The Pipeliners Podcast is excited to deliver a series of episodes with Giancarlo Milano of Atmos International. In this third episode of the series on leak detection, Russel Treat and Giancarlo discuss Negative Pressure Wave technology.
In this episode, you will learn how this technology is used to determine whether there is a leak in a pipeline, when and where this technology is helpful in specific environments to protect high-consequence areas, and how this technology allows you to conduct leak detection without a metering requirement.
In the next episode of the series, Russel and Giancarlo will extend the conversation beyond leak detection to rupture detection and the important considerations required in those circumstances.
Negative Pressure Wave for Leak Detection: Show Notes, Links, and Insider Terms
- Giancarlo Milano is the Senior Simulation Support Engineer at Atmos International. Connect with Giancarlo on LinkedIn.
- As part of this series with Giancarlo, enter to win our free book giveaway contest for the “Introduction to Pipeline Leak Detection” book by Atmos founders Michael Twomey and Jun Zhang.
- Leak detection systems include external and internal methods.
- External methods are based on observing external factors within the pipeline to see if any product is released outside the line.
- Internal methods are based on measuring parameters of the hydraulics of the pipeline such as flow rate, pressure, density, or temperature. The information is placed in a computational algorithm to determines whether there is a leak.
- Normal Pressure Wave technology uses highly-sensitive pressure sensors placed along a pipeline to send data back to a centralized location to determine whether there is a leak.
- Computational Pipeline Monitoring (CPM) is a method for leak detection that uses complex instrumentation and computer analysis to determine the size, scale, and location of a leak.
- The rarefaction method for leak detection monitors the rarefaction waves produced during a leak. When the substance in a pipe leaks, it creates sound waves that can be detected and analyzed.
- API 1130 is a recommended practice published by the American Petroleum Institute and incorporated by reference into the U.S. pipeline regulations in 49 CFR 195.134 and 49 CFR 195.444 for how pipeline operators should design, operate, and maintain their computational pipeline monitoring (CPM) systems. While this standard was not discussed during the podcast, it is a critical document for any pipeline operator with CPM-based pipeline leak detection.
Negative Pressure Wave for Leak Detection: Full Episode Transcript
Russel Treat: Welcome to the Pipeliners Podcast, Episode 26.
[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. We appreciate you taking the time. To show our appreciation, we are giving away a customized YETI tumbler to one listener each episode.
This week, our winner is Dustin Burch of Southern Star Central Gas Pipeline. Congratulations, Dustin. Your YETI is on its way. To learn how you can win the signature prize pack, stick around to the end of the episode.
Giancarlo, welcome back for the third episode on leak detection. We will talk about negative pressure wave today.
Giancarlo Milano: Hello, Russel. Absolutely, we are. It’s a pleasure to be back on this third episode on the leak detection series. Thank you for having me once again.
Russel: Why don’t you start by telling the listeners what is negative pressure wave?
Giancarlo: This is an internal CPM, computational pipeline monitoring and leak detection method. It’s used for the purpose of being able to detect whether there’s a leak in the pipeline or not.
The way that the technology works — this is technology that’s been around since the early ’80s — is that the system uses highly-sensitive and highly-frequently sampled pressure sensors along the pipeline in order to determine if there is a pressure drop from within the pipeline that could be related to a leak.
The information that is sampled by the pressure sensors are collected locally within a data acquisition unit, and then that information, it’s sent over from all the different locations where there’s pressure sensors to a centralized location where the information is further analyzed to determine whether there’s a leak in the pipeline or not.
As far as the concept itself, it’s based on the rarefaction methods where, when there is a pressure drop, that pressure drop is going to travel in either direction, both upstream and downstream from the location of the leak.
Eventually, that pressure drop will be picked by these highly-sensitive and highly-frequently sampled pressure sensors, and will be able to use that pressure drop to identify that a leak has occurred within the pipeline.
Russel: You probably don’t know this, but when I first got introduced to Atmos, I had the opportunity to go to your headquarters over in Manchester. I got introduced to…I wish I could remember the name of the gentleman.
Down in the basement, there was a gentleman, and he was doing the R&D on this. He had a maze of clear plastic pipe, and he was pumping water through it, and had transmitters hooked up to it, and was working on understanding the algorithm, how do you identify a leak using this technique?
Giancarlo: All right.
Russel: I know about it notionally, because I didn’t really get to take a deep dive. It was very intriguing to me.
Giancarlo: The name of the person that you probably met, his name is Andy Hoffman. He is the one that was doing all the research and development for this technology back when we started to implement it within our company. That’s probably who you met during those times.
There was a lot of research being done. There’s still a lot of research being done in order to improve the system and make it even more reliable than it is today. There’s been a lot of that. One of the things about this type of technology, as I mentioned earlier, it’s been around since the beginning of the ’80s, for quite some time.
The issue back then was that the digital signal processing, and the telemetry of such data was not there back then. Although it was available, it couldn’t really be used for the purpose of leak detection as it is today.
Russel: I remember. I was very fortunate I got to go out to one of the local pubs and have a pint with a number of the Atmos folks. I was quizzing Andy about this. Being a guy who knows a lot about instrumentation and telemetry, I was really trying to wrap my brain around how are you going to actually make this work in the real world?
I understood what he was doing in the lab. He was sampling data at like 10 milliseconds. He was using instruments that were picking up pressures at hundreds and thousands of PSI. I had a hard time wrapping my brain around how that’s actually going to work in the world.
Maybe you could tell us a little bit about how you got this out of the lab and into actual commercial use.
Giancarlo: From the research and development, we realized that we couldn’t use the conventional pressure sensors that are used for the process for the purpose of this technology, because we needed something that was more sensitive.
We also realized that in order to be able to capture the way that the pressure wave travels through the pipeline system, we had to measure it or sample it a lot faster than 5 or 10 seconds. Just gathering the data from your conventional PLC and your conventional telemetry, that type of rate is not achievable.
We worked with a few vendors that develop pressure sensors, and we found a few that worked for us. We’ve started using those in our technology. Then the company saw a little bit of a transition, because we switched from purely software into some hardware, as well, where we started doing our own data acquisition units in house.
As you mentioned, this is a data acquisition box that has the ability to sample at very high frequencies. For the purpose of leak detection purposes, we are sampling the signal at 60Hz. The pressure change itself that we’re picking up from these pressure sensors, it’s various very small changes, as you mentioned, in the hundreds of PSI. Very small changes in the pressure variation.
Russel: For our listeners that maybe have a background in instrumentation or electronics, that sort of thing, if you think about this, this wave, you can actually see it in the signal, but it’s muted by signal noise, and it’s muted by transients, and it’s muted by the accuracy of the transmitter.
If I can get this data and sample it, when I have a leak, what happens is I’ve got a pressure drop, but I can hear it. You think about putting your ear on the pipeline, you can actually hear this stuff. You can hear that flow.
What you do is you’re listening to that signal and you’re picking up things that are behind the noise. To me, it’s interesting technology, because it’s not intuitively obvious what’s necessary to do this well.
Giancarlo: Yes, it’s not. It’s very interesting when you’re looking at the raw signal from the pressure sensor. You can very easily see how a pressure wave travels through it, either from the natural noise of the pipeline — operations or just vibrations that are causing the pressure movement on the pipeline — or from a leak or an operational condition.
The thing about using these pressure sensors and the frequently sampled data is that you’re able to see a very nice change on how the pressure travels through it. If you’re sampling at 5 or 10 seconds, all you’re going to see is a step change in between measurements.
When you’re sampling a little bit faster, you can see a good, nice, gradual change of the pressure as that wave travels through it. Obviously, it’s going to depend on the type of fluid that you’re moving, whether you have a gas, or a liquid system.
The way that the pressure wave moves through the system is based on the speed of sound of the medium. We are able to see how this pressure wave travels through a system, whether it’s a liquid or a gas pipeline.
It is a technology that can be applied to either pipeline. Even in multiphase pipelines, we’re able to use it, provided that we are able to see that pressure wave movement from the location of the leak.
Russel: The other thing that I was asking Andy about when we were talking about this is, where my mind was going is that if you take something simple like a straight line of pipe.
There’s a lot of places where I’ve got one or two or three miles straight runs of pipe that are in very high consequence areas, river crossings and things of that nature where I want to know if I have the least little bit of a leak, and I want to know at the moment I have it.
What I made up in my mind was that I needed to get this data from both of these sampling locations like either end of the pipe, and I needed to put it together to do my analysis. As I understand this, the way this works, each one of these systems operates independently, is that correct?
Giancarlo: Right. They operate independently from the point of view that each location, it’s sampling its own data from the pressure sensor. Because the data is being sampled so fast, we have to timestamp it. When that is done, then we’re able to get the pressure measurement along with the time in which that pressure change was seen.
Although the information is gathered individually, all of that data is sent over to a centralized location where data from each location is put together and it’s analyzed through an algorithm where it’s comparing all the pressure sensors.
If you think about it, what you’re doing is aligning all the pressure data with respect of the exact same time, and then making sure that you observe where the first pressure drop was first seen, followed by where it was seen secondly and thirdly as that pressure wave travels through the system.
It’s going to depend on how many instruments or pressure sensors you have in the pipeline, but you are able to take all of them and analyze in a centralized location in order to make them work together for the purpose of detecting the leak in the pipeline.
Russel: One of the real tricks is this. You said time syncing. You said a mouthful there. When you think about this, I’m doing samples at 60Hz. I’m getting 60 samples a second, and I’ve got a very accurately time synced, multiple places where I’m grabbing this data to be able to do that host level analysis.
Giancarlo: Correct. There is a lot of analysis. One of the analysis that you have to make sure that you do, it’s one of the things that we’ve covered in our previous conversations is operational changes. When you have an operational change or you have leak, you could see a pressure drop.
In the case of operational change, let’s say you stop a pump at a station. You’re going to see that same pressure wave travel through the system in the way of a negative pressure wave. That pressure wave is dropping as you move down the pipeline.
When you have a leak, you have the same. How are we able to differentiate this pressure drop from it coming from a leak or coming from an operational condition? The timing is very important. To know the medium that we’re working on is very important, as well. Also, the pattern of the pressure waves travel through different instruments.
All that is analyzed in order to cope with operational conditions for the purpose of avoiding false alarms. There’s a lot of algorithms within the system to cope with all of these changes.
Russel: That leads us into the API 1130, the four characteristics. We’ve got sensitivity, reliability, robustness, and accuracy. How is this in terms of sensitivity? What size of leak, and how accurate is it at identifying a leak location?
Giancarlo: When we talk about the sensitivity, we have a highly-sensitive system. We are able to look for smaller leak sizes in the range between 0.1 to 1 percent in a matter of minutes rather than hours, or up to 60 minutes as we talked about in the previous sessions for the RTTM and the statistical volume balance. You will have a more sensitive system.
Russel: To relate that to what we’ve talked about in the last couple of episodes, that means that I’m looking maybe as much as a 10 to 1 improvement in identification of a smaller leak. I can get to a leak that’s almost a 10th of the size of what I could otherwise and I’m getting there maybe a factor of 10 to 1 in terms of speed of detection. I’m almost 10 times faster. That’s a pretty significant improvement. I also asked about leak location. What about leak location?
Giancarlo: Leak location, there’s also a huge improvement in the location, because by being able to sample the pressure sensors a lot more frequently, we’re able to very closely identify when that pressure started to drop due to the leak.
When we’re able to do that and identify that at different locations, we’re able to get a better time of when the leak started, and also where it started. When it comes to the location of a leak, we’re looking to anywhere within a matter of meters. We have seen some examples where the leak trail has been anywhere between five meters to 30 minutes [meters] during the leak trials.
Russel: You said five meters to 30 minutes.
Giancarlo: Five to 30 meters, yeah.
Russel: Five to 30 meters. That’s pretty huge. Being able to identify something within five to 30 meters, that makes it a whole lot easier to do your dig and find your leak.
Giancarlo: Correct. For both of these, the sensitivity and the accuracy, we have to take that rule that I mentioned in previous episodes. The larger your leak size, the faster your detection time and also the better your locations. This is also going to apply to the system. If you have a large leak, that’s going to be easier to identify that pressure drop as we move through the sensors.
Also, you’ll be able to locate it faster and give you a more accurate location, because you’ll have a more recognizable pattern of that pressure drop at different locations for the purpose of triangulation of that leak. It could be five, 30 meters, it could be 200 meters, it could be two kilometers. It’s really going to depend on the length of the pipeline and how separate those pressure sensors are from each other.
Russel: Yeah, that makes perfect sense. More sensors in a smaller area, I’m going to get better numbers. Fewer sensors in a larger area, I’m going to get numbers that aren’t as good just because of the way all that works. That makes perfect sense. What about false alarms? What about reliability?
Giancarlo: It is able to generate some false alarms for the purpose of being able to properly identify some operational changes if the system has not been tuned appropriately. It does require quite a bit of data to tune these parameters in order to avoid the false alarms. A well tuned system will be able to provide you with very low false alarm ratio, but it does require a lot of analysis of the pressure signals of the different locations.
Also, being able to identify the different operations during the tuning process, or making sure that all the operations have taken place. If there’s a new operation that has not been seen before, it could generate a false alarm during that, which will need further tuning and further optimization of the system in order to get rid of that false alarm in the future.
Russel: I guess it’s just like any other CPM. You’ve got to understand what you need to do to tune it and what you need to do to maintain it and how you’re applying it across the pipeline and across various operating scenarios. It’s really no different. It gives you more accuracy and more sensitivity and better performance from a time of detection standpoint.
Giancarlo: That is correct. All of these systems need to be tuned and optimized. In order to do that, we need to see the operations and as recorded by the systems, and then they need to be tuned and optimized for those particular operations. Something that we had not seen before is going to require further optimization in order to make sure that we don’t have any false alarms, that we have a more robust system.
Russel: By comparison to the other two things we talked about, Real-Time Transient Models and Statistical Volume Balance, what’s the level of effort to tune and maintain a negative pressure wave approach?
Giancarlo: I would put the Negative Pressure Wave somewhere in between the RTTM and the Statistical Volume Balance system. One of the things is that you do need dedicated hardware. The pressure sensors need to be installed in the pipeline in order to pick that up. Then, you’ll also need to connect those to the data acquisition for the purpose of gathering the data. Then you’ll need to connect those boxes to the pipeline communication system, to the telemetry in order to send that information to the centralized location.
There’s quite a bit of effort that needs to be put on at the beginning in order to get all of these key elements in place. Once the information is there and it’s been gathered, then the data needs to be analyzed just as any system based on the different operations, based on the day to day and tune for false alarms during operational conditions.
It would require a little bit more effort at the beginning due to the fact that you have to install all this hardware, but once it’s there, it’s the purpose of just gathering the data, analyzing it, and tuning it.
Russel: I think this is really interesting because we’ve talked about RTTMs. We’ve talked about Statistical Volume Balance. Now, we’re talking about Negative Pressure Wave and we’re beginning to talk about actually doing installation of equipment in the field as being part of the project, not just configuration of software at the hose.
It’s interesting to me. It’s a pretty significant improvement in performance, but I would assume there’s also a significant or material increase in cost, as well, because now, I’ve got to buy these data collection pressure transmitter devices and install them.
Giancarlo: Correct. The reason why we started looking into these type of technologies was to bring our solutions for leak detection. Not only have the Statistical Volume Balance, but also for another leak detection methodologies to other users that maybe don’t have a full meter in the pipeline.
It may be common sense for you and I here in North America that when you have a pipeline, you should be measuring how much volume comes into the pipeline, how much volume leaves the pipeline.
However, that’s not the case in the rest of the world. We’ve come across many operators that don’t have any flow meters on the outlets or they use conventional volume as balance system because they don’t have any flow meters installed. What do they do then for the purpose of leak detection? This is where these type of technologies able to come in to help those customers out.
When it comes to the point of having an additional system that bases on a different technology, redundant type of leak detection system, a Negative Pressure Wave system along with a Statistical Volume Balance or RTTM system would complement each other. You could get the best of both worlds. Maybe if you’re using the Statistical Volume Balance you’ll have less false alarms, but you have excellent location and maybe a faster detection time. There are benefits to that technology when it comes to what can we do in order to improve our leak detection system. That’s where Negative Pressure Wave system comes in.
Russel: Yeah. I think you’re making another excellent point and maybe this is another episode we can do sometime down the road, but talking about leak detection as a program versus what do I need to do for leak detection. I get asked all the time by pipeline operators, “Well, what do you do for leak detection?” That way of asking the question kind of implies, “Well, here’s my solution.”
Really, the solution is about a program because it’s part of it the tool or tools that I select and then what infrastructure do I have to put in place to maintain and operate those tools. Certainly when you have more than one tool and those things are redundant, then you start having some opportunities. You say, “Well, I’ve got one leak signal. Do I have a second?” Then, that can help to eliminate false alarms and create a higher reliability overall program. What else would you want to say about negative pressure waves that you think the listeners ought to know?
Giancarlo: I think one of the things that we didn’t quite cover as far as the APIs that say quantify the leak size. That’s one of the things that the system is lacking because we’re not measuring how much volume is coming into the system or how much volume is leaving the system. When it comes to quantifying the leak size or the amount of volume that has left the pipeline, that’s not something that we can measure as far as volumes in or volumes out into the pipeline.
This type of technology relies on the size of the pressure drop in order to make an estimation of the leak size at the location. Based on that leak size and when the pressure drop started to be seen at the location of the leak, then it would be able to estimate a volume for the system. That’s something that’s got to be said, something that’s got to be known for this type of system.
Many times when you’re combining or using negative pressure wave, if the users do have flow meters along the pipeline, the technology could be used with those signals, as well, in order to quantify the amount of product that has left the pipeline from the moment that the leak started. There are ways to improve it and enhance it if necessary.
Russel: I guess one of the things you said too that I hadn’t thought about is we work primarily in North America is there are a lot of pipelines that don’t have metering on them. They basically have a pump on one end and a tank on the other, particularly in the liquids world. The ability to be able to do leak detection without metering, that’s a big deal. The other two types we talked about absolutely require good metering.
Giancarlo: Yes. In here, you have a pressure sensor that is very sensitive. It’s able to pick up those changes and provide you with a good leak detection method. Something else to also take into consideration, Russel, is that because these pressure sensors can really be installed anywhere in the pipeline, this type of technology could be installed around high consequence areas, which in today’s time in the industry, it’s a very discussed topic.
To be able to do leak detection on, let’s say a river crossing, maybe from across a town or an area of interest where we want to make sure that we monitor that segment. One could install pressure sensors in between that area and monitor just that region.
You could have a CPM for the whole system, but then you could use this type of technology in that particular region to make sure that you are fully covering that area. It’s about as the user starts to look at the technology, concentrating on what they need and where do they need to emphasize their link detection in their pipeline and pipeline crossing. All of that needs to be taken into consideration.
Russel: Absolutely. Again, I think you make a great point. Using this kind of technology because of its speed of detection and increased accuracy in terms of location can be very helpful in high consequence or environmentally-sensitive areas. Because in those kinds of cases, even the smallest of releases could be a very big deal. That certainly creates some additional value if somebody wanted to look at it for that.
Look, I think that’s probably a good place to wrap this episode. I encourage you to go to the Pipeliners Podcast website. Go to this episode and to the show notes. There will be a very clear link. You can click through and register to get a copy of the book, “Introduction to Pipeline Leak Detection.” The first five listeners that get there are going to get a free copy, so hurry up. With that, Giancarlo, thank you very much and thank you Atmos for sponsoring the book giveaway.
Giancarlo: Thank you for having us, Russel. It’s always a pleasure.
Russel: I hope you enjoyed this week’s episode of the Pipeliners Podcast. I’ve been enjoying the ongoing conversation with Giancarlo Milano and our conversation about leak detection. I hope you have, as well. Just a reminder before you go, you should register to win our customized Pipeliners podcast YETI tumbler. Simply visit pipelinepodcastnetwork.com/win and enter yourself in the drawing.
If you have ideas, questions, or topics you’d be interested in, please let us know on the Contact Us page at pipelinepodcastnetwork.com or you can reach out to me directly on LinkedIn. My profile is Russel Treat. Thanks again for listening. I’ll talk to you next week.
Transcription by CastingWords