In this episode of the Pipeliners Podcast, Dr. Keith Leewis joins the show for a practical, experience-driven discussion on determining fitness for service in pipeline systems. Building on prior conversations, the episode explores how operators evaluate integrity, assess risk, and make informed decisions using established industry practices and standards. The discussion offers context and perspective without diving too deeply into technical detail, making it valuable for anyone involved in pipeline integrity and operations.

Determining Fitness for Service Show Notes, Links, and Insider Terms

• Dr. Keith Leewis is a pipeline safety and integrity expert with over four decades of experience spanning engineering, materials, operations, inspections, and regulatory compliance, supported by extensive industry leadership and academic research roles. Connect with Keith on LinkedIn.
• Fitness for Service (FFS) is an engineering assessment process used to determine whether a pipeline with defects such as corrosion, cracking, or thermal damage can continue to operate safely at a defined pressure.
• API 579 is the American Petroleum Institute’s Fitness-For-Service standard that provides analytical methods to evaluate defects like metal loss, cracks, and overheating and to predict current and future pipeline performance.
• ASME B31.G is an early industry standard developed to assess remaining strength and failure pressure of pipelines with corrosion-related wall loss using simplified geometric assumptions.
  • Modified B31.G is an enhanced version of B31.G that uses three-dimensional corrosion profiles to provide more accurate predictions of pipeline failure pressure.
• Remaining Wall Thickness refers to the amount of steel left in the pipe wall after corrosion, which is a key input in fitness-for-service and failure pressure calculations.
• Predicted Failure Pressure is the calculated pressure at which a pipeline segment with defects is expected to rupture, based on material properties, defect geometry, and operating conditions.
• ILI (In-Line Inspection) is the use of instrumented pipeline inspection tools (“smart pigs”) to detect corrosion, cracking, deformation, or other integrity threats without excavating the pipeline.
• MFL (Magnetic Flux Leakage) is an ILI technology commonly used to detect metal loss by measuring disturbances in a magnetic field caused by corrosion or wall thinning.
• Ultrasonic Inspection (UT) is an inspection method that uses sound waves to measure wall thickness or detect cracks, typically requiring a liquid couplant to transmit the signal.
• EMAT (Electromagnetic Acoustic Transducer) is an advanced ultrasonic inspection technology that generates sound waves magnetically without a liquid couplant, making it suitable for detecting cracks—especially stress corrosion cracking—in gas pipelines.
• Stress Corrosion Cracking (SCC) is a crack-type integrity threat caused by the combined effects of tensile stress and a corrosive environment, often resulting in axial cracks along the pipe.
• Axial Cracks are cracks oriented parallel to the length of the pipe, commonly associated with SCC and requiring angled ultrasonic or EMAT inspection techniques to detect.
• Time of Flight is the measurement of how long an inspection signal takes to travel through the pipe wall and return, which is used to calculate wall thickness or identify defects.
• Hydrostatic Testing (Hydro Test) is a pressure test using water to verify a pipeline’s ability to safely operate at a specified pressure by identifying leaks or failures under controlled conditions.
• MAOP (Maximum Allowable Operating Pressure) is the highest pressure at which a pipeline is permitted to operate safely, often adjusted based on test results or identified defects.
• SMYS (Specified Minimum Yield Strength) is the minimum yield strength required by specification for pipeline steel and is used as a baseline for pressure limits and safety factors.
• 1.25 Pressure Test Factor refers to the industry practice of hydrostatically testing pipelines at 125% of MAOP to provide a safety margin and validate structural integrity.
• Brittle Failure is a rapid, catastrophic pipe failure with little deformation, historically associated with older steels and a key driver behind modern hydrostatic testing requirements.
• Crack Arrestor is a design feature, such as thicker pipe sections or steel rings, used to stop the propagation of long-running fractures in pipelines.
• The PRCI (Pipeline Research Council International) is the preeminent global collaborative research development organization of, by, and for the energy pipeline industry.
• MAT-8 Software is a PRCI-developed engineering tool used to evaluate crack severity and predict failure pressure for crack-related pipeline defects.
• PHMSA (Pipeline and Hazardous Materials Safety Administration) is the federal agency within USDOT responsible for providing pipeline safety oversight through regulatory rulemaking, NTSB recommendations, and other important functions to protect people and the environment through the safe transportation of energy and other hazardous materials.
• Integrity Management is the systematic process of identifying pipeline risks, assessing threats, monitoring condition, and implementing mitigation measures to ensure safe operation.
• Probability of Failure refers to statistical methods used to estimate the likelihood that a pipeline defect will result in a leak or rupture over time.
• Pressure Reduction is a risk-mitigation strategy where operating pressure is lowered to extend the safe life of a pipeline segment with known defects.
• Corrosion Rate is the calculated speed at which metal loss occurs over time, typically derived from comparing ILI results from multiple inspection runs.
• CO₂ Pipelines are pipelines transporting carbon dioxide, which present unique integrity challenges due to CO₂’s thermodynamic behavior, decompression characteristics, and extreme cooling during release.
• Charpy Impact Test is a laboratory test used to measure material toughness and resistance to brittle fracture, particularly at low temperatures.
• Supercritical CO₂ refers to carbon dioxide in a state where it behaves as both a liquid and a gas, complicating pipeline design, operation, and failure behavior.
• Outside Force Damage is mechanical damage to a pipeline caused by external impacts, such as excavation equipment, that can alter geometry and complicate integrity assessments.

Determining Fitness for Service Full Episode Transcript

Russel Treat:

Welcome to the “Pipeliners Podcast,” Episode 419, sponsored by EnerSys Corporation, providers of POEMS, the Pipeline Operations Excellence Management System, operations and compliance software for the pipeline operator to address safety program management, control room management, and field operations. Find out more about POEMS at enersyscorp.com.

[background music]

Announcer:

The Pipeliners Podcast, where professionals, bubba geeks, and industry insiders share their knowledge and experience about technology, projects, and pipeline operations.

Now your host, Russel Treat.

Russel:

Thanks for listening to the Pipeliners Podcast. I appreciate you taking the time. To show the appreciation, we give away a customized YETI tumbler to one listener every episode. This week, our winner is Keith Leewis, also our guest. Congratulations, Keith. Your YETI’s on its way. To learn how you can win this signature prize, stick around till the end of the episode.

This week, we’re going to speak with Keith Leewis, again, and we’re going to talk about determining fitness for service. Keith, welcome back to the Pipeliners Podcast.

Keith Leewis: Thank you very much for having me back.

Russel:

I’ve been enjoying our conversation. For the listeners, this is our third episode. We talked about the fundamentals of metallurgy, then we talked about issues around old and new fuel, and now we’re going to talk about determining fitness for service.

Keith, I’m going to ask you to do this again, but if you would, give us a quick intro. Tell us a little bit about yourself and how you became a metallurgist pipeliner person.

Keith:

Thank you very much. When I was in high school, I decided that I didn’t want to work like some of the rest of my class, so I went to school. In the school I went to, we got to pick the one during first year, and metallurgy seemed to answer most of the physics questions that I wanted, so I ended up there.

After that, I ended up making steel. Then I had the opportunity to go back to school and do some more research. I ended up getting a doctorate and becoming a professor there for a little while in steelmaking.

Then, also, the economy being what it was, I had to convert from 300 tons down to small grams to become a welding engineer. I started a welding course there for our undergraduates as well.

Then, I worked in the industry, worked for TransCanada building pipe. Also doing not only as a metallurgist and welding engineer, but helped start with their integrity. Then, I lucked out and worked for the Gas Research as we talked to what’s now PHMSA about integrity and helped write B31.8S, which is the integrity standard for construction of natural gas.

Then, as a consultant, I had the opportunity to work with INGAA and other folks, working between the regulatory and the science end of it. Then, I’ve had the opportunity to be involved in a whole variety of things with ILI and also corrosion, and helped with the new standards and the other API standards, and those kinds of things.

I’ve been lucky to see a whole variety of the integrity background, including the time we blew a 36-inch pipeline up twice in order to calibrate our thermal radiation models.

Russel: [laughs]

Russel: That’s always the funnest research is when you get to do destructive testing. That’s always the funnest part. That’s the polite word for blowing things up.

Keith:

Yeah. We had a viewing site for not only the Europeans and others who came over to watch it but for the local retirement community, and sold hot dogs there.

We cut the pipe with dynamite, and we used a lot of fireworks to make sure that the fire ignited. Otherwise, we would have wasted a couple of million bucks not having any radiation to pick up to see what was wrong with our theories.

Actually, we had a problem because we forgot that when you do it at the beginning, it looks like a mushroom cloud. Our initial radiation wasn’t quite correct, so we had to change the software to take that into account.

Because when you release something like 10 atmospheres of gas, it all comes out in a big amount, and then it slows down a bit as the gas is starting to come out of the two ends of it. The big part is the top of the mushroom, and the other part is the stem on the mushroom. We were two miles away, and you could feel the heat.

Russel: Oh, wow.

Keith: That was one of the fun times I had, yeah.

Russel: Do you have videos of that? I’d like to see that.

Keith: I do not own the videos, but I do have a couple of pictures I took.

Russel: That would be interesting to see. That’s the kind of thing that if you’re lucky, you never see that other than in a research project, right? [laughs]

Keith: You are not supposed to see it, no.

Russel:

No, no. No.

[crosstalk]

Keith: We had to stop the trains and the airplanes as well, so their safety rules. Anyway, it was good. I hope other people who are listening have a chance to do other really nifty things during their own career, right?

Russel:

Right. That’s one of those significant events. You do not get to do that every day, but it can be a…What’s interesting about that stuff, though, is not…I mean, the doing of it is interesting, but the event itself, just preparing and setting up and executing the event is probably 10 percent of the work.

The other 90 percent of the work is all the analytical work you are doing afterwards to verify and understand what the data is saying.

Keith: It was probably more 50-50 because it took us about a year to do that. We had people down there who were talking to the locals as well. We had to run a separate line, put some more valves in, and things of that nature in order to do it. If you don’t do the preparation properly, you’re guaranteed to screw it up. It doesn’t matter what you do.

Russel: I did a little bit of work in college with the Texas Transportation Institute. At that time, they were doing different kinds of guardrails and crashing cars into them. They could set up one of those tests in about a week. Then they would spend a year just crunching the data and trying to figure out what it told them.

Keith: Yeah, but the actual crash only took a few seconds.

Russel: [laughs]

Russel:

Exactly right. Took about a week to set it up, about a minute to crash it. Then you had a year’s worth of analysis. Of course, they were working in a facility where they ran those things frequently, a little different than what y’all did.

Look, I wanted to talk about fitness for service. Once again, we are running down the rabbit holes, but it’s all good. I want to talk a little bit about fitness for service, talk a little bit about API 579. Maybe, if you would, tell us a little bit about API 579 and what it is, and how it is used in practice to operate and maintain pipe.

Keith:

Essentially, 579 is quite comprehensive. It covers a whole variety of things, from heat, overheating the pipe, because you can get thermal changes if you heat the pipe up and lose the strength. You are not allowed to do that.

A lot of the original work was done on remaining wall because you get corrosion, and you can lose wall thickness. Then there are other aspects in there for cracks, for example, so you can analyze for cracks as well.

There are different ways you can look at the information in order to predict how it’s going to perform. Then there are ways in there to estimate how it’ll perform in the future, so there will be continuing deterioration, either crack growth or wall loss, and so there are ways to estimate that as well.

Corrosion is probably the one that most people are familiar with. ASME B31.G is one of the original ones that goes back to about the ’60s.

In those days, when you are lying on your back underneath the pipe, you take out your tape measure, you can measure the length, and if you’ve got a board, you can put it in there and measure the deepest part of the corrosion.

With those two readings, you make a guess that it’s thumbnail-shaped, so it’s an eclipse, and you can determine the amount of wall that’s most likely missing. Then with tests that Battelle did, you could correlate that to the anticipated or predicted failure pressure.

Russel: This is even prior to the MFL tools being used, the…

Keith:

Yeah.

[crosstalk]

Russel: mechanism. You can actually look at the pipe and physically identify and measure the metal walls.

Keith:

Yeah, we still have to do it now, but now we can use various techniques to do it. You can run a laser beam over it so you can actually get the depth. You can use optical now, too. You’ll see the optical will give you the depth as well.

It takes the external diameter as a given, and it’ll give you the difference from what it measures. I have one that uses eddy current that will measure the gap between the surface and the steel underneath. You can get a profile that way. You can get a three-dimensional profile rather than a cord across it like a single line.

Then B31G, for example, became Modified B31G. It improved the predicted failure pressure. In this case, you’d use a three-dimensional grid of depth. As you go down the center line, you look off to the side and you pick the deepest part. That becomes the center line there. That gives you a better value. The Modified is what many people use now.

Then we did some work for TransCanada Pipelines and we came up with the probability. It’s called Psqr. It will actually run lines through the three-dimensional wall loss and pick out the one with the biggest depth and give you estimates of what the failure pressure would be. That’s probably closest to the actual failure strength because it has less uncertainty in it.

Those are the various ones that you will find now in the Fitness for Service or PI 79 for corrosion approaches. Most people do that in order to find out. They’re running ILI pig through it. They’ll get the internal or external corrosion, the wall loss, and they will calculate a predicted failure pressure.

Then the next thing they do is they’ll take the measurements from the ILI over time, because you tend to run them every five years or every 10 years because that’s what the regulations suggest, right?

Russel: Right.

Keith:

Actually, no, 10 years is what we suggested, 7 years for gas is what Congress suggested. We can go down another rabbit hole and I can explain that, but we better not.

[laughter]

Keith:

They’re not an engineering, they’re an estimate.

Anyway. You go back and look at the previous runs and you can estimate the amount of corrosion, amount of wall loss you have per year. Then you can then use that corrosion rate to estimate and use statistics as well. You can rate when it’s probably going to fail.

You should get your butt back there before then so you can decide how you’re going to fix it. Either cut it out or put a sleeve on it or do something else, like, drop the pressure.

Because, typically, we drop the pressure 20 percent, so to 80 percent. 80 percent is actually .8, or is 8/10, which is four-fifths. If you flip it upside down, you get five-quarters. Five-quarters is 1.25. 1.25 is what we test the new construction pipe to with the hydro test now.

In other words, if the existing pressure that you…surviving with that defect, is you set that as the new pressure test, and you can go up to 80 percent of that. It’s your new MAOP, and you can drop the pressure to that as well. Just a simple thought.

Russel: Yeah, no, it’s interesting, Keith. I’ve never really thought about this. I know enough about hydro testing to understand what it is. I know that it is the preferred mechanism of verifying serviceability for a certain pressure.

Keith: Maybe the only one.

Russel: Yeah. Well, yeah, it’s there are other approaches, but it’s the only way to actually verify — right? — because you’re putting that pressure on the entire system and anything that won’t hold that pressure will get identified.

Keith: Yeah.

Russel: Then I’ve always wondered how they came up with 1.25 times max yield as the hydro test standard. You just walked me through that, basically, is how they came up with that.

Keith:

Before that, it was 110 percent. Hydro testing started to become a tool that got used 100 percent, probably by 1970. During the ’60s, people started to use hydro testing. Before that, they used to test the pipe up to the MAOP and so, that way, they would check for leaks, so for well leaks and that kind of stuff. Then they started to take it up to 110.

If you remember back in the ’50s we had construction hydro tests that had a brittle failure that ran for something like seven miles, and so that was why they started the Pipeline Research Council International to try and figure out what to do with pipe so we didn’t have that happen again. That also started the hydro testing work as well.

The pipe, in effect, was brittle. Once it got started, it just kept running until it ran against something that would slow it down a bit. You can stop it by putting heavy wall in, which will then do it, or you can put composites or steel rings around the pipe as a crack arrestor as well. Anyway, that’s slightly different.

We went from 110 percent to 125 percent about 20 years ago. This SMYS is 72 percent is roughly they tested the steel to 90 percent in the mill and we wanted to have 80 percent of that as sort of a safety. We want it operated at 80 percent as a safety. So, .8 times .9 is .72.

That’s how we pulled that out of our safety thoughts. I was going to say something else, but it’s not nice to say that…

[crosstalk]

Russel:

Well, you’re determining safety factors or engineering factors.

[crosstalk]

Russel: It’s based on historical experience and some industry-level engineering judgment.

Keith:

Actually, back to the vintage pipe, there were time when we only tested the pipe milled to 60 percent of the SMYS, of the minimum yield.

Most of that, you won’t find. There’s probably some, maybe 10 percent of the mileage and service that goes back to that, but things changed over time. We went to higher and higher tests in the mill. Now, the mill takes it up to about 95 percent, but it’s only a 5- or 10-second burst. It’s not eight hours. Anyway, that’s our…

Russel: That’s different, too. Taking it up to that pressure versus holding that pressure is different.

Keith: Yeah, and we hold the pressure for a new construction.

Russel: I want to ask another question about this. Why water versus nitrogen? Because you can test with nitrogen as well, but why would you use water instead of nitrogen?

Keith:

Water is essentially incompressible, and gases are not. What happens when you find a leak? The pressure inside the pipe goes down rapidly with water, so there’s nothing driving that little hole or crack to grow the same extent.

If you’re pressurizing it with a gas, then the gas at the edge of the crack can still have the full pressure for quite a while because the gas has to escape out the hole, so that crack can run.

Plus, there’s a lot more energy in there. You think about compressed air in your tire, for example. You don’t want that thing to blow up on you when you’re filling it up. If you’re filling it with water, then that pressure, the driving force, would go away really quickly.

You have a lot more energy inside the tire or the pipe when you’re doing it in terms of how it reacts according to the physics.

Russel: That makes sense.

Keith: One of the problems with CO2, for example, is it doesn’t decompress as fast because it could be partially about the triple point. The crack will run faster, and you could lose more joints out of that failure with CO2.

Russel: CO2 is a strange animal. I’ve done a lot of work with CO2 in my career. It’s a strange animal.

Keith: I don’t know it that well.

Russel:

Usually, it’s supercritical when it’s in a pipeline, meaning it’s not really a liquid, it’s not really a gas. It behaves like a liquid, but it’s spongy like a gas. The other thing is it doesn’t exist as a liquid at atmospheric pressure. If you have a rupture and it goes to atmosphere, it’s going to convert to gas and solid.

The whole energy dynamics of how that works and what is caused because of that is different. It’s got a strange thermodynamic properties. It’s interesting, though. It’s interesting. It’s why it’s used a lot in enhanced oil recoveries because those properties help get the oil out of the rock.

Keith:

It’s a solvent in some ways.

[crosstalk]

Russel: Yes, exactly.

Keith: If I can put a plug in, we just put CO2 into the gas standard B31.8 that came out this year.

Russel: That’s interesting. I know API is about to release their standard for CO2. It addresses some of the unique issues around CO2 as it relates to operation safety.

Keith: They use it as a liquid. They pump it around as a liquid, so their issues are slightly different. That’s why they’ll be different in the standards.

Russel: You also get into issues around how the steel behaves and such, particularly if you have to vent, because you’re going to super cool that steel if you vent it.

Keith: And it becomes brittle. Even if you blow it out the stack, because you’re blowing the line down, you end up with dry ice all over the place and really cold material.

Russel: Minus-129 Fahrenheit.

Keith: We don’t test Charpy’s that low.

Russel:

We’ve talked about fitness for service, we’ve talked about hydro testing, and the rationale for why we test at the pressures we test at. We’ve talked about metal loss, how we find it, and so forth. There’s a lot going on these days, in particular with cracks and finding cracks. We ought to talk about that.

I want to ask a question. Would it be true to say that cracks are a bigger risk in newer steels than older steels because of the nature of their low ductility, or lower ductility, or is that a wrong way to think about it?

Keith:

Generally, yes. We did some work for PHMSA for one of their R&D ones. We plotted our results up as the X would be the predicted failure pressure over the actual failure. If it was 1, there’s a good chance it goes pop, and if it’s .5, you have quite a safety level.

The other one, we used PIR, which is the diameter times pressure, to give some idea of the extent of damage. When you plot that, the closer you are to the origin, the safer you are.

You would get leaks, and then we’d find a boundary in there where corrosion would rupture, but cracks didn’t. Then we found another boundary when you get further from the origin where both corrosion and cracks would rupture. We did that for both old and new steels as well.

It’s interesting. When I think of it, I think, for cracks it is like a C-clamp. The metal around it helps it, so you have to work a little harder to get it to go. With corrosion, essentially, you just got a thin section, and it stretches.

Russel: Aah.

Keith: OK?

Russel: Yeah. I am going to play this back and make sure I understand what you are saying. If I have a metal loss, basically, what’s going to happen is it’s going to act like a balloon, almost. It’s going to expand and eventually burst.

Keith: Yeah.

Russel: Whereas if I have a crack, it’s going to be more like two sets of rock that are impinged against one another. They’ll hold it until they slip, and when they slip, all that energy gets released at one time.

Keith: There’s still a gap between the two rocks, but the rocks are conjoined and are helping support it. It’s like a C-clamp.

Russel: Interesting.

Keith:

There’s a lot of calculations for the corrosion so you can get…There’s modified B31G, and every pigging company can do that for you. As an engineer, you can write your Excel spreadsheet, like I have. That kind of stuff.

For cracks, it’s probably best to use the MAT-8 software. That’s the one PRCI worked on with a number of lads in the field. That software is probably the best one to use to get failure pressures for cracks.

I actually don’t own a copy of it, but lots of other people have. That was the one we used when we plotted it out. That paper, by the way, is in last year’s PPIM, some of the diagrams and stuff. When we plotted it all out to give you…We were surprised, too, because we thought cracks were going to fail before corrosion, but it was the other way around.

Russel: We ought to talk a little bit about how you find…We’ve already talked about use MFL to find corrosion or metal loss.

Keith:

I was going to put another plug in for that diagram because all you need to know is the diameter of the pipe and the pressure. If you don’t know your steel, you can use the SMYS to put it in there. If you don’t know if your pipe will leak or rupture, then you can get a pretty good guess at it, especially for the older stuff.

That was the plug I was going to put in, what I find, when we wrote that, Dan Ersoy and I, were the advantage of using that analysis rather than the old rule of thumb, “If it’s above 30 percent yield, it’ll rupture. If it’s below 30 percent, it won’t,” because we knew there were ones below 30 percent that actually do rupture. Just another thought as we go through this.

Russel:

The other thing is we talked about MFL and how it’s used to identify metal loss, but it’s really not very good at finding cracks. People have moved to ultrasonic tools for finding cracks, which works well in liquids because of the interface and the carriage of the signal through the liquid into the pipe but doesn’t work as well in gas, for the same reason.

What’s an EMAT tool? How’s that used to find cracks? I did a whole series on this, back when I first started the podcast, with a guy named Marc Lamontagne, who’s a PhD ILI guy. He walked me through all this stuff. Sorry, Marc. I’ve forgotten. That’s been a while. I’ve forgotten that because I don’t use it every day.

Keith: Marc’s a good lad. I’ve worked with him, too.

Russel: He’s a smart dude, man. That’s for sure. Fun to work with. Great guy.

Keith:

Essentially, as you pointed out, when we use ultrasonics on pipe, we usually have to put grease or water between the piezoelectric and the pipe in order to get the signal to go through and come back to the piezoelectric in order to pick it up. You get some idea of what’s at the other end. You get a reflection off the front for the steel and one off the back.

Then if you put it at an angle, it’ll bounce off the back wall and up against to the front wall. If there’s a crack there, then some of that crack…The crack’s planar. Some of that will reflect back, so that you can pick it up that way.

If you’re using no couplant, then you have to use another technique. An EMAT uses magnetic. It puts a magnetic pulse into the pipe surface. That pulse goes in and bounces back, just like you would with an ultrasonic one, in terms of using a piezoelectric for that. Piezoelectric is like a quartz crystal.

You can then get a signal in and out. If it’s a normal beam, you can measure the close surface and the back surface, which, in the case of a pig, the back surface is the outside of the pipe. By using the signal and the timing, you can get how much wall is missing on the outside or how much wall is missing on the inside.

Russel: You’re identifying missing wall thickness inside the pipe when maybe the top and the bottom are not missing anything.

Keith:

No. They should be missing something. What you’re using is time of flight. You have a signal. You put a signal in. You should get a reflection from the front. Then a little while later, you should get a reflection from the back. That’ll give you the wall thickness of the pipe.

As the pig moves along, then if the signal from the front takes a little longer, you have wall loss on the inside. If the signal from the outside takes less time to come back, you have wall loss on the outside. It’s all time of flight, like ultrasonics, whether piezoelectric or not.

Now, you can also change the magnetic approach so that you can put the…You can change the surface of the wall to get a noise pulse which will go at an angle. You can go at an angle, say, 30 degrees to the pipe wall, and it’ll go out. Then you’re listening for a return. The return will give you an idea of how far it went.

You have an expectation. If it bounces off the weld, for example, then you know how far that is from geometry. If it bounces off in a different way from that weld, you can tell whether it’s the weld, the change in the property in the material, or if there’s a crack there.

It’s easier to do it for the outside surface. If it’s from the inside, it bounces off the wall and then up to the inner surface of the pipe. If there’s a crack there, it’s got to go back and hit the wall and come back. Of course, the sensor’s moving with the pig. It’s a large mathematical one in order to do that.

Let’s keep it simple. We’re just looking for stress corrosion cracking on the outside. There are two problems then. One, stress corrosion cracking in most cases is due to the hoop strain. That means the cracks run in the same direction as the pipe is going, what we’ll call axial cracks.

That means you’ve got to put the sound in so that it goes in and bounces off the outside wall and comes back so that you can pick it up. That means you have to put a whole bunch of these sensors all around the internal circumference in order to do that, or you can have sensors which would cause the sound and sensors which would listen to the sound coming back.

Don’t forget, with EMAT, you’re looking for movement of the surface, which is the sound. It’s really small. You have a magnetic field and you’re looking for the change in the magnetic field in a magnetic field.

If you cause the signal to go and then listen, and it’s slightly different than if you cause it and you’re listening with another one. The crack, for example, would stop the signal going to a listening one, but it would reflect it back to the one that caused the signal.

Russel: It’s that stuff, to me, without pictures is hard to visualize.

Keith: That’s why we have computers to do this today.

Russel:

Right. That’s why we have smart guys that are doing the first-stage data analysis before it goes to the operators. They’re picking the data up off the tool and figuring out how to interpret it all so they get something meaningful to the operators. They’re trying to figure out what to do with it from an integrity standpoint.

The couple of things that I would take away from this conversation is that to run an effective program, you need to know what you’re looking for, which means you need to understand what your risks are.

Because I can’t afford to run all of these tools all of the time, so I have to be pretty thoughtful about what tools I’m running and where I’m running them based on what I think I’m looking for because of what I believe the risk to be.

Because there’s other things, beyond just corrosion and cracks. You’ve got weld. You may want to look at your welds. You may want to look at repairs. You might want to look at outside force damage, geometry, location, all those things impact. Understanding what I’m looking for and what I can get on a single tool run is key to all of this.

Keith: Yeah, especially if you have a dent because the dent screws up all your geometry calculations. If you’re looking for a crack in the dent because somebody hit it with a backhoe, then maybe you’re not using the right tool.

Russel: And, some of these tools don’t work together. I might be able to put a positioning tool and a geometry tool with an MFL, but I’m probably not going to run an EMAT and an MFL on the same tool.

Keith: They do it now.

Russel: OK. The last time I asked this question, which has been a bit, the answer I got is “Well, we’re working on it.”

Keith: They figured it out.

Russel: You and I both know that there’s a difference between “I have it working” and “I have it working at a level that the operator is willing to rely on it as primary information.”

Keith: Correct. What people forget is the operator doesn’t encourage this. Then the tool doesn’t get better. The operator is part of the R&D program, too.

Russel: That’s right.

Keith: Actually, EMAT…

Russel: There’s a difference between getting data because I need it and getting data that I’m going to factor into a program that’s going to be subject to regulatory scrutiny.

Keith:

You can also think of running an ILI as part of your monitoring. You might decide to do that before you dig the pipe up and replace a pup in it.

[crosstalk]

Russel: There’s all kinds of economic considerations and all that, too.

Keith: What I’m going to say is that they can get cracks, like SCC cracks, about a millimeter deep now. They can pick them up reliably, which we couldn’t do several years ago as well. That means you’ve got a far better idea of what’s happening.

Russel: I think that’s right. We could do a whole nother long conversation about this, Keith. Maybe we should at a future time.

Keith: Should get Marc involved.

Russel: Yeah. If you get Marc and you, you guys are going to hurt my head.

Keith: If it’s good for the audience, then it might be worth it.

Russel:

It’s not a bad idea, actually. Look, listen, I so much appreciate you taking the time. As I mentioned on the other podcast, we’ll link up and try to decode all the jargon and everything on the website. We’ll provide some links.

You can also find Keith’s contact information at pipelinepodcastnetwork.com/guests and just look Keith up, or you can just go to the search at the top-right of the page and look for Keith Leewis. I will say that Keith has a strange spelling of his last name. It’s Leewis with two Es.

Keith: Just remember that it was Dutch. They left the U out.

Russel: There you go. It’s always easier to find a Leewis that has two E’s than a Lewis that only has one. There you go.

Keith: That’s right. Ray hasn’t figured that one out yet, Ray Lewis at ROSEN.

Russel: [laughs]

Russel: Exactly. Thank you so much for your time. It’s been great to talk to you. We will try and get connected at PPIM come February.

Keith: I’ll look forward. Actually, it’s the second last week in January this year coming up.

Russel: All right.

Keith: February would be a nice try, but not quite there.

Russel: I missed it? OK.

Keith: We need to use atomic bombs rather than hand grenades. [laughs] Close.

Russel: Thanks, Keith. Good to talk to you, man.

Keith: Take care. Bye.

Russel:

I hope you enjoyed this week’s episode of the Pipeliners Podcast and our conversation with Keith.

A reminder before you go, you should register to win our customized Pipeliners Podcast YETI tumbler. Just visit pipelinepodcastnetwork.com/win and enter yourself in the drawing. If you’d like to support the podcast, you can leave us a review on Apple Podcasts, Google Play, or wherever you happen to listen.

[background music]

Russel:

If you have ideas, questions, or topics you’d be interested in, please let me know on the Contact Us page at pipelinepodcastnetwork.com, or reach out to me on LinkedIn. Thanks for listening. I’ll talk to you next week.

[music]