In this episode of the Oil & Gas Measurement Podcast, host Weldon Wright talks with Emerson’s Oil & Gas Manager, Stan Calame, about evolving trends in upstream measurement and the challenges shaping the future of the industry. The conversation explores how advances in digital technology, data-rich measurement devices, and edge computing are transforming field operations, while also examining workforce shifts, training needs, and the importance of maintaining data accuracy. Together, they offer insight into how technology and experience must work hand in hand to ensure reliable measurement in modern oil and gas production.

Upstream Measurement Trends Show Notes, Links, Insider Terms & Full Episode Transcript

• Stan Calame is the Oil & Gas Manager at Emerson. Connect with Stan on LinkedIn.
  • Emerson helps manufacturers achieve top quartile business performance through the industry’s broadest portfolio of technologies to measure, control, optimize and power their operations – and the experience and expertise to solve their toughest problems. From reliable, easy-to-use, and innovative products and software to responsive, expert services, Emerson combines advanced technologies, industry-leading expertise, and an insatiable curiosity about the world around us to create sustainable solutions for the essential customers we serve.
• Upstream – The segment of the oil and gas industry focused on exploration, drilling, and production activities occurring near or at the wellhead, prior to transportation or processing.
• Midstream – The industry segment responsible for transporting, storing, and sometimes processing hydrocarbons between production and refining or distribution.
• Downstream – The sector covering refining, processing, distribution, and sale of petroleum products to end users.
• Measurement – The process of quantifying oil, gas, and water volumes or flow for operational control, revenue allocation, and regulatory compliance.
• Wellhead – The assembly at the surface of a well that provides structural and pressure-containing interface for drilling and production equipment.
• Production Separator – Equipment that separates a well stream into its component phases—oil, gas, and water—allowing for accurate measurement and processing.
• Turbine Meter – A mechanical device that measures fluid flow by detecting the rotational speed of a turbine within the stream; traditionally used for liquid measurement.
• Orifice Run / Orifice Meter – A differential pressure device that measures gas flow using an orifice plate; widely used in custody transfer and regulated measurement.
• Coriolis Meter – A high-accuracy flow meter that measures mass flow and density based on the Coriolis effect; provides detailed diagnostics and multi-phase detection capability.
• Ultrasonic Meter – A flow measurement device that uses sound waves to determine flow velocity and calculate volume or mass flow, common in gas measurement applications.
• Shale Boom – The rapid growth of oil and gas production from shale formations due to advancements in horizontal drilling and hydraulic fracturing technologies.
• Haynesville Shale – A major natural gas-producing shale formation located primarily in Louisiana and East Texas.
• Decline Curve – The rate at which a well’s production decreases over time; crucial for estimating reserves and forecasting economics.
• Fracking (Hydraulic Fracturing) – A stimulation technique that injects pressurized fluid into subsurface rock formations to create fractures and enhance hydrocarbon flow.
• Telemetry – The automated collection and wireless transmission of measurement data from remote field devices to central systems for monitoring and analysis.
• SCADA (Supervisory Control and Data Acquisition) – A control system used to collect real-time data from remote equipment for operational monitoring, control, and analysis.
• Edge Computing – Processing data locally at or near the source (e.g., at a well pad) to reduce data transfer loads and enable faster decision-making.
• Data Fatigue / Paralysis by Analysis – A condition where excessive data volume overwhelms operators, making it difficult to extract actionable insights.
• Lease Operating Expense (LOE) – The ongoing cost of operating an oil or gas well, including labor, equipment maintenance, and utilities.
• Flow Computer – A specialized device that calculates flow rates, volumes, and other parameters from meter signals and applies industry-standard calculations.
• Chromatograph (Gas Chromatograph) – An analytical instrument used to determine gas composition, enabling accurate energy content and flow measurement calculations.
• Single-Run Flow Computer – A flow computer managing data from one measurement run; early units in the 1990s began integrating pump control and tank-level monitoring.
• Data-Rich Devices – Modern field instruments capable of providing large amounts of operational and diagnostic data in addition to basic flow information.
• Two-Phase / Three-Phase Flow – The presence of multiple physical phases (gas, oil, water) in a production stream, which complicates measurement and separation.
• API Gravity – A measure of oil density standardized by the American Petroleum Institute; higher API values indicate lighter, more volatile oils.
• Allocation Measurement – The process of dividing production volumes among multiple wells, zones, or royalty owners based on measurement data and production ratios.
• Separator Efficiency – The effectiveness of a production separator in fully separating oil, gas, and water phases, impacting measurement accuracy.
• Facility Engineer – An engineer responsible for designing surface production facilities, including separators, tanks, and flowlines, to handle varying production rates safely and efficiently.
• Installation Practices – Field techniques ensuring proper alignment, torque, and orientation of equipment such as meters and valves to avoid mechanical strain or measurement error.
• Calibration – The process of verifying and adjusting measurement instruments to ensure accuracy within defined tolerances.
• Diagnostics – Data and signals from smart measurement devices that identify operational conditions, faults, or process abnormalities.
• Artificial Intelligence (AI) – The use of algorithms and machine learning to interpret operational data, predict failures, or optimize production performance.
• Machine Learning – A subset of AI that enables systems to improve predictions or pattern recognition based on historical data without explicit programming.
• Human-in-the-Loop – The concept that human expertise remains necessary to validate, interpret, and act on automated or AI-generated insights.
• Custody Transfer – The point at which ownership of hydrocarbons changes hands, requiring highly accurate and auditable measurement standards.

Upstream Measurement Trends Full Episode Transcript

Weldon Wright:

Welcome to episode 49 of the “Oil & Gas Measurement Podcast,” sponsored by EnerSys Corporation, the provider of Muddy Boots field operations software, assisting oil and gas operators to implement accurate and repeatable measurement activities across their entire operation, while also simplifying data collection from the field.

Find out more about Muddy Boots at enersyscorp.com/software/muddyboots.

[background music]

Announcer: Welcome to the Oil & Gas Measurement Podcast, where measurement professionals, bubba geeks, and gurus share their knowledge, experience, and likely a tall tale or two on measurement topics for the oil and gas industry. Now your host, Weldon Wright.

Weldon:

Welcome. My name’s Weldon Wright with FpvPrime Measurement Consulting. I have Stan Calame on the phone with me, the oil and gas manager over at Emerson.

We’re going to talk a little bit today about upstream, some of the issues there. We’re going to cover a lot of ground. As an underlying theme to several of the podcasts before and after this one, we’re probably going to touch a little bit about how our industry’s hunting for new people, smart people, into our industry.

More important than that, we’re recording this the first week in September. We have launched into football season. Both of us hail from that same little school down in College Station. Although I think that Stan was a little bit better, I was at A&M during some of the dark Jackie Sherrill years. Stan came in and did a little bit of Slocum.

Stan, welcome to the podcast. Tell us a little bit about yourself, and what you do over at Emerson.

Stan Calame:

Thanks, Weldon. First, I got to start by saying whoop, and the Aggies are 1 and  so far, so we’re not going to get too far ahead of ourselves, but that’s always a great way to start September, so I’m pretty stoked.

My career path, it’s a long one but it’s pretty simple. I got out of A&M in 1989 and went to work for Emerson straight out of school and been there ever since. Celebrated 36 years this June. Spent a little time in Baton Rouge, Louisiana, up and down the ship channel, and migrated out to the East Coast in Raleigh, North Carolina.

Then the holy land was calling me back to Texas. Came back, had my kiddos, and have been here ever since in the Dallas area.

My role as oil and gas manager, I like to tell people it’s as much cat herding as it is anything else because I have the privilege of working with a number of our different products, and we have a direct sales channel and a representative channel, our impact partner network.

Emerson, as you probably know, has a lot of different products, from dump valves to DCS — pretty much, if you can measure it and control it, we make it — which is really fun. It’s kept me engaged for decades. It’s quality stuff. It’s been a good ride to work with customers.

Of that 36 years, about half of it, I would say I’ve been hydrocarbon conscious, because I really didn’t get into the oil and gas space until about 2008 when the shale boom was hitting.

The Haynesville shale was popping off. I had always been in the Dallas area and been responsible for the eastern part of the state. As the shale boom moved its way into the oil space, I found myself frequenting Midland, Odessa quite a bit, which is sort of the center of the universe.

Weldon: Holy land?

Stan:

Yeah, the holy land. I still spend a lot of time in Midland, Odessa and parts west of that in the desert out there. I’ve heard people say it’s the three Ds of oil and gas are dull, dirty and distant, and that certainly applies to some of the places that we go out there.

It’s been a great industry. It’s been a great learning curve. I still feel like I’m learning every week, it seems like. I think I try to adopt that philosophy because I think if you’re not learning, you’re probably not trying hard enough because I don’t think we’ve figured everything out. Even with AI, there’s still a lot to discover.

Weldon: I would agree 100 percent. Anybody that says they know everything in their industry is someone you should not be listening to.

Stan: [laughs]

Stan: That’s specific.

Weldon:

I call myself semi-retired with very, very heavy on the semi. I, really, drive bird dogs around going hunting these days. Still, you can’t stick your nose in this industry without being exposed to stuff that’s new and different. That’s what I would like to talk about as we start off this morning.

I’d like for you to talk to us a little bit about an overview level of what have you seen change in this industry? How have we changed technology, equipment, and people-wise?

As I mentioned earlier, I know, as we talked prior to actually starting our recording today, we share a common concern. I think everybody in this industry that’s been here a while shares it. That is, this industry, from the wellhead to the burner tip, as we like to say, is hurting for high-quality new folks coming into the industry. Right?

Stan: Absolutely.

Weldon: Let’s talk about that, how you see things change, have changed, and where you think we’re headed.

Stan:

I can think back, if I go back in the time machine, some of the first times I would be on a well site I would see a production separator, and on the liquid side of the separator I saw a turbine meter, and on the gas leg I would see an orifice run, and you go to a field office and there might be 15 or 20 trucks because you had lots of humans out there maintaining those pieces of equipment.

It was a busy time. Over the years I’ve seen, I guess with the advent of shale, just the characteristics of the wells, the huge decline curve that there’s been a huge emphasis on getting quality data, accurate data, data that you can trust in the very early part of production. Because in many times that’s where the well is paying for itself, whether it be a gas well or an oil well.

It’s a huge decline curve. You really need confidence in the data because, a lot of times, you’re going to be planning on fracking or developing that rock based on the results of the well you just produced. It’s a pretty quick churn, from what I can see.

I’m not a reservoir person or downhole guy, but I know that I’ve seen an intensity of the need for good-quality data. I think what that’s done, it’s forced the advent of adoption for newer technologies.

I mean, newer, they’ve been around for decades. Probably the two largest sectors that I’ve seen adoption are in Coriolis measurement, both on oil and gas and ultrasonic measurement.

Those have been around for years, but where we used to see those primarily in the midstream space. Those have been migrating for the last 20 years into the upstream space, and now are pretty much commonplace because of the data that they give you, aside just from the process data, but they give you insight into really what’s happening in the pipe, a window into the process, you might say.

With that comes a little more complexity from a technical standpoint. You’ve got, and across the board, like you said, if you ever listen to Mike Rowe, he really does a good job of amplifying that message that we’re short on the trades and the generations that we have today were raised on tablets and screens. Nothing wrong with that. The skill sets are different now.

You need to be able to digital data rather than analog data. As humans, we’re evolving to that space where we’re more accustomed to digital, instant data and that’s how we make decisions. I think that’s driven a profile of workers and it’s also driven the technology. Just the advent of that.

Weldon:

What you said earlier about when you started in the industry, you’d go out there to maybe 20 people on a production site. I started in this industry back in high school as a junior and senior hooking up pump-off controllers and pump jacks in the Austin Chalk formation.

It was all, there was a pressure gauge at the top of the well and we gauged the tank and there was a gas meter. We struggled in those days because there wasn’t digital information transfer. Right?

Stan: Right.

Weldon:

Now we have telemetry, now we suck some of it in.

You mentioned something earlier, before we started recording that I’d really like to focus on for a couple of minutes. That is, as you talk about, we’ve made this change to digital technology. New folks coming into the industry want to be data-related. They want to work on that tablet and not stick their head inside of a motor controller. Right?

Stan: Right:

Weldon: You mentioned that we have so much more data now. You said a couple of things I would like you to loop around back to. One is that our technologies are data-rich today, but the other is we have so much data. Your comment, before we started recording, was that we’ve always got to be cognizant of our first question needs to be, are those numbers correct? Are they accurate?

Stan: Absolutely.

Weldon: Can we look back around and talk about that some more?

Stan:

Sure, yeah. I see it all the time. Wherever there’s a number, there’s somebody asking how do I trust that number? Or, they’re comparing it to something else. The great thing about these devices these days is that they’re data-rich. The bad thing about that is we don’t have a pipeline big enough to push the data through.

The primary concern of operators is to operate safely and keep people safe and make sure we’re producing and generating the maximum revenue. The devices are, if you want to think about it, they’re little data servers.

I’ve had an arc of, I guess, understanding early on, because a large part of my career I was selling Coriolis, and it’s just so data-rich. When you’re in a plant connected to a DCS, you can get one-second data, that’s no problem.

But when you’re on a well pad, and if I give somebody 25 registers I’d say, “Yeah, you need to pull these once a second,” and I couldn’t understand the look in their face when I was telling them that and they’re thinking, “Look, you’re going to get lucky to get one poll every 15 minutes.”

It took me a while to figure that out that the device I was selling was not the only thing that they were contending with. As good as that message was, they couldn’t afford to pull that much data.

Now, I think we are at a spot now in technology with edge computing where we don’t have to send all that data through a SCADA system. In fact, there’s no need to. It’s clutter and it creates data fatigue, I think — paralysis by analysis.

I think if you can manage that data on the edge of the pad in an industrial computer, you can do so much to leverage that stranded data. You can start to see when your separator might not be efficient. If you’re getting slugs of liquid and you’re starting to get gas in your oil or liquids in your gas, you can actually see that on a second-by-second basis.

If you can act to mitigate that when it occurs, then you dramatically improve, number one, your measurement because it’s going to be accurate, and, number two, you’re going to mitigate expenses that are accumulating because of lease operating expenses.

You’re not going to be reworking oil that’s got a bunch of water in it. You’re not going to risk shutting in your gas because it’s got liquid in it. There’s ways now that you can detect that in real time.

The other thing that works against this is we have fewer humans doing this work than we used to. Companies aren’t hiring human resources at the same pace that they’re punching holes in the ground. I don’t think they ever will.

In fact, the ratio is probably going to become more wells to humans as we go. Technology is such that they’re drilling wells faster, completing them faster, and the well-per-human ratio is probably going way up, I would say, in almost every producer you talk to.

Weldon: I agree.

Stan: The only way to manage that is through real-time data, and not just data, but you have to take something meaningful from that. Otherwise, you’re just drowning in ones and zeros, and that’s not sustainable.

Weldon: You have to process the data.

Stan: You have to have actionable information to be able to make decisions to maximize your operation, if you ask me.

Weldon: You have to process that data. You have to consolidate, get that data down to be more concise, more pointed to where the fewer people we have can act upon it. At the same time, we’ve got to be aware — we all see, we all look at the YouTube videos and stuff about AI gone astray — AI can only be as intelligent as the learning tools we give it to learn with the information we feed it. Right?

Stan: Exactly.

Weldon:

If we don’t have the industry legacy knowledge for that to help train these AIs, how are they going to replace the folks we had in the field?

I know we’re focused more on the upstream part of the industry in our discussion here today, but I know in measurement, when we talk about oil and gas measurement, upstream is a little bit special when it comes to the techs that are out in the field.

Once we get in the midstream world, a measurement tech is primarily concerned with measurement. He may worry about a tank level here and there, but our plant I&E folks are worried about instrumentation of the plant that midstream measurement guys test the meters out the wellhead.

When we get into the pipeline, and our measurement techs on the pipeline are very much focused almost exclusively on measurement devices. Maybe there’s a valve actuator sometime, maybe there’s an analytical equipment.

When we’re on the wellhead side, the guy on that wellhead that is taking care of the measurement equipment is touching a lot of stuff that’s not just strictly measurement. Right?

Stan: Oh yeah.

Weldon:

I think we forget that, in the oil and gas measurement portion of the industry, we’ve got to be aware that, especially on the upstream side, there’s numerous devices that need to work alongside of, and in many times they work best with our measurement equipment.

I can remember back in…What are we going to talk about here? Are we going to talk about ’92 or ’93? The first of those single-run flow computers that…Oh my god. They were going to do pump control also.

Stan: [laughs]

Stan:

It was an amazing thing. A few years after that, we brought the first condensate and water tank levels in via that. That was just revolutionary. You alluded to this earlier. We were in the world there to where we were proud of ourselves we had hourly information. We were bringing in four or five points.

A really high-dollar installation, maybe there’s an online chromatograph, and we’re bringing in 15 pieces of information every hour. You alluded to this earlier, the new devices we have today, these data-rich end devices, there’s massive information there, isn’t there?

Stan:

Oh, yeah. Maybe a keyword there is being able to contextualize, because you got five pieces of data and they’re all giving you a terabyte of data a day, and so how do you stitch all that together and understand the message that it’s telling you?

We see this all the time in some of these shell wells where the oil is very volatile. You’re talking API gravities in the 50s, maybe even the 60s. It’s almost NGL. To your point, you have to operate a dump cycle almost like a control loop, or at least you have to understand what each component is going to do to impact that measurement.

In the old days, you may just dump liquid to a tank and then gauge the tank. Nowadays, you’re commingling many wells, and if you want to allocate properly to those royalty owners, you have to have confidence in each one of those meters.

Guess what? The liquid may flash if you exert too much pressure drop on it. You have to control maybe the rate at which you dump. That’s determined by the valve that you select and the level control that you use.

If you want to have good measurement, you got to look at it as a system. It’s more than just a meter. It’s how’s the entire system working? To your point, I think the closer you get to the wellhead, the more jack of all trades you have to be.

You have to be masters of those because most of the times when I get trouble calls, there’s nothing wrong with the instrument. It’s really telling you something is going on in the process.

Sometimes, the lease operators, they’re calling a measurement guy to troubleshoot a measurement problem. He gets there or she gets there, and then they say, “No, your level’s too low. You got gas blowing through the meter,” or, “You got a liquid carryover,” or maybe the separator’s filled up with sand, and you’re not getting good separation because you don’t have enough resonance time.

The meter is giving you a symptom, but it’s a symptom of a bigger problem. The meter’s not the problem. It’s the process is what’s going on it. Now we have the ability to extract that insight out of that data. It’s really exciting, but you’re right, it takes…

Weldon: It’s exciting and dangerous at the same time.

Stan: Yeah. [laughs]

Weldon:

We talk about we have smarter devices out in the field and our data-rich technology. These Coriolis meters, these ultrasonic meters that provide us greatly increased diagnostic capability in the field, or even operating data outside of just diagnostics.

We can do things with this technology that we have never done with the old orifice meter. Now that doesn’t mean the orifice meter couldn’t do some of those things. We just haven’t done it. As we’ve moved into these high-dollar measurement devices, don’t mean to be beat up on any salesmen at any specific company because this is industry-wide…

Stan: [laughs]

Stan: Sure.

Weldon:

the push has always been. I’ve sat there in that office and made the purchasing decisions. I’ve written the specs.

Worked for a while for a pretty good-sized pipeline company as a director there. That salesman shows up in your office talking about the new technology, “You’ve got to try our new model Z of brand X meter. It’s smarter. You’re going to need to calibrate it less. You’re going to need people on-site less.”

Management has very much bought into our morphosis of technology at the wellhead and on the pipeline. We’ve morphed into more complex, higher-dollar measurement equipment, but management has bought into that with this drive that it means we’re going to need less people. We’re going to need to go to every site less often to calibrate it. There’s going to be less maintenance. There’s less cleaning.

When we’re doing that, that falls right into what you were saying earlier about so many of our problems are process-related right now. If we’re not putting people on the site to really see what’s going on on the site, we’re missing a big piece of that.

Using the technology, using those edge devices in devices to process and pre-process data for us, that’s only as good as how intelligently we build a system. Right?

Stan:

Absolutely. To your point, we’ve talked about the shortage of human resources. You can’t afford to get there and spend a bunch of time troubleshooting. You need to know what’s going on when you get there because at the rate people are producing today, downtime is…In oil and gas, if you’re not producing, you’re not making money.

Whenever you do deploy a human to check on a location, if you have a problem, they need to have at least a top three ideas of, “What do I need to remedy this problem? Do I have the parts on hand to fix it?” It’s just like any other thing that you would call, get repaired, diagnosis is the key to resolving the issue. You have to know what the root cause is.

If you’re hiring good people, if you’re massaging that data and contextualizing the data that you’re getting from all these devices, if you do that well, then you can accomplish that.

You can arrive at location understanding what needs to be done to mitigate the problem and to move on to other things because you just don’t have the luxury of sitting on a location for a day and trying to figure out what’s going on.

I see people all the time. If they don’t think the meter’s working, they’ll just rip and replace, and if that doesn’t fix it, all that time you’ve just invested, you lost because you still haven’t fixed the problem.

Weldon:

As you mentioned earlier, a lot of the time, this is not a measurement hardware problem. We claim we’re sending single phase, we’re sending only gas or only liquids through a meter. Yet, we really have two-phase or three-phase. I use three-phase fairly often. We need to clarify that a little bit.

When we say that in the measurement world, when we talk about wellhead…This is for my listeners that are in the pipeline side and the downstream side of our industry. When we say three-phase in the context of measurement at the wellhead, we’re talking about natural gas, oil, and water are the three phases we talk about.

We’re not talking about solids, although sand can go through a meter sometimes. When we say three-phase, we’re usually talking about water in addition to the gas and oil. Now we can handle those things if we’ve selected the right equipment. If we’ve designed our installation and if we’re operating our installation properly, we can handle those, right?

Stan:

Right. The thing that I’ve seen with facility engineers, and I really respect and appreciate their task because they’re forced to design a facility and a vessel that’s capable of not only accommodating the huge initial production, which by the way, is when the well is paying for itself, but you also have to have it properly sized whenever the well is depleted.

You can’t spend the type of money upfront to get a huge vessel because that’s not a wise deployment of capital. There’s a Goldilocks effect there. It’s got to be just right. Sometimes it’s going to be too small for initial production and probably too big for depleted production, but you got to find a spot where the economics make sense.

To help you with that, this technology, it can actually fight through some of that multi-phase measurement. You’re going to have a little higher uncertainty, but the technology is such now that you can run oil and water through a Coriolis meter, and you can tell how much oil and how much water is in it. You can still get good allocation measurement.

Yeah, you’re going to have to clean the oil up downstream, but at least you can deal with that initial production upfront, and get reasonably good measurement so that you can feed that to your reservoir people and your production people.

Because again, like I said earlier, they’re using that data not only to calculate revenue, but they’re deciding what they’re going to do on the next completion and how they’re going to drill and frack the next well, so that’s important data. If it’s wrong, then, with bad data, you’re going to make bad decisions. It’s always good to have high-integrity data, I would say.

Weldon:

One of the key underlying things — and you mentioned the importance of that facility engineer designing that — it takes a village. Nobody can do this, and any one person in the chain or one piece of equipment in the chain that’s not doing what we expect it to do, and that’s the keyword, what we’re expecting it to do, we get in trouble.

We rely on engineering, the facility engineer, pipeline engineer, whatever we’re calling them. We rely on that to design a facility based upon good information.

We’ve got to know, now we’ve moved further upstream of the upstream process, now we’re looking at reservoir engineers, we’re looking at information from the actual well drilling logs, even we’re looking at the initial well test, just for that facility engineer to have the right information to design a good installation.

Then we have to select the right products. Then we have to install those products correctly. Just because the engineer drew it the right way and we’ve selected the right product, we don’t draw in every valve, every nipple, every slope. Installation and the knowledge and experience to install properly is important.

A good friend of mine tells a tale of a very, very expensive Coriolis meter on an offshore platform, and the many trips to fly many people out to that platform. All it was is the flanges didn’t line up right, and they had got a chain hoist and pulled a piece of pipe back into alignment to bolt up a Coriolis meter. They had done that when they replaced the Coriolis meter twice, so I digress.

Stan:

You’re right. It’s not just like a piece of pipe, and if you have a skill set that’s…Their tools of their trade are 48-inch aluminum pipe wrenches, and they just put a cheater pipe on it and let’s stretch that pipe and line it up.

With a piece of pipe, that’s fine, but when you have a device that’s designed to resonate at a certain frequency, if you put external torque on it, yeah, you’re going to impact that. The meters today are highly immune to that, but yeah, it’s possible. I see it. You’re right.

Weldon: You got to get it installed correctly, right?

Stan: Oh, absolutely.

Weldon:

Then the techs operating out in the field, they’ve got to be properly trained. They’ve got to be properly trained, and they have to have the time to maintain, calibrate it, operate it properly.

Then we need the right data processing. In today’s age, that likely includes a lot of pre-processing right there on the well pad. Then we need that quality control afterwards. Someone needs to be looking at the data that comes out of it. You’ve got to get that entire village together before you get good information at the end, don’t we?

Stan:

I spend probably 50 percent of my time in a training explaining mode where I’m explaining to people what they’re looking at. Because you have more data, that comes with some responsibility. You have to be able to analyze the data, or at least have something like maybe an edge computer that can boil that down and at least start to contextualize that for you, what is that telling me?

Maybe you need to be looking at two or three different pieces of data. I think where we’re going with AI is we’re still going to need humans to connect those dots together.

I think we’re getting closer to be able to look for correlations in data, and I don’t know if it’s machine learning or if it truly is artificial intelligence or not, but it’s automating that workflow that’s going on in your cranium and automating that on a device. Ultimately, a human’s still got to look at it and say, “Is that the right data or not?”

Regardless of where we go with AI, there’s always going to be that requirement, but it’s going to take training. The more intelligence we get, whether it’s artificial or not, the more we’re going to need to train people on how to judge whether or not that data’s good or not.

Weldon:

AI is already getting its fingers into our business. Overall, I think that’s a good thing. I think we’ve got to have it, with the direction our industry’s going. I just think we’ve got to apply the intelligence to do it.

When I got into this industry, “Oh, computers are cool. How do we know the darn thing is right?” That’s how I got into this industry. I started in this industry programming flow computers, so a whole lot of that, “OK, we made a change. We have to validate that change. We have to understand that change.”

I was working for a manufacturer, Bristol Babcock. Long gone, but any rate, every one of our customers, when we had a firmware update or a new model of flow computer, they wanted to validate that in their own lab.

Who has their own flow lab anymore? First of all, we’ve never had those on the upstream side. On the pipeline side, very few companies are operating their own flow labs and doing the kind of testing they were doing before, because it’s a generational thing.

My generation is computers were new. They were there as I was getting into the industry, but they weren’t standalone. PCs were there. I went to school. Originally, I was going to program business software. Computers in the oil and gas world, “Prove it’s right. How do I know it’s right? I want to see from my own two eyes that we get the right answers out.”

Today’s day and age — this is a generational thing. I don’t mean to be derogatory in any way — computers are part of life, “It’s a computer, it must be right.” That’s the change we made. That’s the thing that scares me as we move into AI, this assumption that it’s a computer, it must be right.

We’ve gone from a computer that was carefully programmed by, hopefully, someone that understood not only the programming language, but the task he was programming for. We’ve gone from that, teaching a computer what to do, to giving the computer access to a whole bunch of data and hoping it figures out the right process.

Stan:

I still find myself going back and and doing manual calculations on something. That’s all I can prove. If I see something that’s odd, then I’ll have to physically make a calculation just to prove to myself that I can trust that number. I think there’s always going to be an element of that.

How we maintain people that know how to do that, that’s probably the challenge here, is how do we maintain some level of discipline to check that automation, I guess.

Weldon:

That may be a good thought to land this on. How do we maintain the right people? How do we get those sanity checks as we move forward to this next phase of measurement in the oil and gas industry?

Stan, I appreciate you being on the podcast with me today. Any other last thoughts or anything you have before we wrap it up?

Stan:

Just along that thing, if there’s any folks that are contemplating entering the industry, I’ve seen some really sharp folks hired in our industry to start analyzing this data.

There’s a future in energy, I would say. It can get political, but I think hydrocarbon is a friend of humanity. I would encourage you to help us produce and generate as much as we can because I think it’s key to human flourishing. It’s rewarding as well. I guess I’ll end it there. “Gig ’em, Aggies” is the only thing else I can say.

[laughter]

Weldon: Gig ’em, Aggies. I’m with you. Thanks a lot, Stan. You have a great day.

Stan:

You bet. My pleasure.

[background music]

Weldon:

Thanks for listening, folks. I hope you’ve enjoyed the podcast and that we’ve given you some things to think about. Please spread the word about our podcast by telling your boss, your coworkers, industry peers, and anyone else you think that might benefit from it.

We will have Stan’s contact information, along with a full transcript of this episode, in the show notes on our website, pipelinepodcastnetwork.com.

If you have comments or questions about this episode, suggestions for future topics, or if you would like to offer yourself up to the podcast microphone as a guest, send me a message on LinkedIn, or shoot me an email at weldon.wright@fpvprime.com.