This month’s Pipeline Technology Podcast episode sponsored by Pipeline & Gas Journal features Micheal Nadler discussing his Pipeline and Gas Journal article over a difficult ILI and how he successfully overcame that challenge.
In this month’s episode, you will learn about the specific tool that was created to conquer the tough ILI, as well as why it was necessary and what tasks the tool performed while inside of the pipeline.
Challenging ILI of a Crude Pipeline Show Notes, Links, and Insider Terms:
- Michael Nadler is a Project Manager for Challenging Pipeline Diagnostics at ROSEN Germany. Connect with Michael on LinkedIn.
- Pipeline & Gas Journal is the essential resource for technology, industry information, and analytical trends in the midstream oil and gas industry. For more information on how to become a subscriber, visit pgjonline.com/subscribe.
- ILI (In-line Inspection) is a method to assess the integrity and condition of a pipeline by determining the existence of corrosion, cracks, deformations, or other structural issues that could cause a leak.
- Pigging refers to using devices known as “pigs” to perform maintenance operations. This tool associated with inline pipeline inspection has now become known as a Pipeline Inspection Gauge (PIG).
- Non-piggable pipeline is a portion of pipe that cannot accommodate a pig device, making it more difficult to inspect for defects. A pipeline may be non-piggable because of extreme bends, its composition, or changes in diameter.
- Tethered tools are used for ILI on pipe that lacks flow or lacks launch/receive capabilities, typically on legacy pipe. The tool is tethered to a source and typically enters and exits the pipe at the same location. It can be battery-powered using copper wiring to charge the unit.
- Bidi Pigging (Bi-Directional Pigs) are designed to be used for hydrostatic testing, displacement of water or air, removal of debris, liquid evacuation and/or product separation.
- Ultrasonic Testing (UT) is a type of NDT technique based on the propagation of ultrasonic waves in a tested object or material.
- Time-of-flight diffraction (TOFD) inspection is an ultrasonic technique that allows for the location and sizing of defects.
- FFS (Fitness For Service) Assessments are quantitative engineering evaluations that are performed to demonstrate the structural integrity of an in-service component that may contain a flaw or damage, or that may be operating under a specific condition that might cause a failure.
- Corrosion Growth Analysis identifies the cause and quantifies the rate of corrosion activity in a pipeline.
- Related Links:
- Article: Why the Inspection of Loading Lines is not just about the Tool
- Webinar: The Exploration of Tethered Technologies
- Video: Challenging Pipeline Inspections Video Library
- PGJ Article: Options for Internal Inspection of Difficult-to-Check Pipelines
- PGJ Article: Mastering Inspection of Challenging Pipelines
Challenging ILI of a Crude Pipeline Full Episode Transcript:
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 Treat: Welcome to the Pipeline Technology Podcast, episode 26. On this episode, our guest is Michael Nadler with Rosen. We’re going to talk to Michael about his June 2022 Pipeline & Gas Journal article titled “Challenging Tethered ILI of a Crude Pipeline in the Caribbean.” Michael, welcome to the Pipeline Podcast.
Michael Nadler: Hey, Russel. It’s great to be here. Thanks for having me.
Russel: Before we dive in and talk about your Pipeline & Gas Journal article, I’d like to ask a little bit about your background, where are you from, and how you got into the current role, and what you do, and all that.
Michael: I’m located in Germany, which is not a typical oil and gas area. When I did my electrical engineering degree in mid ’90s, for whatever reason I ended up with an ILI provider called Pipetronix. They’re now part of Baker Hughes.
When I started there, I was assigned to a team working on a specialized inspection tool for the Alyeska pipeline. That was technically very interesting. Best part of it, five months later, I was sitting in a plane headed for Alaska. That was my first job as an engineer and a really great start.
I really liked it right from the start, all the technical stuff of course, and then meeting people from all over the world, and going to crazy places where you usually don’t get, like the North Slope in Alaska. I was a young engineer, was hooked right from the start. Since then, I’ve been here and I still like it a lot.
Russel: It’s interesting, you mentioned you’re in Germany. I don’t know if Americans know this, but there’s a lot of pipeline technology, particularly inspection technology that comes out of Germany. There’s some really good engineers there. You guys are certainly leaders in this domain.
Michael: Yeah, there are a lot of inspection companies here, and I work for some of them mainly for developing and deploying ILI tools with all sorts of techniques. At ROSEN, I work as a project manager in the challenging diagnostics group. We look for difficult to inspect or so-called unpiggable pipelines and try to find solutions where standard tools can’t work.
This is super interesting. It’s probably the most fun part of the ILI world. Every project is unique and every project has something new to learn.
Russel: Your article is entitled “A Challenging Tethered ILI of a Crude Pipeline in the Caribbean.” Maybe tell us a little bit about this project and the nature of the challenge.
Michael: That was for an oil loading line which is used to pump crude oil from tanks out to a oil tanker, 30-inch, about three miles long. Loading lines, they pose some specific challenges to pigging, and this was one of these typical loading lines.
When you inspect a pipeline for corrosion, cracks, or whatever, usually you use ILI tools of course, and you have pig traps. But there are lines, that don’t have these facilities. You cannot pump in the direction you want, or you only have a single entry to get into the line, but there’s no trap on the other end. Loading lines are typically like that.
Most of them, they start somewhere onshore in a tank form, valve station, or something, and then they go out into the sea for a certain distance and basically end up at a loading buoy that swims on the serface and where these tankers can connect to.
Just below this loading buoy, there’s a PLEM, a pipeline-end-manifold, on the sea bed. This is typically the end of the part of the pipeline you can pig with the inspection tool. From there, there’s no piggable root up to the buoy.
It’s like a deadend. To pig that type of line you need to have a bidi tool, but that also doesn’t always work.
Russel: Michael, the thing that’s interesting to me about this is the idea that I’ve got a deadend pipeline. Because the tools are typically designed, you put them in, flow, moves them, and there’s a way to take them out at the other end.
It doesn’t really contemplate these loading lines that are going out to buoy and effectively dead end. I guess the question I would want to ask is how much of that is there?
Michael: There are hundreds or not thousands of loading lines in the world. Wherever oil tankers need to be loaded or offloaded to any onshore facility, there’s a loading line like that.
You always have these types of lines within tank farms, for example. A loading line, or something like that, to fill the tanks in the tank farm because they just end in the tank. There’s typically no pig trap on the receiving end either.
There are a lot of these pipelines, and I suspect some or many of them are just never properly inspected. Maybe they do just spot checking from external or through other means, but there’s no regular pigging or inspection program for many of these lines. This article shows that there are technical solutions for these types of pipelines.
Russel: If you talk about terminals, most of that kind of piping is above ground. When you start talking about loading lines, particularly loading lines that are going to ships, that may be above line, but often it’s not.
This is an area, at least in terms of what I read in the trades, I’m certainly no expert, but it seems to be a growing area because we’re getting larger and larger ships. Rather than having to create the docking facilities, we’re just taking the crude out to the ships.
Michael: Exactly. Yeah.
Russel: There’s a lot of this, I think. Give us a little bit of the scope of this particular project. What was the size of the line, length of the run, and all that type of stuff?
Michael: As I said, these deadend pipelines, you can inspect them with bidi pigging if you have the proper ILI tool, of course, for bidirectional pigging. In this case, it was not possible because there were restrictions on the maximum pressure of some part of the system. We could not pump a tool in and then pump it back.
Another option is of course to have some temporary pig trap installed on the subsea end. That is quite common. It’s done often, but it’s quite costly because you have to have this temporary equipment, you need supply boats, divers, and so it costs a fortune.
What we proposed to do here is to use a tethered tool that is being inserted from the onshore end, then travels to the subsea end, and then you just pull it back. You can do everything from one end. I guess subsea tethered tools are the perfect solution for this kind of setup.
Very efficient. You don’t need all this specialized equipment, and it is also cost efficient.
Russel: How long was this line that you were inspecting?
Michael: Was three miles. 30-inch line, three miles long. Only the first few hundred meters were onshore and the rest was offshore. That’s a typical setup.
Russel: Interesting. What kind of tool did you run in terms of the kind of data you were capturing, and what was your approach to designing the tool?
Michael: There was a tool based on ultrasonics for corrosion mapping. You measure the wall thickness and measure corrosion with a fine resolution. It covers the complete circumference, with a tight resolution.
Also you have a measurement for every one or two millimeters in axial direction. You get the full scan of the inner pipe surface of this line. The tethered tool was a self-propelled tool, so it has a strong tractor on the front, a tractor module that can pull the tool.
Then it has the UT part for the recording, and it has an integrated cleaning system. I guess I’ll come to this later with more details. Then in the end, of course, you have the umbilical cable, which supplies the tool with power and communication to the operator onshore.
It is also used to be pulled back when you come back to the launcher, you have a big winch reeling up that cable and pulling the tool out. That is a typical setup we have here.
Russel: Michael, I’m looking at the picture you have in the article. I see the unit on the front and I would guess the motor drive part, or the tractor drive part, of this was actually at the front of the tool, correct?
Michael: That’s correct. Yeah. That’s the tractor.
Russel: Were you able to reverse that tool and push it back out or did you have to rely on the tether for retrieval, or was it some combination?
Michael: We can do both. When you go out, you only have the tractor as a means of propulsion, obviously. When you come back, the main means of propulsion is to pull on the cable. You can use the tractor to push a little bit from behind to make it easier.
The system, the safety limits, and everything is designed such that if the tool fails, if your communication fails, power fails, or the motor doesn’t work anymore, you can always get the tool out just by pulling on it. So, when the power goes down, all the propulsion wheels they collapse and go into fail-safe mode so that they have a minimum friction, and then you can get the tool out. This is the safety net we always have.
Russel: In this particular line, did you have a lot of issues with paraffin? Because you’re talking about cleaning. How did you deal with cleaning around running this tool? I would think you have the same problem with a cleaning tool, because it’s dead in the pipeline. I’m going to push all the junk to the end and how do I get it out of the line?
Michael: Exactly. Yeah. Inspection is one issue, but cleaning is probably even a bigger issue because this line has never been cleaned before because there’s no proper means of cleaning. If you pump cleaning pigs in and you don’t get the tools out, all the paraffin clogs up at the PLEM, and so it doesn’t really help.
It’s never been pigged before. We designed a cleaning system or cleaning module integrated into the inspection tool, that could clean the pipe wall so that the UT signal is good enough, or the echoes are good enough for good for UT recording.
But we cannot bring all the stuff out because there was no flow. To pull it out or push it out would create too much friction. The cleaning system, it just did remove the stuff that is on the pipe wall, remove that from the wall.
In that particular line, there was also some soft wax, more or less a quite liquid type of soft wax. That is not an issue, but on the pipe wall there was compressed, hard, brittle wax mixed with some corrosion deposit or whatever, and it was sitting on the pipe wall obstructing the UT signal and this is what we had to get off.
Russel: Let’s talk a little bit about the cleaning tool, because I want to understand what that thing was actually doing. Was it just running across the wall or was it motorized in any way to scrub the wall?
Michael: No, it was not motorized. It has mechanical little wheels that slice off this hard, brittle stuff on the pipe surface so it falls down, or it looses off the wall and then floats away or something.
Russel: Basically, you’re taking what’s on the pipe wall, putting it into the fluid and leaving it there?
Michael: Yeah, exactly.
Russel: I would guess that there’s an operational consideration after you run this tool about that oil I have in that line is slightly contaminated. Is that correct or am I wrong about that?
Michael: I don’t know the details of it. We discussed it with the pipeline operator who has to actually make the decision. They have the option to flush the pipeline, to flush it clean. Of course, they need a tanker on the other end to accept this stuff. I don’t know how that worked out.
Russel: Wow. I don’t know. I’m completely speculating. I would think you’d want to sample and test it and see what you have and then make a decision about what you’re going to do. I’m sure that would all occur after the tool run.
Michael: Yeah, that comes after the tool run and we are not really involved in that kind of decision.
Russel: Sure. Understand. Yeah. I always have to find the edges of what people know.
Michael: Yeah. Obviously, if you compare this kind of cleaning tool, to like running proper cleaning tools where you have a full selection of soft ones, aggressive ones, and different brushes and everything, you can, of course, have a much better cleaning effect.
There was no option to do this. The only option we had is just to clean the wall, not everywhere, but just where we needed it to have a good UT measurement and then leave it behind. With tethered tools, the friction is a key parameter.
Because when the tractor goes in, it does not only have to pull the weight and the friction of the tool itself, it also has to pull the cable. The cable has weight and friction also, but the key thing here is the capstan effect when you go around bends.
Russel: Tell me, what you say, capstone effect?
Michael: Capstan effect is when you pull a cable or a hose around a bend, it creates additional friction. The more bends you have, and it also depends on the angle, the higher the friction is. If you try to pull a cable around, let’s say five or six 90-degree bends, you have to pull really hard.
This is what the tractor had to do when we went in, because onshore obviously you have bends, road crossings, and all sorts of stuff. This is also the reason why we could not use this cleaning tool when we were going in because friction would be just too hard for the tractor, for pulling everything.
When we went in, we had no cleaning. We just did the UT recording, and we had the option of having live data on our computer screen. While we go in, we see the UT data in real time. We can see how good it is and what locations it is bad because of wax.
Then we can make a decision where cleaning would make sense for the return run, when we come back. When we come back, we’re pulling on the cable, and we have a lot more power on the umbilical winch than compared to the tractor.
The tractor, by the way, it also suffers from the slippery surface due to the parafin. Coming back, we have a lot more power, and then we can switch on the cleaning tool.
Russel: Did you actually run the tool in and out more than once?
Michael: No, just once. Once in, without cleaning, and out, with cleaning.
Russel: Oh, OK, but you were collecting data both ways?
Michael: For both ways, yeah. That is standard procedure for our tethered tools. You always have two datasets. In this case, it was additionally beneficial and interesting, because on the live monitor, when coming back, we could immediately see the effect of the cleaning.
Russel: Oh, wow. Yeah, that makes sense. You actually get a sense of how good your cleaning was.
Michael: Yeah, exactly.
Russel: Which you don’t get with standard cleaning rigs.
Michael: Yeah. [laughs]
Russel: It’s interesting.
Michael: That’s one operational aspect of tethered tools that also helps sometimes a lot. You see immediately the effect of things. If something goes wrong, you see it immediately. You can even stop and try to fix it, like by resetting the tool, or going forward and back a little bit, stuff like that.
Russel: That’s great. That was the question I was going to lead to, because if I’m an operator, I’ve got an umbilical, and I can actually see the data collection, if I see something in the data collection, I have the ability to go back and look at it two or three times if that’s what I think I need to do.
Michael: Exactly.
Russel: That’s also normally available to you in an ILI tool.
Michael: No, not really, no. Another couple of other little things that are not possible with normal ILI tools: You can add additional instrumentation, especially, for example, TOFD system for checking the weld, and then you can stop at the weld, go forth and back a little bit, do a few measurements, and then continue on.
You have interaction with the tool, and because of the tractor and the winch, you can go forth and back. You have a lot more options than with free swimming tools.
Russel: I would think there’s applications just for that feature.
Michael: Yeah.
Russel: Interesting. I guess the next question I want to ask is in terms of analyzing the data, because it’s unusual to get more than one data set for a particular spot on a pipeline. That creates some interesting questions in my mind about how do you manage and what do you do with that data because it’s unique from other tools.
It’s almost like you’re running multiple tool runs.
Michael: Yeah. Both runs are initially treated the same way. They’re both recorded, then processed, and statistics made and everything. The analysts then pick the better of the two as a main data source for their analysis and reporting.
They always can check in the other one if a certain feature anomaly looks different or has a better reading for whatever reason. If the line is not clean, like in this case, it just increases the chances of getting a good reading of a particular place dramatically if you have been there two times.
Now even with two different modes of operation, one with cleaning, one without, that gives you a lot of more confidence in what you see more than once.
Russel: To me, this is really fascinating. One of the questions that comes up for me is how far out can you run an umbilical self-propelled tool? In this case, you went a little over three miles. What’s the operating limit of that type of thing?
Michael: The longest cable we currently have in our portfolio is 12 kilometers.
Russel: 12 kilometers. About 10 miles?
Michael: Yes, about 10 miles. We’re working on getting even longer. It’s not just the length, but it also depends a bit on the number and type of bends. If there are a lot of bends, there are restrictions.
We recently had a project in the North Sea. It was very challenging in terms of bends. We made calculations how much the friction will be and how far we get and everything. Estimating friction is not so easy. Obviously, in real life things turn out to be different sometimes.
Russel: There’s all kinds of variables too, like what’s on the inside of the pipeline and is it slippery or not, right?
Michael: Yeah.
Russel: Anybody that’s ever pulled wire through conduit knows about friction, right?
Michael: Yeah, exactly. In that case, we rebuild a very elaborated test loop, actually on the roof of our building and back down again, just to test the number of bends. That all worked out. Also in this project, the same principle applies.
When you go in, you always stop at certain intervals and do a pullback test. You test the real measurements versus your estimations. You have a gauge on your winch, where you can measure the force you need to pull.
Russel: You’re monitoring cable load at the winch. Every time you go through a bend or some kind of pipeline feature, you’re, “OK. Let’s see if we can still pull back.”
Michael: Yeah, exactly. We don’t want to go so far into a pipeline or around so many bends that we cannot pull it back. Even if the tool fails, we always want to be in the position to just pull it back without a lot of intervention.
Russel: I call that an orthogonal failure. It’s just basically I’m going to plan. The idea is knowing when I do my planning and engineering up front, that at this point my pullback load ought to be X. I do a pullback and it’s 1.5X. I probably need to rethink how far I’m going down that pipeline, right?
Michael: Yeah, exactly. That is what we’re doing. If the measurements deviate from our estimations and our calculations too much, we re-do the calculation, and maybe make a decision not to go in further or we say it’s okay, we can do more. That is part of the standard procedure.
Russel: What else about this project do you think it’s important to talk about?
Michael: Another challenge here was: when we go in I said it’s hard for the tractor to pull because of the load and the bends and everything, but also the slippery surface due to the wax. So, typically, we have speeds of four or five meters per minute. But here, it was so slippery and there were so many bends on the onshore section that the speed dropped down to one meter per minute, which is…
Russel: 3.3 feet per minute, basically
Michael: It was very slow, and the further we went in, it went slower and slower. When we came back on pulling with a winch, we had about five times the speed. The speed was not the only concern. What we didn’t really know how much additional friction it would cause to scrape off all this dirt from the wall.
We didn’t know, would it accumulate in front or within the tool and create additional friction, which we had to push like tons of wax with the tool. We did all sorts of studies, estimations, and calculations. We also had test pipes where we put some mortar, like cement, on the inside of the pipe and scraped it off, and tried to find out what’s going on.
It’s a lot of assumptions, a lot of guessing.
Russel: Oh my gosh.
Michael: Yeah, we didn’t really know. In the end, it worked out nicely. We did not have to abort the cleaning or anything. We monitored this load cell on the winch constantly, and everything was within the safety limit. We were prepared if things go wrong, that we just had to stop the cleaning.
Because, obviously, the priorities are to get the tool out without any trouble. Everything worked as planned.
Russel: Michael, I’m listening to you talk about this. A lot of times with these conversations, I’m always processing. I try to think about what I would want to take away.
What I would say, and this probably applies to all out-of-the-box type challenging ILI projects, is upfront being very clear about what your assumptions are and having a plan for validating and confirming those assumptions as you actually run your tool.
Michael: Plan what to do.
Russel: As you’re talking about that, I could see getting some distance out, turning the cleaning tool on, and pulling back for a while just to confirm that what happens is what you thinks is going to happen.
Michael: Yes. You need to have a plan.
Russel: In understanding where you can do that to get a good confirmation of your assumption, and yet still do it in a way that you can recover the tool. My takeaway is, just as you’re doing this type of thing, be very clear about all your assumptions and have a plan for validating every one.
Michael: Exactly. Yeah. You need to have assumptions and calculation, need to validate them and need to know what to do if there’s a deviation. That all needs to happen in real time. During the run, the inspection crew on site, of course they can pick up the phone and call someone.
But, depending what time of the day it is, you cannot just call for an engineering meeting and do some assessment for hours. You need to have a really clear plan with a decision matrix for the crew in the field to make the right decisions.
Russel: This would be my council to younger engineers, is that you need to understand what the operator’s risks and costs are associated to doing this project. Because what happens is when you understand their risks and the potential costs, you end up doing more thorough, detailed, and comprehensive engineering upfront.
It can be justified by the customer because of the risk they’re mitigating.
Michael: Yep. That’s correct.
Russel: Michael, let’s talk a little bit about the data analysis. Can you give us an overview of what occurred there?
Michael: Yep. As I said, when we looked at the live data during the run, we could see a positive effect of the cleaning when we would turn the tool or the cleaning on.
That data after the run was immediately sent to the analysts with both sets of data. They could confirm that we actually had a really positive effect of this cleaning module. When we compare the two data sets, the analyst has calculated that we had about 30 percent of degraded data when going in without cleaning.
70 percent was good data within the specification, and 30 percent was bad. When we came back the good data went up to 86 per percent. That means for 86 percent of the pipeline surface, we now have the full measurement accuracy of the tool, which is a big improvement.
Russel: Yeah, certainly a big improvement. I’m not knowledgeable. What would be standard for good data? What level would you be looking for?
Michael: If we have the option of traditional cleaning with a lot of cleaning tools, we would aim for like 95 percent or something, but here obviously we didn’t have this option that we could only increase it from 70 to 86 percent.
Apart from these, beyond these numbers, we have two data sets and even in places where the measurement is not perfect because we have two data, you can still get something out of it.
We have been in this pipeline actually before, a couple of years back, without any cleaning options, and we had also similar results concerning good versus bad data. We also did two recordings. So, in total we have now four recordings: one with cleaning, three without. And putting these all on top, of course you get a quite good set of information about where corrosion actually is, how deep it is, how big, and everything.
Then besides the typical inspection report that basically consists of a list of anomalies and statistics, we did some integrity services. Which was called FFS (fitness for service) that basically checks if the pipeline is fit for continued operation, or if you have to decrease the maximum allowable working pressure, for example.
The other thing we did was corrosion growth analysis. Again, looking at the old data from a couple of years back and the new data and see if we see the same corrosion, and did it grow over the last few years? Try to predict future growths and then we can also give a lifetime prediction for the pipeline or segment of the pipeline.
Again, because the quality of the data increased due to the cleaning that made all these analyses more reliable. Because of the four data sets, we have so much information that we can actually do this, even if the data quality is not perfect due to the cleaning issue here.
It all depends on the input data, and then you have all sorts of options to make something good out of it.
Russel: Michael, I would assume that you guys, as typical in ILI data analysis, there’s the working with the raw data and then there’s the processed data that you provide to the customer. In providing that final information to the customer, were you able to show them, “Here’s multiple passes,” or did you distill that down into a final finding?
Does that question make sense? I may be way out in the woods here. I don’t know.
Michael: No, it makes sense, but I have to think if I’m able to answer it probably. What happens is, when the data is clear, we just report what is there. We just report anomalies with the depths, widths, and height.
If it’s not clear, we can then look at both sets or several sets of data and explain where the information comes from and what the confidence is in the data. For example, the main data set was, for whatever reason, not good at that location, but we could pull some information from another data set and that indicates certain things.
Even if the confidence is not a hundred percent, we can get some information and help to better understand the condition.
Russel: That makes sense to me. Listen, Michael, this has been awesome. I’ve certainly learned a ton. It’s amazing. I am not an ILI guy. I’m not an integrity management guy, but I’ve had many of these kinds of conversations doing this podcast. I’m learning and I really value the opportunity to talk to experts like yourself and talk about some of these interesting projects that you do.
Thanks for coming on, and hope to have you back when you do another one of these interesting projects.
Michael: Thanks, Russel. It was great to talk to you. I hope that was interesting to your audience and, yeah, would appreciate to come back with some other interesting story.
Russel: I hope you enjoyed this month’s episode of the Pipeline Technology Podcast and our conversation with Michael. If you’d like to support this podcast, please leave a review on Apple Podcast, Google Play, or your smart device podcast app wherever you happen to listen.
You can find instructions at PipelinePodcastNetwork.com. If there’s a Pipeline and Gas Journal article where you’d like to hear from the author, please let me know either by going to the contact us page at PipelinePodcastNetwork.com, or reach out to me on LinkedIn. Thanks for listening. I’ll talk to you next month.
Transcription by CastingWords