This week’s Pipeliners Podcast episode features Dr. Jun Zhang, the CEO and founder of Atmos International, discussing her career in pipeline leak detection.
In this episode, you will learn about Dr. Zhang’s educational background, how she started Atmos International, the challenges she faced when starting a company in the pipeline industry, how leak detection technology has grown over the past 25 years, and her predictions for the future of pipeline leak detection.
Leak Detection: Show Notes, Links, and Insider Terms
- Dr. Jun Zhang is the CEO and founder of Atmos International. Connect with Dr. Zhang on LinkedIn.
- Atmos International is the world’s leading pipeline leak detection and simulation provider. Dr. Zhang founded the company in 1995 in Manchester, U.K.
- Atmos was recently awarded the Queen’s Award for Enterprise in the Innovation category. The award recognizes outstanding achievement by a U.K. business.
- Smart Awards for Innovation are grants available by the U.K. government for game-changing and disruptive ideas from business, through a regular, competitive application process.
- Leak Detection is the process of monitoring, diagnosing, and addressing a leak in a pipeline to mitigate risks.
- 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.
- Shell is a British-Dutch multinational oil and gas company headquartered in the Netherlands and incorporated in England.
- The Hague is home to the Shell global headquarters in The Netherlands.
- BP (formerly The British Petroleum Company plc and BP Amoco plc) is a multinational oil and gas company headquartered in London, U.K.
- ExxonMobil is an American multinational oil and gas corporation headquartered in Irving, Texas.
- Fault Detection is the process of being able to monitor a system, observe when a fault has occurred, identify the type of fault, and locate the fault with accuracy.
- AI (Artificial Intelligence) is intelligence demonstrated by machines in contrast to the natural intelligence displayed by humans.
- IoT (Internet of Things) is a system of interrelated devices with unique identifiers with the ability to transfer data over a network without requiring human-to-human or human-to-computer interaction.
- IioT includes Industrialized devices suited for creating data, connectivity, data delivery, and consumption. This term encompasses sensors, edge computers, controllers, communication devices, PCs, and HMIs.
Leak Detection: Full Episode Transcript
Russel Treat: Welcome to the Pipeliners Podcast, episode 141, sponsored by Burns & McDonnell, delivering pipeline projects with an integrated construction and design mindset, connecting all the project elements, design, procurement, sequencing at the site. Burns & McDonnell uses its vast knowledge and the latest technology with an ownership commitment to safely deliver innovative, quality projects. Learn how Burns & McDonnell is on-site through it all at burnsmcd.com.
Announcer: The Pipeliners Podcast, where pipeline professionals, Bubba geeks, and industry insiders to share knowledge and experience about technology, projects, and pipeline operations. Now, your host, Russel Treat.
Russel Treat: Thanks for listening to the Pipeliners Podcast. I appreciate you taking the time. To show that appreciation, we give away a customized YETI tumbler to one listener each episode. This week, our winner is Mike Legere with Energy Transfer. To learn how you can win this signature prize pack, stick around ‘till the end of the episode.
This week, our guest is Jun Zhang. She’s joining us from Atmos International in the U.K. to talk about a lifetime in pipeline leak detection. Jun, welcome to the Pipeliners Podcast. I’m so very glad to have you as my guest.
Jun Zhang: Thank you very much, Russel, for having me.
Russel: First off, let me ask, if you would, why don’t you tell the listeners a little bit about yourself and about your background?
Jun: I’m Jun Zhang. I’m the CEO of Atmos International. We provide pipeline leak detection and simulation technologies to the pipeline industry.
Russel: I might add that Jun is based out of Manchester, England. It’s morning for me and afternoon for her. My voice is a little rough because we’re having a lot of pollen in the air in Houston. If I’m hacking a bit, I apologize. Jun, if you would, tell me, how did you get involved in pipelining? How did that come about?
Jun: I did my Ph.D. in fault detection and diagnosis for chemical processes at Manchester University. In 1988, Shell International wanted to apply leak detection to their operational pipelines, but they could not find anything in the market that matched their needs. They decided to develop their own leak detection system. They hired me to work on that because of my research background.
Russel Treat: What’s your educational background, Jun?
Jun: I had a BSc in Electrical Engineering, an MSc in Process Control, and a Ph.D. in Fault Detection.
Russel: Interesting. To my mind, particularly when you start talking about what you did in your master’s and doctoral work, that’s quite specific.
Jun: When I was choosing a research subject, I wanted it to be something practical so I could use it in industry when I finished my research. I was very fortunate that I managed to convince my supervisor to do this research with me. It was the first time for him to do it as well. He actually learned it at the same time as I started my research.
Russel: Your Ph.D. was actually in a discipline that didn’t widely exist at the time. Is that what you’re saying?
Jun: Exactly. No one else did it before [laughs] in our university. I had to ask my supervisor to do it just for me.
Russel: That’s fascinating to me, Jun, because there was a point in my career where I was thinking about going back to university to pursue a Ph.D. As I began to investigate that, one of the things I found is that typically Ph.D. candidates do not get to define their discipline.
It’s defined for them based on whatever research the university is doing at the time. To me, that’s a big deal that you were able to accomplish that.
Jun: You’re right. I was very lucky. I still think because of what I chose to do, that has more or less formed my career. What I’m using today for leak detection, it’s still the statistical analysis which I used for my Ph.D. research. I’m really fortunate to be able to use that 30 years later.
Russel: Isn’t that fascinating, right? You, in effect, invented an entire new technology and commercialized it. I admire that. I admire that. It’s a big thing.
Did you go direct through your educational program, through the Ph.D., and then enter the workplace or did you work prior to entering into your Ph.D.?
Jun: I went straight to BSc, MSc, then Ph.D. Then I went to work for Shell. Because of what I did for my research, I managed to invent the statistical leak detection system within a few months when I was working for Shell Research in Amsterdam.
I had the opportunity to test it on a few operational pipelines within Shell. That included a propane line, ethylene line, and the natural gas pipelines. I went straight from my research to application in Shell and had the technology developed based on the research.
Russel: One of your first projects was applying this approach to an ethylene pipeline?
Jun: No actually the first one was on a propane pipeline in Shell U.K. The second one was an ethylene pipeline.
Russel: That’s about as difficult as you can make it, at least given what I know about ethylene.
Jun: You’re right. We didn’t choose the easiest one to try because it all depends on the availability of the operating company and if they could innovate. We always withdraw fluid from the line so we could flare the propane from the pipeline for the leak trial. That’s why Shell chose that pipeline to start with.
Russel: Just the hydraulic characteristics of an ethylene pipeline, just talking about level of difficulty. If there’s a scale between 1 to 5, you picked a 5.5.
Russel: That’s just tough.
Jun: It was our natural choice, but it’s what was practical at that time and what was available for us.
Russel: It gave you an opportunity to prove and practice your research.
Jun: Exactly. Once we could make it work on an ethylene pipeline, we know it will be quite easy to make it work on a crude oil or petroleum pipeline.
Russel: Exactly, just because the hydraulic characteristics are simpler on that stuff. You went straight from your Ph.D. to Shell. I assume that Shell hired you for this purpose, to get involved in their leak detection program and efforts. What was driving Shell at that time to look into leak detection and try to come up with a creative way to do it?
Jun: Shell wanted to be environmentally friendly and tried to minimize the consequences of leaks, but they could not find any product in the leak detection market. That’s why they wanted to carry out research and develop their own technology. That’s why they hired me.
Russel: Interesting. Obviously, you left Shell and you started Atmos. How did that happen, that you decided you actually wanted to take this technology and bring it to the broader market?
Jun: It’s a very interesting question because I was really enjoying my career within Shell. I was progressing very well. When I was in The Hague, which was the Shell head office, in 1995, the International Pipeline Manager came to see me. He wanted help to make the leak detection system work on an operational pipeline, which was our own technology.
I thought that was a good time for me to leave Shell to start Atmos International. That’s when I left Shell to continue with the development and the delivery of the statistical system.
Russel: Were you able to acquire the technology from Shell and then use that as the foundation for the business?
Jun: Yes, we had a license agreement for Atmos to commercialize the Shell statistical system worldwide.
Russel: Being someone who has done these kinds of things myself, you’ve now done two things that are both extraordinarily challenging to accomplish for all kinds of reasons. The first is convincing a major university to allow you to do a Ph.D. in your chosen discipline.
Now, you’re convincing a major international oil and gas company to license out their software to a start-up business venture. As I’m having this conversation with you, Jun, I’m learning to respect you even more than I already do. That’s a pretty incredible story, just that far.
Jun: I think I was very fortunate that the Shell International Manager was very open, even though he knew I was an employee at Shell. When I decided to quit to continue the technology, he had full confidence in my technical ability to make it work.
I was also very confident, having implemented the technology myself on a few pipelines. I know I could make it work. That also gave me extra confidence to go ahead with a new business.
Russel: I think, Jun, that for anybody who’s listening to this and has had the idea that, “I want to start a business.” You’re making an extremely important point. You had a technology which had been deployed in the field and had proven its ability to accomplish its objectives.
You were starting with something that was already known to work and had a referenced install base. That’s pretty extraordinary, particularly in this kind of, what I’d call, heavy engineering or high technology type of thing. It’s so very important that that is true. You can’t really get going in a company until you have at least that.
Jun: You’re right. Also, I think we are fully aware as well that the pipeline industry needs a good, reliable leak detection technology. The customers need the problem to be solved. If we can make it work and solve their problem, help them protect their pipeline and minimize the consequence of pipeline leaks, then we will be successful.
Russel: Yeah, exactly. As you were early in your career, starting Atmos, what were some of the challenges you faced?
Jun: I suppose there were many challenges at the beginning. I probably wasn’t idle enough in order to worry about any of them. Also, I do think my experience in working for Shell for about eight years has helped me succeed in many ways.
Why do I say this? When I first started my job at Shell Research, it was in 1988. That was 32 years ago. Although I invented the statistical leak detection system within the first few months, I actually did not get a high rating from my management on a review because all the criteria for annual review were based on the behavior of a white male in those days.
My rating as a Chinese woman was very low. That was actually very disappointing for me. I did not want to accept that. In order to impress the managers, I started to observe how the well regarded male colleagues behaved and started to adapt myself to the local culture, also to avoid being regarded as a technical expert only, without management potential.
I also convinced my new boss at Shell U.K. to sponsor me for an MBA through distance learning. With the U.K. management support and my determination to demonstrate my high management potential, I actually managed to get my rating changed, which was also very difficult in those days.
Once your first boss ranked you, it’s very hard for the second boss to change it. It also took a lot of courage for my U.K. manager to overturn the rating and put me on a fast track for career progression within Shell. That was a very good experience. Without that challenge within Shell, I don’t think I would have learned so much about the oil and gas culture. It has also made me more adaptable.
I believe this is also an essential skill nowadays as we must adapt to changes all the time. The current coronavirus outbreak is a good example. This is unprecedented for the world. We have to be resilient and adaptable to overcome the severe impact that it has on all of us.
Russel: Jun, you talk about early in career being a Chinese woman and coming straight out of an educational track into an industry rack and challenge of that. I had a very similar experience myself early in the military where I was working with many senior officers and was having much difficulty getting my job done because relational skills and…
There was a whole ‘nother level of — I’m going to say — manners. Emily Post kind of, how do you carry yourself? How do you talk? How do you sit? How do you dress? Those things that I had never really been exposed to. I was fortunate in that I had some mentoring.
They said, “Yes, Russel. You’re very smart. Yes, Russel, you’re good at what you do, but you’ve also got to be good at these things.” I set about learning all that. In a very short period of time, it had a major, major impact on my career path.
I would say to anyone, we all have cultural challenges, regardless of what our background is because we’re never a perfect fit when we show up at the door. [laughs] Hence, it goes to your point about adaptability and how important that actually is.
Jun: You’re right. I also say it to my young colleagues nowadays. I said, “I wish I had a mentor when I started my career instead of having to find out everything myself without anyone to turn to, asking for advice.”
Russel: There’s a lot of us now, older folks in business, that have stepped in all the potholes and made many of those mistakes. If somebody will come and ask we could actually help them avoid it.
Jun: Yeah, that’s also why I’m very keen in mentoring young people myself to give them the support to make life easier for them from the very beginning.
Russel: They’re going to have a whole different set of challenges than we had. The world’s a different place. That’s what I would assert.
Jun: You’re right. The early challenge when I started Atmos was to convince potential customers that our pipeline leak detection technology will work on their pipelines because there were quite a few systems in the market that did not deliver what the vendor had promised.
The market had a very bad reputation and no customers believed any vendors when we started. To demonstrate how our system worked, we built a portable pipeline system that pumped air in a small plastic tube with flow and pressure sensors and a valve that could be opened to simulate leaks.
We built all of them in a suitcase, could carry it around the world to do a demonstration in order to convince the potential customers that our system would indeed work.
Russel: How big was that suitcase?
Jun: It was a Samsonite case, which is one of the medium-sized ones with a compressor inside and two flowmeters, two pressure sensors, and a 270 meter long, 4 millimeter diameter tube.
Russel: Here’s the next question. How did you get that through airport security?
Jun: Actually, because it had no battery. It was power-driven. It seemed to have gone through security largely without any issue. I had been stopped in Brazil, going into Rio. In many other countries, it just went through like my personal suitcase.
Russel: Fascinating. That’s really awesome. It goes to your creativity. I’ve got a picture of that. Do you still have that suitcase?
Jun: We do. We still have a few versions of it in different offices.
Russel: That’s awesome. That should be inside a glass case in reception at the home office, right?
Jun: [laughs] We should build another model for it.
Russel: Yeah, exactly. Were there any other challenges that you faced as you were getting started?
Jun: I think that the main one, is to convince the customer it would work when the market was so corrupted. Also, to have a few customers who would believe in us and having a system installed. It helped a lot. Of course, Shell, and BP, and ExxonMobil — they were great customers at the very start of our business.
Russel: Atmos, from when you started it back in…What was the year you actually started Atmos?
Jun: 1995, so it’s 25 years ago.
Russel: What would you say that over that 25 years, you guys are now an international company? How many countries do you have customers in at this point?
Jun: We have about 160 customers in 60 countries now.
Russel: What would you attribute your success to?
Jun: I think the most important thing for our success is to help our customers solve their problems. Pipelines provide the safest form of transport for moving oil and gas around but they still could leak due to various reasons.
Detecting leaks reliably would help pipeline companies reduce the environmental impact significantly. If we can provide a reliable solution for these customers, I think that would be one of the important success factors.
To make sure we continue to improve the leak detection technology, Atmos invests about 30 percent of our revenue in research and development every year. We have also won a few Smart Awards for Innovation from the U.K. government. That has certainly helped us with both the company’s reputation and the financial support.
Russel: I was very impressed when I saw a press release come out where I saw that Atmos had been recognized by the Queen. I don’t know exactly what the award was. All I saw was Atmos and Queen. I’m like, “Okay, that’s just really, really cool.”
First off, that’s a very prestigious award. Secondly, you’ve got to be in the U.K. to win something like that.
Jun: You’re right, we have won the Queen’s Award for Enterprise in the category of innovation this year. It was a great recognition for our innovative work and the achievement in pipeline leak detection.
Russel: Congratulations. That’s a credit to you and to the entire Atmos team.
Jun: Thank you very much, Russel. I’m very proud for this achievement. We thank all of our customers for their continued support over the years and also thank our team for their outstanding work.
Russel: You haven’t mentioned this. I know you plan to get to it, but the team at Atmos, you have, at least in my experience, given the people I’ve met. I’ve met quite a few. You have a level of technical expertise and teamwork that, at least in my experience, is a bit unprecedented.
It’s quite an impressive team that you’ve been able to assemble that’s made this organization such a success.
Jun: You’re right. I don’t think we would have achieved anything without our fantastic team within Atmos. We’ve been very fortunate that we have such an amazing team of over 180 professionals around the world.
We have very dedicated professionals who are totally committed to providing the best customer service in the pipeline industry. We have very high caliber research and development innovators in our company, as well. We’re really fortunate in that we need to thank all of my colleagues, all their commitment and their contributions.
Russel: Making such a huge investment into R&D and having met some of the other PhDs that are on your team, you guys are doing some really interesting stuff. I’m telling you what my opinion is, at least in my experience. You have taken what you did in university and research and have perpetuated that and built a team to continue it. Some of the stuff you guys are doing is quite interesting, I think.
Jun: It’s very interesting. I think some of my friends would joke with me. We built the biggest laboratory in our industry. We’ll carry on doing research like in the university. I think it’s because our passion is in innovation.
We know our customers want to detect a drop of oil within a second without any false alarms. They want to do it cost-effectively. We know the aim for getting this perfect system is never going to end.
We’ll have to continue to innovate, getting better sensors, getting better software algorithms using the latest technology that we can get a hold of — like AI or Internet of Things — and the better sensors, and so on.
We have to continue to innovate in order to improve the technology all the time.
Russel: That’s a great segue because the other question I wanted to ask you is what do you think the future is for pipeline leak detection? In particular, all this advanced math that’s getting so much airtime these days.
Jun: I think the pipeline industry will certainly benefit significantly from new technologies. We are living in a very interesting time with the development of sensors, the Internet of Things, and big data. The pipeline industry would get a lot more applications from these new technologies.
In terms of whether artificial intelligence or machine learning could be used for leak detection, we still have doubt about it. This is because most of the pipelines do not leak. That means there are very limited leak data, available for machine learning to learn from it.
In addition, every pipeline is unique with its own hydraulic behavior and metering performance. The flow and pressure that we have from one leak in one pipeline can be totally different from another leak in another pipeline. You cannot use this leak data to teach machine learning to detect the leaks.
Russel: I think that’s what a lot of people don’t understand about artificial intelligence, extraordinarily good at seeing something and recognizing that it’s seen it before. That requires a large training set to tune the model and get to a level of reliability.
To your point, there’s not a lot of leaks. Even on some of the very largest pipeline systems, they might have one leak every two or three years. It’s just not enough data to train an AI.
Jun: You’re right. I think some pipelines have never leaked in their lifetime, as well. It’s contrary to what people might perceive. Pipelines are actually very safe.
Russel: Yeah, absolutely. What do you think AI might be used for in pipelining?
Jun: The AI can be used for a lot of other applications on a pipeline, such as pipeline operation, efficiency improvement, data validation, preventive maintenance of equipment, and so on. There will be a lot of potential applications on a pipeline.
Russel: You mentioned Internet of Things, which I’ve done a number of podcasts on because that’s the space that I play in. How do you think IoT is going to impact leak detection?
Jun: It would make it a lot more cost-effective and help improve the performance of leak detection systems if we have a lot more IoT available.
Russel: How would that happen, in your opinion?
Jun: If you have a lot of sensors at different locations on a pipeline network, you can detect smaller leaks with much more accurate location estimate and hopefully with less false alarms, as well, because you have a lot more data to verify a change in flow and pressure, whether it’s caused by operational change or a leak if you have more sensors along the pipeline.
Russel: Right. By using IoT, which is by its nature more cost-effective, you can get more sensors. I would assume you can also get more data from the sensors themselves. Then, from that, you can do a better job of not only detecting, but probably more to the point, validating and eliminating false alarms.
Jun: Exactly. I think one of the main challenges at the moment for pipeline companies who invest in leak detection is the cost of sensors. A flow meter could cost you anything from $10,000 to $250,000, depending on the quality of it. That could be one of the obstacles for implementing a leak detection system on a pipeline.
Russel: Exactly. For many of the leak detection approaches that are volume-based or mass balance based, they’re highly dependent on the meters. Then, they become constrained by the accuracy or the uncertainty in the meters themselves.
That is a problem that can be mitigated by adding other highly accurate pressure sensors and other types of things. To me, it’s fascinating. It’s an area we’re going to see a lot of change in the future.
Jun: Yes, in addition to the AI. Of course, you just mentioned using pressure sensor. You could have other technologies for leak detection, like using pressure wave or other methods for detecting leaks as well.
Russel: Right, exactly. Jun, let me ask you this. At the end of this, you were going to leave a word of encouragement or a word of guidance for the next generation of pipeliners, what would that be?
Jun: Very good question. I will say have confidence in the technology and we can make pipelines much safer, even safer than what they are today. Keep it to be the best form of transport.
Russel: What I would say, Jun, is even for me, when I started getting beyond technology and focusing on pipeline safety and what it means to apply the technology to get to a result, I found my passion again. That’s what I hear you saying, is focus on making the world a better place to live in.
Jun: I’m very happy that I have started my career in the pipeline industry. I think it has a long term future for all of us, because it’s still the best for transport.
Russel: Absolutely. Jun, thank you so much for coming on the Pipeliners Podcast and sharing your story. I really appreciate it. It’s been a pleasure.
Jun: Thank you very much.
Russel: I hope you enjoyed this week’s episode of The Pipeliners Podcast and our conversation with Jun Zhang. Just a reminder, before you go you should register to win one of our customized Pipeliners Podcast YETI tumblers. Simply visit pipelinepodcastnetwork.com/win to enter yourself in the drawing.
If you would like to support this podcast, please leave us a review on iTunes/Apple Podcast, Google Play, or whatever smart device you happen to use. You can find instructions at pipelinepodcastnetwork.com.
Russel: If you have ideas, questions, or topics you’d be interested in, please let me know either by reaching out on the Contact Us page at pipelinepodcastnetwork.com or directly to me on LinkedIn. Thanks for listening. I’ll talk to you next week.
Transcription by CastingWords