saying this is
15:04
an issue, but that seems more for pig ability, not necessarily for integrity. Yes, that's correct.
15:09
It's a pig ability issue. There was testing that had already been done.
15:15
Many, many years earlier, I think the first kind of well-thought-out set of tests was done by a
15:25
pair of engineers, Balanos and Ryan. They were working for Columbia Gas Services in Columbus,
15:37
Ohio, and they performed pressure testing on samples of pipe that they,
15:44
you know, uh introduced dents and of specific proportions and depths and some were smooth and some
15:50
were some had gouges in them uh you know and put in caps and uh pressured them up and so what they
15:59
found was that that really deep dents, the pressure would just push the dent out and out.
16:04
And as long as it didn't have a scrape or a gouge in it, it didn't really affect the burst pressure
16:11
of the pipe. So, you know, then that was in 1958,
16:17
59, sometime like that. And that was published in an oil and gas journal article.
16:24
Then in 1960, Bob Iber and George McClure who were involved in a lot of fracture control studies
16:32
that came maybe a few years later. So they were at Battelle.
16:37
They did a study for the American Gas Association, the Pipeline Research Council,
16:44
which was those two were tied together at that time, where they did a lot more testing,
16:51
looked at a lot of broader range of variables.
16:57
And what they found
17:00
was that dents as deep as 9% or 10% of the pipe diameter didn't,
17:05
as long as they were plain dents, did not affect the burst pressure of the pipe. So that 6%
17:12
criterion is not a direct integrity-related parameter.
17:19
It's, as you said, Chris, it's for pickability. So I wonder, Mike, so this is around the time when
17:26
Kieffner and Fowler, from stress we're doing some of the early testing work too and that went into
17:35
1156 if i recall that was the api standard on rock stents that like appeared and then was retracted
17:41
um do you know the depth limits of that test because there's some reason in my mind i have six
17:46
percent as the depth limit from that test too that i thought they worked in so i'm wondering if the
17:51
six percent limit which then appeared in 1995 and is around the time that keefner and Fowler doing
17:58
their test at stress if it worked its way and then limited their test to 6% incidentally too.
18:06
Perhaps. I don't recall that.
18:14
And I'm not sure which came first, now that you mention it. I don't recall the publication date of
18:22
the 11-56, but it might have been around then.
18:28
And they were, to be honest, I think they were more concerned with the process of dents with
18:38
scrapes or gouges and validating the process of buffing those out to a smooth contour.
18:44
So there was an equation in the Canadian standard at that time that was a sort of a conservative
18:53
version of... the original b31g equation but nobody really knew where that came from and so uh api
19:06
or api funded a study between carried out by stress engineering and keithner and associates to at
19:15
least try and validate that at which they did so they showed that
19:23
that you could remove the mechanical damage by buffing it out within limits. And they defined that,
19:28
hey, look, if the damage is really bad, like if you have to remove more than 40% of the wall,
19:34
that can't be your only repair because there's evidently invisible damage that just buffing it out
19:40
doesn't take care of. But that's a different issue. So what I'm wondering still, though, is,
19:45
you know, like one of the ongoing debates on 1183, not ongoing, but one of the things we run into
19:50
is that currently. if you have a dent with metal loss,
19:55
we assess the burst pressure from the metal loss using B31G, right? Which is consistent with that.
20:01
You keep referencing the B31.8 injurious dent portion. And it basically says that for dents up to
20:06
6% deep, go ahead and assess the metal loss according to B31G and then assess the dent to the
20:13
other criteria. Yep. And that has effectively established a limit of 6% for dents with metal loss.
20:19
Because we have no code guidance that says if it's 7% with metal loss,
20:24
you can still assess the corrosion using B31G. And it put a hard limit in 1183.
20:31
And it's worked its way even still today into some of the dent ECA procedures. Like if we have a
20:36
dent with metal loss and it's a 7% deep dent with metal loss, we have no test data,
20:43
no code data to support analyzing greater than 6%. Well, actually,
20:48
it's funny you mentioned that. In the mid-early or mid-90s, say sometime between 1990 and 1994,
20:57
the EPRG had a recommendation of 7%. Did it? Yeah.
21:04
Now I'm going to have to go look. Now you made me go look. I feel like I'm going to actually have
21:07
to follow up with these questions in the audience. You know, I feel like I'm actually going to need
21:12
to ask. Kiefner when we talk to him, say, hey, did y'all test greater than 6%? Get the answer to
21:19
that on that 1156. And then I'm going to have to go look and see if EPRG had a recommendation of
21:23
6%. See if somebody else out there had a recommendation for greater than 6%. Okay. But here's the
21:29
other thing related to something that you mentioned, which is corrosion in a dent. And B31.8
21:35
allows you to evaluate corrosion in an otherwise...
21:42
plain dent the same as that that exact same corrosion in an undented pipe correct so does the
21:50
current edition of asme b31g which got totally rewritten in 2009 and has been successfully updated
21:59
every couple years since and so uh and that was based on those same tests that were done carried
22:08
out by stress engineering and keefner and associates where they looked at uh just plain metal loss
22:16
not where the metal loss is removing a gouge okay so there there does seem to be a difference
22:23
between metal loss due to corrosion versus metal loss where you're trying to buff out true
22:31
mechanical damage that has you know that traumatized microstructure at the grouch and and that
22:38
could have a crack and and stuff like that um so
22:43
Yeah, that's where those come from. One last thought. We spent a lot of time talking about dents
22:48
with metal walls, right? But the seam weld portion is still 2%. Is the seam weld portion with 2%
22:55
just purely carryover from the 1963 definition for construction and it's just never been updated?
23:05
No. I do also recall a discussion at B31.8 about...
23:13
the problem with indentations affecting seams and girth welds.
23:22
And seams and girth welds don't necessarily have the same tolerance for strain as smooth metal.
23:31
And so I think at least for the construction portion,
23:44
The feeling was that, of course, the construction part for the gas code didn't change. It's still
23:49
2%, period. For operations and maintenance,
23:56
you can...
24:00
You can invoke instead a strain criterion, which we haven't talked about yet.
24:07
It's not as liberal as for plain pipe,
24:13
but you can go to a higher strain. But it still recognizes that growth welds don't have the same
24:20
kind of strain capacity. But somewhat arbitrary in terms of the limit that was put on it.
24:26
I guess that was my thought. It could be. And of course,
24:30
The B31.8 code allows a pipeline operator for maintenance purposes to come up with their own
24:40
criteria for whatever they're doing based on an engineering analysis.
24:46
The catch-all. I forget what it is. We debated that before. There is that catch-all phrase.
24:52
Okay. This is a good point. We're going to pause. So audience, you heard it. 6%.
24:57
I think many of us expected that it was a tool passage issue. Mike confirms it is indeed a tool
25:04
passage issue, but dating way back and borrowed even from earlier versions. So we come back.
25:10
We're going to talk about maybe a little more of a fun topic when it's near and dear to my heart,
25:15
which is where does 6% strain come from? First guess from Chris was 6% date,
25:21
6% strain. Why not? You'll find out when we come back.
25:29
All right. Welcome back, audience, as we continue our conversation on the hidden truth for dents
25:37
and pipelines with our guest, Mike Rosenfeld. So, Mike, you closed us out at the end of the last
25:41
section with the origins for the 6% deep criteria. in B31.8.
25:48
And you mentioned that, again, the depth origins come out of the construction portion of the code
25:53
all the way until 1995, which is when 6% gets introduced. And per your recollection,
26:00
it seems to be based on an obscure survey of ILI vendors and depth-based. But that's not all the
26:07
story, right? There's something else that you didn't share with the audience? Yeah, there's a
26:12
little bit of a plot twist, which is that... 1986,
26:18
which is nine years earlier, the B31.4 standard changed their dent depth acceptability criterion
26:27
from 2% of the diameter in 12 inch or larger pipe to 6% of the diameter in pipe larger than 4
26:38
inch nominal pipe size.
26:42
So this was, as I said, nine years earlier. And I have tried to investigate this with,
26:50
you know, people who currently are or formerly were involved with B31.4 going back into the 90s or
27:00
in one case who was a member of the committee at that time. And none of them have been able to
27:06
state what the basis for that change was now maybe they did maybe the b31 four committee did their
27:17
own survey of ili companies and they were concerned about pickability and maybe b31.8 actually
27:23
picked up on that from them um i don't i don't know but uh they may also have been uh um tired of
27:34
of i don't know cutting out Yes, there were only 2% of the pipe diameter. I'm arguing that
27:40
somebody got some construction influence on the code at that time. That's what I'm thinking too.
27:45
Consensus came from consensus. We need to stop fixing these. I will say this much. The ASME
27:50
committee, at least today, Mike will probably defend me on this. I feel like the ASME committee is
27:54
pretty guarded in their ability. They don't make rapid changes.
28:00
They don't succumb to external pressure too often in my opinion. That's just my opinion of the 31
28:06
.4 and 31.8 committees.
28:09
On the code committees, we're kind of lazy and don't really want to make changes. Is it lazy?
28:17
Is it lazy or is it more? Yeah, it's a combination of if it's not broke,
28:24
don't fix it or being conservative. And also...
28:29
Things exactly like this where we don't know why it says what it says. So you got to be careful
28:33
about making changes because. So do we just need another big dent empirical testing program?
28:39
That's what this seems like. I think we just need another one. Maybe somebody says we need to do a
28:43
JIP and just test a whole bunch of this stuff. Don't go there. And then we can debate between six
28:47
or seven, right? Six, seven. I went ahead and asked.
28:51
I went ahead and asked Gemini. And Gemini blames it on a buffer.
28:57
That, you know, what Mike said is that dense less than 10% don't affect a burst pressure of a
29:02
pipe. So they just apply to safety factor. You do like nobody in the audience knows what you mean
29:06
by Gemini. Because we all use Copilot and ChatGPT. I'm sorry. So why would you do that? Like nobody
29:10
uses Gemini. But you know what? So Gemini is Samsung's version. But you know what nobody talks
29:15
about? Apple's version. So thank you very much. All right,
29:22
Mike, this is a more fun conversation for me. The origins of strain. And the limits in B31.
29:29
Let me guess. The limit is 6%. Yes, it is. Of course it is. Why not? Why not?
29:36
The mark of the B666.
29:41
Because they're so bad for you. Dents are just so bad. Oh, my gosh. That might be the first
29:45
religious reference on the podcast. That's so funny.
29:49
No, so that our audience is familiar, the 6% limit. So going back to the first editions of P31A,
29:56
and everybody knows this pre-2018, the code allowed you to evaluate dents with strain starting in
30:04
2003. And the limit was originally 6%. Right. Voila.
30:10
Yeah, well, I mean, first of all, I think the question is, why are we even looking at strain?
30:16
That's fair. All right, take us back. Right.
30:20
and who was doing it and i there were a lot of people uh involved with that and uh i i believe that
30:30
the first so first of all why strain well because uh you know a ductile material and most most line
30:40
pipe actually is ductile even older line pipe materials true are are ductile and a ductile material
30:48
nails from
30:51
TMS, too much strain.
30:56
Hold on. When you came on for PPIM, didn't you have your own acronym back then too?
31:03
You had a hardness. That's right. He's trying to get away from it. Now he has a TMS.
31:09
I love this. Too much strain.
31:14
Well, prove me wrong.
31:20
In around 1994, Dave Warman, he works at Crest now.
31:26
He was working at Iroquois Pipeline, Iroquois Gas Transmission System,
31:31
rather. And the Iroquois line was faced with some kind of big,
31:39
dense challenges. Yeah, left over as a result of construction,
31:45
you know. the the contractor didn't do a super job of preparing the ditch bottom there were some
31:51
also some big rocks in the backfill and they tested to a high stress level so you know when you
31:58
have the weight you've got a pipe sitting on a rock and then you add the weight of a meter of dirt
32:05
over it and then you add the weight of water in the pipeline and then you take it up to a hoop
32:10
stress of you know 100% SMYS or higher you're going to have areas where the around that contact
32:20
there where you exceed the biaxial yield strength of the material during the test and the pipe is
32:26
going to wrap itself around that rock so you end up with big big honk of dents and so they had some
32:33
of these things and so um dave and uh another guy named jim justice who was a trans canada engineer
32:44
iraq was actually a trans canada project they were at that point starting to think about strain as
32:54
an alternative for accepting dents rather than the depth alone and what their criterion was uh a
33:05
three percent a strain of three percent and what they would do is they uh developed a process in
33:13
the field for uh examining dents using a like a contour gauge to uh and then tracing that profile
33:22
onto a piece of paper or just comparing it against a say a curvature pre-drawn curvature did that
33:29
could indicate what this what the strain level was because if you take a a straight piece of pipe
33:35
straight a straight plate and you bend it to a curvature you're introducing strain all right so
33:41
there's a relationship between the curvature and the strength uh and their strain criterion was
33:47
three and that was based on um uh and uh,
33:55
an attempt to be conservative against a strain limit of 10%. Where does 10% come from?
34:01
Well, when you're doing a qualifying, uh, a welding procedure or a welder, you do the root band and
34:07
face band test. And for typical ranges of pipeline wall thickness, that's a strain that Ben test is
34:14
a strain of around 10%. So they wanted to go to 3% of that.
34:21
Uh, and they wanted. and what they were doing too is they were only measuring strain either in a
34:27
longitudinal axis or a transverse accents they weren't trying to combine them or anything like that
34:33
and the strain was purely based on change in curvature and i got involved with that for a couple of
34:42
reasons one was that i was able to provide dave with a a an asme code based justification for that
34:55
three percent and because they wanted that for to try and persuade the dot that these dents would
35:05
were going to be okay and the relationship was that nowhere in the code at that time uh was uh
35:14
strain levels up to three percent prohibited and or even higher for that matter didn't talk about
35:21
it
35:24
And when you look at cold field bends, the allowed bending radius for pipe came to strain levels
35:34
typically between 2% and up to 3%. And nobody gets it all bent out of shape about straining your
35:44
pipe up to 3%. So if it's okay for a cold field bend, why can't it be okay for an indentation?
35:52
that was the reasoning there then the other reason i was involved too was that the presence of
35:59
these dents and the rocks you know in the in the ditch and stuff like that led into some
36:08
contentions that the pipeline was not constructed in accordance with its uh you know for
36:14
construction permit and so there was a lot of other uh there was a big sideshow associated with
36:23
that. So at this time, this originates because they need an alternative criteria other than depth
36:31
for a pipeline that has a lot of dents. And what they resort to in the field is uniaxial,
36:36
one axis or the other, comparisons of curvature using a contour gauge. You know, it's fun. I
36:41
actually got one of these in my email box, I swear, yesterday. I have not seen a legitimate one.
36:46
And I got a dent field report with a carpenter's contour gauge. And they were comparing the
36:52
curvature to printouts on a page. And I was like, oh, my God. I've talked about this but never
36:57
actually seen it. In all fairness, it was, I'll be honest with you, it was somewhat effective.
37:04
Better than some calculations that I've seen from optical scans, I will say. So this goes,
37:09
and y'all need a basis. Y'all's first rational starting point for a basis is 3%, which you
37:13
rationalize as being we've been bending pipe this way empirically forever with no problems. All
37:19
right, so we're at 3%. How did you get above 3? Well, I know why you got above 3 because they asked
37:23
you to, but how did you get above 3? Well, right. So, you know,
37:29
a next step in that process was that I was approached by Rich McGregor,
37:36
who I knew, he was a former Iroquois engineer. I knew him during the time of that dense work at
37:44
Iroquois. He was now at Great Lakes Gas Transmission, also a TransCanada.
37:51
And he had a situation involving two big dents in the bottom of one of their pipelines where the
38:04
dents were quite deep, like 9% or 10% of the pipe diameter. The dilemma for them was that they
38:10
were out in the middle of a river. And it was going to cost a couple of million dollars to go
38:14
remediate that. And he asked a perfectly reasonable question. Do we really need to repair them?
38:22
And so that's where, you know, you start getting a little more serious about trying to evaluate the
38:29
strain in a dent. And so, I mean, Dave and I had both together and separately been going around
38:38
measuring, using a contour gauge to measure strain and, you know, in one axis or another.
38:45
But then this brought up. the question of doing that using what you can get out of a high
38:52
resolution geometry or caliper type tool and so they were using a high resolution caliper tool to
39:00
do that and so it it was using that data and some manual processing of that data it was possible to
39:12
show that even though these dents were quite deep diameter wise the actual strain levels were
39:18
pretty low because they were only, the strain levels were maybe two or 3%, just because they were
39:25
spread out over such a large area on the pipe. And so I think this really gave some permission to
39:35
contemplate large dents on a strain basis.
39:40
And the other thing too, is that some pipeline operators gas and li
39:47
situations where the pipe is sitting actually on a very sharp pointy rock. And so that acute
39:56
contact can lead to a very sharp indentation with a crack.
40:02
And it leads to a punching or sheer crack that leaks. And yet the dent depth is within,
40:12
say, the 2% acceptance limit. So what that showed was that depth and dent strain are not at all
40:26
related. Not good metrics. Yeah, and also right at about that time,
40:32
so this was, that was probably in 96 or 97, and probably in between the stuff that Dave and Jim
40:41
Justice were doing at Airquab and that, I'd been approached by Jim Cox,
40:48
who was a metallurgist with Plantation Pipeline. That's a hazardous liquids pipeline.
40:56
And Pat Porter, who was with Tubascope Vetco at the time,
41:03
to come up with an algorithmic approach to using ILI data to calculate strains.
41:11
So, you know, these things were all sort of converging at about the same time.
41:17
Around 97, you said? Yeah, which is actually two years after the B318 code had introduced the
41:26
strain limit.
41:29
So these things were all kind of happening at about the same time. And I think at,
41:36
I don't remember who introduced the issue or the question at B31.8.
41:43
It might've been me, but maybe not. Go ahead and take credit for it. Well, I don't want to take
41:49
credit if I didn't do it. I just don't recall.
41:52
But that led to
41:57
The 6% strain limit. Now, where did the 6% strain limit come from? That 6% is a pure coincidence
42:04
compared with a depth of 6% because those two things don't relate to each other.
42:10
But the 6% did come from me because I had done a survey looking at...
42:17
Another informal survey or a formal survey this time? Or you heard something from somebody once.
42:24
It was an informal survey of... foreign operator not the trans canada pipeline system yeah of dense
42:33
on their system actually i think it was the algonquin system um and uh looking at well that and
42:41
some other dense related information for example from colonial and a few others and uh so i looked
42:50
at dent depths and i calculated Dent strains and,
42:56
you know, sort of a biaxial strain, looking at the changes in curvature in both longitudinal and
43:02
transverse axis, and also the amount of stretching in the pipe that you get with a dent.
43:10
And looking at those strain levels and the smallest,
43:16
lowest apparent amount of strain where there was a crack was 12%.
43:27
You know, a typical ASME codes approach was, well, the worst that we saw was 12%.
43:34
So we'll just put a factor of safety of two on that and make it 6%. And voila.
43:40
That's where the 6% strain comes from. And you were looking at what data when you came up with the
43:44
12%? You're about to ask him how many statistical relevance? No, I'm not getting there, bud. You
43:48
know, like a dozen and a half dents or something. A dozen and a half dents.
43:54
Yeah. It's the half a dent that pushed it over the edge. The other question is, what was the data
44:00
from? Where did you get the data from? You said that was ILI data or that was field data? No, these
44:05
were field data or cutouts. Okay.
44:10
That's another source of randomness in that sort of analysis,
44:18
which is that... that you measure in the field or the dent that you measure in a cutout is not the
44:24
same as the dent that you measure with inline inspection. But nevertheless,
44:30
what we had were cutout bite. So the 6% strain limit came from about a dozen dents that you looked
44:36
at. About a dozen and a half. Not a half dent, but maybe about 18. No, that's not 4.5.
44:42
That's 18. Oh, I was wondering the same thing. I was like, it was a half dent. What's a half dent?
44:47
Now it makes sense. In the failure analysis, they cut right down the middle of the dent, so you
44:51
just had half of the dent. It was half dent cut off the field. I was so...
44:57
Thank you for clarifying that. So that's really funny. So there it is. Is that enough dents, Rhett?
45:02
I mean, you know, I... Hey, did at 1183 look at like a couple thousand dents? Maybe like 30,000
45:07
dents? Mike, since you were there. And I've got the strain criteria has been great in terms of
45:14
transforming gas integrity management for dents. And it's persisted now for well over 23 years.
45:20
And it's been refined and made improvements and all that. But I wonder if you now could go talk to
45:27
you in 2000 when you were probably projecting that for the code and said, Mike, do you know how set
45:33
in stone this 6% is going to become? Because 6% still persisted at CSA, still 6%.
45:42
No, I suppose that wasn't obvious to me. It never is,
45:48
though, is it? Well, we put this stuff into codes. That's the truth. I mean, working on code
45:52
communities, we do the best we can. And the funny thing is, we put things in there. It's just what
45:56
makes me realize is how...
46:02
difficult things become to change. I mean, I joke about this. Figure four is probably one of my
46:06
favorite sacred ones ever. Like, you know, there's so many things in the code. And so 12% was
46:11
based off of give or take 18 dents and lowest observed level of cracking. I have heard that before.
46:16
With a factor of safety of two. With a factor of safety of two. Not 1.4, not 1.37, 1.39.
46:22
None of our other standard safety factors. Just two. Got it. Yeah. That's great. Well, how about
46:26
welds? Where'd you get welds from then? Because you put welds at 4%. Right. And...
46:31
that's and you know what that was uh that was also a typical asme codes based decision um and i
46:41
remember sitting around a conference table talking about what should the number be and we were
46:46
debating it should be this that basically um uh um i went around the table and said what what
46:55
number would you be comfortable with for a weld and uh
47:02
The result of that was 4%. I'm just wondering if you did that today. I feel like 20,
47:08
22, 30. That would probably get shot down because we know that's really not going to work.
47:15
5% could have been. Whoever's bias comes to the table is what I'm getting at. That's the fun part.
47:22
So that's so much fun. B318 did in 2018, they did significantly raise. those strain limits because
47:28
a lot of pipeline operators were kind of squawking about. Yeah, they fixed the equation to some
47:35
other things. I mean, I never thought that the error in the equation was that aggressive. But one
47:39
of the limits is we never actually, we have yet to touch the weld limits portion of that,
47:45
right? And you and I have even chatted about that. That actually probably needs to be updated, but
47:49
we really don't have good data for it. It's the truth. And I don't know if we're willing to sit
47:53
around the table and ask that same. That would be fun. You know what? I might do that. My next 31
47:57
.8 O&M committee meeting. Just say, guys. I want to change 4% around the table. What are we all
48:03
comfortable with? Just yay or nay. Just anybody willing to do this. That's the first step. I should
48:07
totally do that. Oh, I know Evan's chair. He's probably, if he listens right now, he's probably
48:12
rolling over. So my goodness. Well, Mike, hey, this has been fun.
48:17
I hope, you know, if you think of some other good stories that the pipeline industry needs to
48:22
preserve, hopefully we've laid to rest some of the myths around dents. So now we know where.
48:29
where 6% comes from twice. It came from about 12% strain, half from about 12 and a half dents.
48:36
So there's this common theme of numbers here. It is. The next limit must be a multiple to six.
48:42
That's what I'm thinking. We also like even numbers is what it seems like. We don't like those odd
48:46
numbers. It just feels odd. Twelve and a half cents.
48:51
But Mike, no, seriously, it's been a lot of fun to our audience. Thanks for joining us. We're going
48:55
to continue talking to this series. And if you have some other, you know what, audience, throw it
48:58
out there. If you have some other good questions, you want to know where the heck does this come
49:01
from in the pipeline industry? Send your questions to us and we're going to see if we can get the
49:05
right people to answer it. But until you hear back from us again, we're going to sign up. on this
49:09
episode of Pipeline Things. We'll see you back in a couple of weeks. This episode of Pipeline
49:12
Things, executively produced by Miss Sarah Etier. Thank you very much to our guests from RSI
49:17
Pipeline Solutions, Mike Rosenfeld, and Form Marketing for Allow Us to Film at their venue. And of
49:22
course, my life partner and co-host, Mr. Christopher DeLeon.
