Thomas Meinzer of NDT Global joins the Pipeline Technology Podcast to discuss his recent article in Pipeline & Gas Journal, “Evolution of Ultrasonic Inline Inspection.”

In the discussion hosted by Russel Treat, you will learn about how ILI has advanced over the past 20 years, the significant improvements in data digitization to support data analysis, how pipeline professionals can use tools to support metal loss detection, the underestimated role of lamination in pipeline integrity, how robot technology is supporting pipeline inspection capabilities, and more topics.

Evolution of ILI: Show Notes, Links, and Insider Terms

• Thomas Meinzer is the Head of UM Data Analysis (Compression Wave) for NDT Global. Connect with Thomas on LinkedIn.
  • NDT Global is a leading supplier of ultrasonic pipeline inspection and pipeline integrity management. 
  • Access Thomas’ Pipeline & Gas Journal article, “Evolution of Ultrasonic Inline Inspection,” in the February 2020 edition of PGJ.
• ILI (Inline Inspection) is a method to assess the integrity and condition of a pipe by determining the existence of cracks, deformities, or other structural issues that could cause a leak.
• Ultrasonic Testing or Ultrasonic inline inspection (UT) is a type of non-destructive test (NDT) that uses ultrasonic waves in the test area. UT uses sound waves to send a signal into a steel pipe to detect the presence of corrosion or cracks within the pipe.
  • Ultrasonic Shear Wave is an ultrasonic testing technique used for weld inspections. An ultrasonic transducer introduces an ultrasonic beam into the test area to detect discontinuities in the weld based on the refraction of the ultrasonic beam. Shear wave testing detects longitudinal cracks, longitudinal weld defects, and crack-like defects such as stress corrosion cracking.
  • Compression Wave Ultrasonic testing measures pipe wall thickness and metal loss. This type of test uses compression waves to send signals longitudinally through the pipe and perpendicular to the surface of the pipe.
• POD (Probability of Detection) is used in NDT to quantify the probability of detecting a specific flaw in a pipeline.
• Laminations are a type of defect that arises during the manufacturing of metallic pipe. Ultrasonic testing is commonly used to detect the presence of laminations.
  • API 579 (Fitness-For-Service) is a standard developed and published jointly by API and ASME that describes several fitness-for-service (FFS) assessment techniques that help ensure the safe and reliable operation of pressurized equipment such as pipelines.
• MAOP (maximum allowable operating pressure) was included in a bulletin issued by PHSMA informing owners and operators of gas transmission pipelines that if the pipeline pressure exceeds MAOP plus the build-up allowed for operation of pressure-limiting or control devices, the owner or operator must report the exceedance to PHMSA on or before the fifth day following the date on which the exceedance occurs. If the pipeline is subject to the regulatory authority of one of PHMSA’s State Pipeline Safety Partners, the exceedance must also be reported to the applicable state agency.
• Distribution Integrity Management Program (DIMP) looks at the various threats that can cause an unintended release of gas from the pipeline. DIMP activities are focused on obtaining and evaluating information related to the distribution system that is critical for a risk-based, proactive integrity management program that involves programmatically remediating risks.
• Clad Pipes are formed from a metallurgical bond between pipe and cladding material. The cladding is done by an explosive bonding process or by welding of clad material on a pipe.
  • CRA-clad inspections (Corrosion Resistant Alloy Inspections) are used to inspect pipelines that are utilized in highly-corrosive conditions, especially in offshore areas. [Read more about CRA-clad inspections on the NDT Global website.]

Evolution of ILI: Full Episode Transcript

Russel Treat:  Welcome to the Pipeline Technology Podcast, episode number two. On this episode, our guest is Thomas Meinzer, Head of UM and ILI Data Analysis at NDT. We are going to talk to Thomas about his Pipeline & Gas Journal article titled, “Evolution of Ultrasonic Inline Inspection,” which appeared in the February 2020 issue.

[music]

Announcer:  The Pipeline Technology Podcast, brought to you by Pipeline & Gas Journal, the decision-making resource for pipeline and midstream professionals. Now, your host, Russel Treat.

Russel:  Thomas, welcome to the first guest episode of the Pipeline Technology Podcast.

Thomas Meinzer:  Hi, Russel. Thanks for the invitation. The pleasure is on my side.

Russel:  I’m glad that you’re here. I had an opportunity to read your article that was in the Pipeline & Gas Journal back in February before I got on the microphone with you here. I think I’m well prepared to ask some questions. Let’s dive in. What do you say?

Thomas:  Yeah, sure. Let me know what you’re interested in.

Russel:  Okay. Let’s start by talking about the history of ILI and what’s been going on for the last 20 years or so. Maybe you could give me a high-level walk through of what’s been going on with ILI technology in the last 20 years.

Thomas:  Happy to do so. We had significant improvements in several areas in the past years, for example, starting with the data digitization. Let me explain the principle about that a little bit. Usually, the ultrasonic sensors are used as transmitter and receiver.

The received signals are transformed into electrical signals, which are digitized and further processed using analog to digital converters, so-called ADCs.

Now, the sampling depths of the latest ADCs provide a real high dynamic range for adequate recording of ultrasonic signals. Due to this, now, much more information can be stored in the inspection tools compared to tools from the past.

Another important topic related to what happened in the past 20 years is the miniaturization and the power management. In general, the progress of inline inspection tools is closely linked to the progress in electronics.

Russel:  I think that that’s self-evident, if you will. If you know anything about instrumentation and such, that as computers have gotten faster and more powerful, as memory has gotten cheaper and uses less power, then that trickles out into other things like instrumentation. Then, those benefits begin to hit those applications.

Thomas:  Yes, exactly. You’re right, that’s exactly the point. Highly integrated electronic components allow for compact size and reduced power consumption at the same time. As a result, an increased number of sensor channels, for example, can be accommodated in less space.

This is a huge advantage. Increased sensor channels are leading to better circumferential resolution. More parallel data processing is enabling higher inspection speed.

At the same time, higher speed is saving money for the pipeline operators in case of an inspection run because their production loss is significantly reduced. On top of all this, less power consumption is also increasing the inspection range. Nowadays, we can inspect longer pipelines with one inspection run compared to the past.

Russel:  You mention in your article that between 1990 and 2010 that the clock speed of microprocessors has increased by about a hundredfold and that the transistor count has increased by almost a thousandfold in the same period.

That’s a pretty huge movement. That’s got to have a fairly significant impact on what you can do with an ILI tool.

Thomas:  Yeah. Let me summarize the advantage from this huge development on the electronic side.

Nowadays, we are storing approximately four times more inspection data than we have stored in the mid or late ’90s. The number of signals which is available to the data analysts after an inspection run has significantly increased, which is, of course, also leading to a better outcome on the reporting side.

Russel:  I guess the analogy, and it may not be a good analogy, but if I take a digital photo and look at what I can do with a digital photo today versus what I could do with a digital photo off of a phone 20 years ago, I don’t even know if they had photos on phones 20 years ago.

Certainly, we’ve seen a huge improvement in that technology. It’s all because the ability to carry more data and to do more processing is moving forward rapidly and continues to.

Thomas:  It’s actually a good comparison, is a digital photo. The resolution on a digital photo nowadays is much higher than it was in the past. Therefore, the picture appears to be more clear.

The same principle happens with inspection data when the analysts are taking a look with their visualization software at specific anomalies.

Russel:  How has this improved the size of the anomaly that you can look at? How better are we able to see what’s going on with the pipe because of these improvements?

Thomas:  The minimum detectable feature size, is directly linked to the resolution grid. We have three different components of resolution on a UT ILI tool.

The first component is the axial resolution. This is the distance between two ultrasonic readings in axial direction. Then, we have the circumferential resolution. This is the distance between adjacent sensors in circumferential direction. The third component is the resolution of the depth measurement, which is associated to the depth profile of an anomaly.

The requirements from operators become more strict over time and with increasing experience. Therefore, there is a high pressure to improve the resolution grid and to make the pictures from the tiny, small pitting and pinhole anomalies more clear so that they can be reported in the inspection report.

Russel:  What are we talking about in terms of absolute sizes? What can a tool actually see with current technology?

Thomas:  A high-resolution NDT robot can detect pitting and pinhole anomalies with a minimum diameter of 0.2 inch at depths of 0.03 inch with POD of 90 percent. That’s really tiny already.

Russel:  That is tiny enough that you almost couldn’t see it with the naked eye.

Thomas:  Yeah, correct, in some cases because we also have to take into consideration that some pipeline surfaces may be quite rough. You need to be able to differentiate between just roughness and a real pitting or pinhole anomaly caused by a corrosion mechanism.

Russel:  That’s pretty amazing. What would that minimal detection size have been 20 years ago?

Thomas:  Approximately 20 years ago, we have been at a minimum defect size in the area of 0.5 inches. That was really a lot larger than nowadays.

Russel:  Basically, anything less than a half-inch you wouldn’t have seen with the older technology.

Thomas:  Yes, exactly.

Russel:  I guess that raises the question, how much more precision are we going to be able to get to as these tools continue to improve?

Thomas:  As I have mentioned, one of the resolution components is the depth sizing resolution. This is actually delivering the true profile of a corrosion anomaly. This current version of ILI inspection robots in NDT, have a depth resolution of 0.1 millimeters. I’m not sure exactly what this is in inches, but it is 0.1 millimeters.

Russel:  0.1 millimeters?

Thomas:  0.1.

Russel:  That is…you’re talking thousands of an inch now. That’s tiny, tiny, tiny.

Thomas:  Really.

Russel:  When I think about that, Thomas, one of the questions I ask is at what point do you start having so much data that it becomes meaningless?

Thomas:  Actually, this point does not exist. You can never have enough data. The more data you have, the more data points you can use for a decision whether signals are reliable or not. The more data you have, the more precise will be your outcome, finally.

Russel:  Right. What about feature interactions? Are we to the point that given the detection capabilities of these tools that we can actually start picking feature interactions between things like crack and corrosion combined?

Thomas:  Based on the continuing miniaturization of electronic components as well as the progress in the development of the sensors, a combination of several independent inspection technologies in one single tool is one of the essential enhancements.

Combinations of different inspection technologies are finally supporting the identification of interacting threats. While an individual defect may not be a major concern, the discovery of multiple types of defects at the same location may be critical for the safe operation of a pipe segment.

With the combined technology approach, we are able to detect interacting cracks and deformations, interacting metal loss anomalies and cracks, also deformation and metal loss anomalies. All different kinds of combinations, depending on the different technology we are using for an inspection. All different types of combinations can be identified.

Russel:  With this lowering power requirement and with the miniaturization of the sensors themselves, I would assume that it’s much more feasible to start putting multiple tools together in a single tool run. Is that a trend that’s occurring? Do you see that expanding in the future?

Thomas:  Yes, absolutely. That’s one of the current important improvement areas. Combined inspections offer several benefits for the pipeline operator.

For example, mob, demob and cleaning activities are required only once. The data sets from the different technologies will be perfectly aligned during that data analysis procedure.

Aligned data sets are very helpful for the data analysis outcome. As well, they are helpful for the identification of the interacting threat. Also important to mention, applying several independent inspection technologies for the same feature type may finally improve the POD for this specific feature type. It may also lead to an increased sizing accuracy.

Russel:  I’m sorry. You used a three-letter acronym on me that I was not familiar with. POD?

Thomas:  That’s the probability of detection.

Russel:  Ah, okay. Got you. So, we’re talking about things getting smaller. I think one of the other things that comes up for me as I think about that is, what’s the smallest pipe size that you can run a tool in? Where are we now versus where were we 10 or 20 years ago?

Thomas:  That’s a really good question. Smaller inspection robot components means, also, advanced benefits to small diameter inspection runs. The advantage of the miniaturization and the reduced power consumption will lead to several significant refinements related to small diameter inspections.

NDT Global robots are available from six inch upwards, up to 48 inch. You may imagine that batteries for a six-inch robot are not that large.

In the past, the inspection range was limited due to exhausting batteries. Nowadays, longer, small diameter pipelines with distance ranges up to, for example, 150 miles can be inspected without any power issues.

Russel:  I could certainly see…I’m thinking right now on the natural gas side, but certainly there’s a lot of four-inch pipe in natural gas that could probably benefit from inspection. Do you see us getting to the point we get the tools that will be able to run in a four-inch pipe?

Thomas:  Certainly, technology-wise, this would be possible. I’m not aware of any thoughts in that direction in NDT Global.

Russel:  Thomas, I’m operating well outside of my knowledge so I just have to ask these questions.

Thomas:  You’re welcome. I’m happy to answer all your questions if I can.

Russel:  I want to talk a little bit about shear wave and compression wave. Maybe you could talk to us about what is shear wave, what is compression wave, and how do they help in terms of finding different kinds of defects?

Thomas:  There are different measurement principles for the main technologies, metal loss detection and crack detection, are based on two different wave types, as you mentioned already.

Let’s discuss compression wave and metal loss first. The so-called compression wave is the most efficient way of transmitting sound through liquids or solids. Compression waves are used for the metal loss detection and travel at high speeds with a minimal loss of energy in the steel.

The UT metal loss technology measures the signal runtime of a sound pulse which is generated by a transducer traveling through a test piece. In case of an ILI inspection, the piece is the pipe wall.

The angle of incidence is perpendicular in case of compression wave. Perpendicular to the pipe wall. From the time measurement, the actual wall thickness is calculated taking into consideration the speed of sound in the steel. That’s the principle of the metal loss measurement.

For crack detection, we are using the shear waves. In the case of crack detection, shear wave pulse is traveling through the pipeline wall at an angle of 45 degrees, like a zigzag, and is reflected either by an external or an internal flaw.

Based on individual patterns of the reflected signal and supported by algorithms, the recorded information is then transferred into crack properties by our analysts. Reference sensors onboard measure the medium properties and the wall thickness to ensure that the crack robot is always operating within the optimum settings.

Russel:  I’m going to try to play this back, just to make sure I understood it. Compression wave is I’m taking the signal and I’m shooting it directly at the pipe. I’m perpendicular and I’m shooting directly at the pipe. I’m reading the reflection back.

Thomas:  Yes.

Russel:  Shear wave is I’m shooting at an angle. If I’m shooting directly at the pipe, I’m going to find missing metal.

Thomas:  Yes.

Russel:  That makes sense because the way the signal reflects would be dependent on how much metal is there to reflect back. If I know what the wall thickness is or should be and I know what my signal is, I can get my metal off.

With the shear wave, I’m at a 45-degree angle. I’m actually sitting here and using my hands. I wish people could see what I’m doing with my hands. With a shear wave, you’re firing at an angle. Consequently, the reflection’s going to be against any cracks versus against metal loss.

Thomas:  Yes, with the shear wave we are more shooting for flaws, which are touching the surface on the internal or the external side. With compression wave, we are shooting directly into the material to measure the thickness.

Russel:  What kind of defects are hard to identify, just in general? How is this being able to miniaturize and combine tools affecting your ability to detect them?

One of the things that we had talked about off-mic was laminations. Maybe you could talk about what a lamination is and how you go about identifying laminations.

Thomas:  That’s actually a good topic because when we talk about the metal loss technology in general, we often forget about the capabilities related to the detection and sizing of laminations.

A lamination actually is a quite common defect in the manufacturing process. We see laminations mainly in seamless pipes, but also in longitudinal welded pipes — they are detected quite often.

A lamination is actually a subsurface type defect that mostly occurs in parallel to the pipe surface originated during the rolling process. As long as a lamination is parallel to the pipe surface, isolated, and does not touch a weld, it is usually not really relevant for the integrity of a pipe.

Laminations may also have a sloping character in the pipe wall or maybe in connection to a weld. In this case, they also should be assessed further because such lamination may reduce the effective thickness of a pipe.

Russel:  Interesting. I’m listening to what you’re saying, Thomas. I’m trying to process it as I’m thinking about it. What this is causing me to think about, I don’t know if it has any relevance at all, but as metal corrodes you’ll see it flake, right? Is this similar to that, but in terms how the steel’s actually manufactured in the first place? Does that flaking nature exist within the steel?

Thomas:  Yeah. Actually, a lamination in the pipe wall is a perfect reflector for the UT beam, as it provides identical conditions compared to the back wall of the pipe. Therefore, the amplitudes from lamination measurements in the inspection data provide reliable information for further evaluations.

Lamination contours like sloping or surface breaking can be detected and sized accurately. The significance of laminations can be assessed afterwards, according to API 579 or, for this assessment, different depths and lengths values are taken into consideration. Those depths and lengths values use can be sized in the inspection data.

Russel:  I guess another question to follow up, that is laminations are fairly common, if I’m understanding what you’re saying. They’re frequently benign as it relates to pipeline integrity.

There are some situations where they can be problematic. The complexity is not only identifying that the lamination exists, bot to identify if this particular lamination is something that you should have concern about.

Thomas:  Yes, exactly. For specific laminations with sloping contours or with surface connection, it’s also important that the analyst is able to differentiate between normal metal loss and such lamination with a sloping contour.

Russel:  Interesting. I’ve never heard about laminations before this conversation. That’s adding to my education here.

Thomas:  I’m happy that we found such a topic because I know that you are well experienced.

Russel:  I talk frequently about inline inspection. I have a pretty good understanding because I’m an engineer. I’m a structural engineer by education, so I have a pretty good understanding of things.

This is certainly not my expertise. I know enough to be dangerous in this domain, is probably the way I would characterize it. I don’t know enough to be competent in this domain.

Thomas:  I get it. The lamination topic is really an important one because when we talk about ILI tools we talk about crack and about corrosion, but lamination actually is kind of underestimated.

Russel:  Right. I think the other thing that’s going on, just in general…We haven’t talked about this, but the fact that I’m able to gather more information. I’m able to gather it more accurately. I’m able to overlay it and align it more accurately is causing me to be able to see a whole bunch of things I’ve never been able to see before.

That causes some other complexities from an overall program standpoint. I’ve got all this information. I’m identifying all these features, but how many of these features are actually problematic? The more features I identify, and inventory, and monitor the more complex this becomes.

Thomas:  Exactly.

Russel:  Of course, the flip side of that is as we start to apply some of the machine analysis and such, we’ll be able to get where we’re dealing with an abstraction, a level away from the detailed data. We’ll become more effective. That’s a whole another conversation, I think.

Thomas:  We are starting, also, activities towards machine learning, automated intelligence for data analysis because, from the high amount of data we are getting with one run nowadays, some automated support is required to get those data analyzed in a sufficient timeframe.

Russel:  Right. It’s certainly different than the process we were doing 20 years ago. With the amount of data that we’re collecting, you’ve got to have a way to visualize that data. You’ve got to do some level preprocessing to be able to get to where the engineers can look at it and start doing their assessments.

Thomas:  Yes, exactly.

Russel:  What are some of the other challenges? We’ve talked about improving accuracy and talked about laminations as a challenge. What about things like higher pressures, higher temperatures? What are some of the requirements there? How are they being addressed?

Thomas:  The NDT Global robots are not only designed to inspect standard onshore pipelines. They also operate reliably under difficult conditions. You mentioned high pressure and high temperature.

For example, tools are able to operate under temperature conditions up to 176 degrees Fahrenheit. I just had to convert this. (laughs) They are also able to operate up to a maximum pressure of 1,740 PSI. As long as the temperature and the pressure is within those ranges, the robot performance is not affected at all.

The good thing is that runs with high temperature and/or high pressure do not happen that often, but in general, the principles would also work under those conditions.

On top of this, other challenges are related to construction techniques, for example. Thinking about clad pipes. Metal loss robots deliver excellent results for CRA-clad inspections. The different material properties of the clad and the carbon steel base material do not affect ultrasonic inspection performance.

The ability to customize the configuration of the NDT inspection robots makes them perfectly suitable for that detection of metal loss anomalies also in CRA pipes. We have a range of parameters to set up the tool according to specific needs. That is really very supportive to widen the range of conditions where the robot can operate.

I could tell you also other challenges, for example. It’s not a challenge we have very often, but several times a year, we are inspecting pipelines with internal coating. There are two types of internal coating like epoxy, fusion-bonded epoxy. Such a coating may be applied after the construction with special tools to avoid further growth of already known defects, or it may be directly applied during the manufacturing process.

With the metal loss technology robots, we are able to evaluate the condition of the internal coating as well as anomalies behind the coating. Contributing factors here are coating thickness and coating materials. For instance, coating thickness should be in a range up to 0.1 inch if anomalies behind the coating shall be detected and sized accurately.

Russel:  I guess that would cause some pretty good complexity in how you’re setting up the robot to gather its data if it’s got to look behind an internal coating.

Thomas:  Yeah. As long as the coating is bonded to the inner pipe surface, the signal gives you both signal values from the coating as well as from the pipe wall behind the coating, so you can evaluate actually both.

That’s very important, because if the coating was applied after the construction due to the fact that the pipeline was damaged already, then the condition of this coating is very important to monitor whether those defects behind the coating are growing or whether they stay with the same depth, length, and width in the pipeline behind the coating.

Russel:  Are you actually able to pick up debonding between the coating and the pipe wall?

Thomas:  Yes. In case the coating is not bonded to the inner pipe surface, then the UT beam is actually stopped behind the coating. It cannot penetrate the pipe wall due to the fact that there is a gap between the coating and the pipe wall. In those cases, we would only measure the thickness of the coating and nothing behind, and then we would know that at this particular area, the…

Russel:  In effect, you create, if you’ve got a debonding, it creates a blind spot in terms of the tool’s ability to see this deal.

Thomas:  Yes, that’s a way we could describe it, yes.

Russel:  Interesting. Wow. It always amazes me, Thomas, the more I learn about any aspect of pipelines, the more I realize how much there is to learn. [laughs]

Thomas:  Yeah.

Russel:  It’s never-ending. Well, we come to the end of our time, but I think we’ve covered a lot of what you had in your article. Is there anything else that you’d want to add so that the listeners can get a bit better view of what you had in your article, or anything you would want for our listeners to take away from the conversation?

Thomas:  Yeah, I agree. I think we have covered the main topics. Nowadays, for the inspection of liquid pipelines, ultrasonic ILI tools offer specific advantages with regard to resolution as well as measurement accuracy.

As we said at the beginning, since the ’80s, these tools have been considerably improved by taking advantage of the progress in electronics, data processing technology, and also data storage capabilities made available from other application areas actually.

The evolution is still ongoing, and we can look forward for further upcoming improvements. I’m sure that this evolution has not stopped.

Russel:  Yeah. I will tell you what my key takeaway from this conversation is, Thomas. That’s simply something you said earlier, and I’m going to paraphrase. Engineers will never have enough data.

Thomas:  Yeah. I fully agree to this, yeah.

Russel:  Thanks for coming on. This has been awesome. I really appreciate the opportunity to get to visit with you.

Thomas:  Thanks, Russel. The pleasure was mine. I was happy to answer your questions.

Russel:  I hope you enjoyed this week’s episode of the Pipeline Technology Podcast and our conversation with Thomas Meinzer. If you’d like to support this podcast, please leave us a review on iTunes Apple Podcasts, Google Play, or wherever you happen to listen. You can find instructions at pipelinepodcastnetwork.com.

If there’s an article that’s been published in the Pipeline & Gas Journal and you’d like to hear from the author, please let me know on the Contact Us page of pipelinepodcastnetwork.com, or reach out to me on LinkedIn.

Thanks for listening. I’ll talk to you next week.

[music]

Transcription by CastingWords