Episode 20: How SEF Energy is Innovating Oil Field Manufacturing with Automation and AI


Adam Honig: Hello and welcome to Make it. Move it. Sell it. On this podcast, I talk with company leaders about how they’re modernizing the business of making, moving and selling products, and of course, having fun along the way. I’m your host, Adam Honig, the CEO of Spiro.ai. We make amazing AI software for companies in the supply chain, but we’re not talking about that today. Instead, today we’re talking to Tim Marvel, who’s the Vice President Business Development and Technology for SEF Energy, an oil field manufacturing company. Welcome to the show Tim.

Tim Marvel: Thanks Adam, appreciate you having me.

Adam Honig: No, it’s great to have you here. Maybe just tell us a little bit about SEF and what you guys are up to.

Tim Marvel: Yeah, so SEF is a private equity backing company. We’re a holding company that holds two companies. One is Downing, which manufacturers really well heads and pressure control equipment for the industry. And then Oso, which is a perforating company. What we focus on is really manufacturing in the United States, one and two, bringing new technologies to the industry. So we do a lot of innovation and when I talk about innovation, we talk a lot about patented innovation.

Adam Honig: So let’s start by talking about manufacturing in the United States. I feel like there’s been more and more of an emphasis for bringing manufacturing back to the US. What’s your perspective on that?

Tim Marvel: Yeah, so when our company was started about eight years ago, one of the thesis that our CEO had was that we could manufacture competitively in the United States. And so we’ve been executing on that over the past eight years. We started with a smaller company in Oklahoma City, and since then we’ve grown. We’ve added about 140,000 square feet of manufacturing facility there. It’s a beautiful facility. We’re pretty efficient and we have a great manufacturing vice president that’s done a great job there, bringing efficiencies to what we’re able to do. One of the nice things, and you’ve seen this in the last couple years with COVID, is that as supply chains have got strung out, we’re able to respond to our customers quickly because we manufacture right there in Oklahoma City.

Adam Honig: What do you think is stopping companies from doing more manufacturing in the United States? Is it more of like a labor cost thing or just from a mindset perspective?

Tim Marvel: Well I think labor costs one, but I would say more in what we see is labor skills, skill sets. One thing I’ll brag on my brother a little bit real quick, but he’s a principal in a high school up in Alaska and one of the things that they’re bringing is they’re bringing in vocational training into the high schools. This is something that’s, I think, critical, machinist skills and that type of thing. And I think we forget sometimes in our country, that those skills are highly valuable and they pay well. 

Adam Honig: I’m happy to report that last night I went to my son’s open house . My son’s in eighth grade, he’s going to high school next year, and they have a whole manufacturing program at his high school. And they’ve got like local companies they work with and teach things, and they’ve got 3D printers and all kinds of great stuff over there.

Tim Marvel: We need more of that. That’s good.

Adam Honig: Yeah. So you’re focused on manufacturing in the US, tell me a little bit about the thinking around automation in the factory then. Is that a big focus for the organization?

Tim Marvel: What we really automate is out on what we do a lot around frack sites. So we do a lot of automation on the frack site. So if you think of a frack site, when you’re fracturing an oil well, if you think of that as a manufacturing site, which it really is, we are automating a lot on that location. So from job to job to job, we’re automating the functions around that and pulling people out of what are danger zones. That’s where a lot of more of the automation comes in, more so than in our manufacturing, although we have a lot of automated equipment.

Adam Honig: And do you provide the equipment that does the fracking or you guys are actually like setting up the fracking situation yourselves?

Tim Marvel: Well what I’d tell you is there’s three systems. Think about when you come out to a frack site, right, your pumps, you have your wire line, and then you have the surface systems and so we provide the surface systems. So the operator, the oil company right, they hire different companies to do these different services. So we provide one suite of those services out there.

Adam Honig: Gotcha. And so going back to the manufacturing of the United States, you feel like you’re seeing a trend of companies following you guys in expanding manufacturing in the United States?

Tim Marvel: Some yeah, I do think there’s a lot more, and you hear a lot of buzz around that, and you see a lot more people coming back, and I think that’s a good thing for our country.

Adam Honig: But it seems like from what you were saying, that skill development is a challenge that people are gonna run into if they do that.

Tim Marvel: Yeah, I think culturally and skill-set wise, I think we need to be encouraging our people that they have two paths, right? You can go to college or you can go and get a vocational, both worthy and needed skillsets in our country.

Adam Honig: Yeah, and I know it doesn’t have to be either or, right? Because I was up at one of our customers’ manufacturing plant, they had a whole room full of University of Indiana engineers in there fixing all the machines that were on the floor, which are constantly breaking.

Tim Marvel: Yeah, you still have to have both. You marry both of those together, right? We have some excellent operational people that really understand the operations around whether or not it’s manufacturing or like I said, treating the frack site that do an excellent job and then you marry that with the engineers that can engineer out some of the processes or engineer out some of the problems or the hazards, that type of thing.

Adam Honig: Well, I talk with a lot of people, I feel like I’m hearing a lot more about people building or expanding facilities in the United States for manufacturing especially. And I think what you mentioned before about being close to the customer, shortening the supply chain so that they’re able to get all the materials and everything that they need locally. It’s just who knows what’s gonna be happening in the Pacific in the future, I think, is on a lot of people’s minds.

Tim Marvel: I’ll tell you another thing that’s really nice about marrying your manufacturing with your engineering. So we are able to engineer something and then we walk downstairs and we’re able to build it, and then our engineers are able to see what issues are, that type of thing. And so your speed of innovation is much faster when you have manufacturing there versus let’s say a six month supply chain that you make something and six months later it shows up and you realize, oops, I probably shouldn’t have done this. And now you gotta go back to the drawing board. So our speed of innovation, which is one of the keys of what we do, marrying manufacturing and innovation together really helps us.

Adam Honig: When people think about oil drilling and fracking, I don’t think innovation is the thing that comes to mind. I think that they think of guys wearing hats and boots and stuff like that. But you’re right, if you look at fracking just in general, like we couldn’t do any of that work 20 years ago. That’s all brand new technology.

Tim Marvel: Well especially in the shell, being able to frack a non porous rock and get oil out of it, that is definitely an innovation the United States has led on. So I’ll give you an example on the drilling side. So you just get an idea of the technology, but we are drilling down 10,000 feet, so let’s say two miles, and then we’re going out three miles from the spot you’re in. So imagine drilling underneath the city while you’re out on a farm. So you drill underneath that city and we can pinpoint within feet of where we want to hit on that wellbore, let’s say a six inch wellbore when you get out there. So there’s quite a bit of technology, lots of pressure, lots of temperature. So when you look at the hurdles from an innovation standpoint, you’re right up there and honestly, with what everybody thinks about it, the space, that’s a challenging environment, but underground’s very challenging.

Adam Honig: Yeah, I’m trying to even picture in my mind how that would work. I’ve got one of those cutaway pictures of the earth with a drill going underneath the city. It’s gotta be actually a lot more challenging underground than out in space. If you’re in space, there’s nothing to really bump into unless you get pretty far out.

Tim Marvel: That’s right. It’s all the same atmosphere. Or the other conditions around you whereas down hole, you are going through all these different formations. And then you have caverns and you have water and pressures and temperature, geothermal, whatever you name it, we got it. So it’s actually pretty fascinating when you get into it.

Adam Honig: It’s so funny. So I spend a lot of time talking to people about software development as well, and people sometimes can’t really understand why we have trouble doing estimates for software development. I always explain, well it’s kind of like digging a pool in your backyard, until you start digging, you never know what you’re gonna hit, right? But in your business, it’s literally like the same thing.

Tim Marvel: Yes, well we don’t always know what we’re gonna hit when we go down a hole. And I’ve worked on different sides of the industry. So on the drilling, there’s challenges and then when we complete the wells to be able to get the oil out of the ground. I think a lot of people think that you just poke a hole in the ground and that oil comes out. Those days ended a hundred years ago, so now you have to go down, you have to drill, and then you have to stimulate what they call the rock around you to get the oil and gas out of the hole.

Adam Honig: I saw you were with Baker Hughes, was it at the start of your career? I did a little bit of work consulting for Baker Hughes at the start of my career and it was the first time I learned anything about oil drilling, except in like a Bugs Bunny cartoon when he drills a hole and gets some oil. And I think that’s the way most people think about it, you know, you just boop, put a thing down and kind of there you go.

Tim Marvel: Yeah, absolutely.

Adam Honig: Tell me a little bit about the innovation side of things. What are some of the innovations that you guys are working on now?

Tim Marvel: I’ll talk about a couple of them, but they’ll give you a good example. So when we frack a well we actually plug off different zones underneath that. I talked about that being three miles out, two miles down, three miles out, right? We go down there, we isolate an area, and then we blow holes in the casing. There’s a steel pipe down there, we blow holes in that, and then we pump fluid down the hole. And when we do that, it’s like a 10,000, 15,000 PSI to fracture all that rock down there. Every time we isolate a certain section of that three miles, we do that, let’s say 80 times. Well there’s transition times on the surface. So in those transition times and all this equipment, and you have people in what’s called the red zone, where you have, like I said, you have pressurized iron. So I don’t know if you’ve ever seen anything that goes off at 10,000 psi, but it’s not a pretty site. You don’t wanna be around that.

With all that’s going on there, we automate a lot of the functions that were going on where, let’s say, every time we frack a zone, it would take us 45 minutes to transition to the next zone, to the next zone, to the next zone. We’ve got that down to about 30 seconds now. And all of that is through automation, right? So we’re automating a lot of the functions, control systems, we’re using tons of data that’s coming in. We’re streaming data to the Cloud real time that we then use to further optimize what we’re doing. We can do remote operations into a frack site. So for instance, we had a time where we had a little issue on site not related to us, I’ll make that clear, where we had to close in a well. We closed it in, it was in Pittsburgh or up in the northeast area and we closed it in from Dallas. So those are the types of things that we’re innovating really around lots of software and control systems that we’re bringing into the industry.

Adam Honig: Gotcha. Are you guys using any AI techniques to be dealing with all that data? Because if you’re capturing all that data, that’s something that the AI can really help with.

Tim Marvel: Oh yeah. We get a lot of calls from our guys around AI, yes. So we’re doing a lot of analysis, a lot of algorithms. We have very high frequency data. So there’s things that we can see in that data that people didn’t know about before. And now we’re learning and we’re using techniques. We have a PhD in signal processing that works a lot with what we do. So when you talk about innovation, this kinda gets exciting, Adam, this is where I get pumped up. You talk about the things that we’re doing and the data that we’re utilizing, that people before didn’t realize that this stuff was actually happening and what you can see in that data, so it’s really exciting.

Adam Honig: I know it’s a little technical, but is there something you can share that maybe a more general audience might be able to understand about the kind of insight that you get from the data. 

Tim Marvel: I’ll give you a couple examples. One, we get a lot of perceptions. Hey, that thing takes 20 minutes to do that operation, right? We get these customer complaints, well then we go in and we actually disaggregate all that data, bring it down, tear it apart, and look at just that function, whatever that is in the mass of all this other data. And then we show them, well, in actuality, yes, we did have one event that was 20 minutes, but out of 310 times that we did this, that was 20 minutes once, the rest of it was 30 seconds. So those types of things changed the conversation with the customers. And then the other thing is, okay, that 20 minutes, why did that happen and what do we need to do to fix that? And one of the things I’ll add to that, Adam, is that we talk a lot about human workflows versus automated workflows. In our industry, when you have a human workflow, they go through a bunch of procedures, checklists, that type of thing. And remember what I said at the beginning, you are working with a lot of different companies out there. So when something goes wrong with human workflows, what do you get? So you try to find the root cause that everybody does this, you know, they’re all pointing at each other. And then once they do find a root cause or think they find a root cause, invariably it has to do with them adding more checklist procedures, we gotta train better, right? And one of the things I always talk around is you can never out-train your problems. I mean it’s impossible. You talked about automation, right? If we have an issue, we get one second data, that one second data that’s coming in, and that’s our slower speed data. We get a lot of faster or higher hertz data than that, but that one second data, we can tell exactly what happened. Even if it’s on us, we know what happened, we can tell the customer this is what happened, and then here’s the key. We look at that and we engineer it out. And guess what, whatever that problem was, doesn’t happen again. Other problems might happen, but that one doesn’t happen again.

Adam Honig: So I’m hearing kinda like a quality control feedback here, that you’re using the data to spot the problems and then instead of adding a human solution, use automation in a way to get rid of that problem. And obviously there’s gonna be other problems, but at least that one won’t happen again.

Tim Marvel: Yes, take your data, take your AI, take your control systems and work through that problem and get rid of it.

Adam Honig: It’s gotta be challenging though to be dealing with a number of different companies all at the same site.

Tim Marvel: It is, but here’s the deal. So what we’ve done, we’ve taken a different approach, instead of marrying all these surface systems together that don’t talk to each other, and the interaction between them is a human being. We’ve taken all those, we’ve connected them, interlocked them through one control system so now we have feedback from every one of those systems. And so we know exactly what’s happening in that surface system. So that goes a long way versus what we had traditionally, we have five, six different control systems out there that don’t talk to each other, that only interact through a human being. So there’s no way to get rid of that problem.

Adam Honig: It doesn’t sound like you’re a manufacturing company, sounds like you’re a technology company at this point. I mean, you guys are almost a software company even. 

Tim Marvel: Well, we have a lot of that. Yeah, we are a technology company and we do a lot around technology. We have some very good PhDs that are in control systems, signal processing, and they do a fantastic job of innovating.

Adam Honig: I think tying it back to manufacturing, though, I think this is something that the United States has a big advantage in. If the future of manufacturing is really the smart manufacturing or the AI, augmented manufacturing, I mean, we might have a shortage of skills in some of the areas, but I think in that area we can do pretty well.

Tim Marvel: I agree. I’ve been very fortunate, like with Baker Hughes, worked around the world. I definitely am a believer in the quality of the people that we have and the innovation that a lot of our people have the ability to think through problems.

Adam Honig: Let me ask you this, one of the things we’ve been talking with people about is changing expectations of customers. Somebody coined the phrase “the Amazon effect,” everybody expects you as a business to be as quick to deliver products and finish projects and stuff like that as Amazon often seems to be. Do you feel like customer perceptions are changing in your industry?

Tim Marvel: You know, it’s interesting. So our industry, there’s a lot of de-risking, right? You know, when you talk about technology, they try to get these processes in place and then they try to do them over and over again. So to break into that cycle, it takes a little bit of convincing to try something like what we’re doing, which is completely different than what everybody else does. So we talk a lot about this when you’re introducing new technology, and Adam, you probably know this because you’re doing the same thing, but when you talk around perseverance, grit, it takes a lot to get people to think through and say oh, you know what, I can do this differently. I don’t know if I answered your question exactly, but that’s what we see as one of the issues is really driving change to what’s traditionally done.

Adam Honig: No, you answered a much more interesting question for me, which is sure, if you’re gonna do something different in the drilling industry, how do you get the customer to feel comfortable with that? I mean, there could be catastrophic results if things don’t work out the way that they’re supposed to, right? That’s gotta be a big hurdle to get over.

Tim Marvel: Yes. But let’s go through what you just said, I think that’s really good. So we talk a lot about the risk associated with the old way versus the new way, they may have gotten used to some of the risks on the old way. Where we have getting rid of a lot of those risks associated with whether or not it’s people, error, that type of thing. One of the things I talked about earlier, we’ve interlocked these systems together. So now if one system has an issue it alerts the other and if the condition isn’t correct, the computer won’t go to the next step. So it actually is a safer system than doing it manually, which is what we’ve done historically.

Adam Honig: Gotcha. So even though change can be perceived as risky, what you’re showing is actually the whole system is designed to lower the risk.

Tim Marvel: Designed to de-risk.

Adam Honig: I assume with your buyers, that minimizing risk is a key element for them.

Tim Marvel: Yes, absolutely. So I’ll just give you some numbers. If a typical well, and I’m gonna get this wrong, but let’s say it’s between two and a half million to 5 million to 7 million, 8 million to complete one well, there’s a lot of risk associated, dollar-wise on that and the returns associated with that. Well, one of the things we have to do is talk to our customers about de-risking that and getting rid of some of those costs.

Adam Honig: Gotcha, fascinating. I’d like to ask people about maybe an unusual or different kind of situation that you’re in. Is there something that comes to mind when I mention that?

Tim Marvel: Really one of the things that we’re innovating is around trying to do a system that you can fracture 24 hours a day, seven days a week. And when I say fracture, meaning fracture the rock. So today, efficiency, on a 24 hour day is let’s say 16 hours typically. Which means that you have a typical frac fleet, let’s say costs between 25 and 40 million dollars. So that’s the capital that you have out on a frack site, so eight hours a day that you’re not using it to do what it’s intended to do is not good.

Adam Honig: You’re just eating that cost.

Tim Marvel: Eating that cost, right? So what we are doing is trying to come up with systems that allow you to get greater utilization of your equipment. And so I mentioned one earlier about going from that 45 minutes to that 30 seconds between the different zones. The second one that I would highlight for you is when these frack trucks are out there, there are 18 to 20 of these frack trucks out there that are pumps that are pumping that 10,000 psi down the hole, right?  But they go down, they last maybe 5, 6, 7 stages or zones and then they gotta pull them out. And typically, what you have to do is you have to shut down the whole frack site, bleed down all that pressure, pull those out, and then we gotta fix all those trucks. What we’re working on is being able to actually pull a truck out through automation while the rest of the trucks are pumping. So imagine a system where nobody can go in the red zone, have to keep people out of there, you have to be able to unlatch and latch quick connects. You have to be able to drain this fluid, bleed off the pressure, and then you have to be able to back a truck back in, reattach, so we have arms going out, grabbing, couplings, you know, coupling these together, clamping them.

Adam Honig: I’m seeing the world’s biggest NASCAR pit. That’s envisioning, yes.

Tim Marvel: That’s right, without the eight people around it.

Adam Honig: Right, exactly, just robotic or somehow, yeah.

Tim Marvel: Yes. So that’s one of the exciting things I think that we’re working on right now.

Adam Honig: Yeah, that’s amazing. Well Tim, this has been awesome, I feel like I’ve learned so much about fracking I didn’t even know was possible before and three miles under and two miles down. That’s crazy stuff, man. That’s so impressive. Really love hearing about that. I love hearing about the manufacturing in the United States, definitely a big believer in that. And hopefully we can get more people coming outta school to be interested in pursuing that as a vocation for sure.

Tim Marvel: Yeah, we’d love that.

Adam Honig: I think that’s what we need here, but no, it’s been awesome, so I really appreciate your joining me on the podcast.

Tim Marvel: Adam, thank you very much for having me, really appreciate it. It was a lot of fun.

Adam Honig: So as a reminder, you can find every episode of the Make it. Move it. Sell it podcast at Spiro.ai/podcast, be sure to subscribe. And I don’t know, Tim, maybe people should give us a thumbs up or a like, or something like that. What do you think?

Tim Marvel: Absolutely. Good review, yes.

Adam Honig: We need a good review from you Mr. Listener or Miss Listener, so thank you for that. And of course you can always subscribe if you like, but I’d like to thank everybody for tuning in and we look forward to talking with you on the next episode.