Skip to main content

Sponsored by: Sigma | Flogistix by Flowco, and the Role of Data in Responsible Energy Production

Summary

  • Vapor recovery optimization: Flogistix uses Sigma and Databricks to build a unified data architecture for real-time decisions in vapor recovery systems, preventing emissions while optimizing production.
  • Emissions reduction through data: Advanced analytics identify optimization opportunities in production and vapor recovery, helping organizations meet environmental regulations while improving efficiency.
  • Universal data access: The solution empowers everyone from executives to field mechanics to access operational insights and drive sustainable resource production using data-driven decision making.

Sponsored by: Sigma | Flogistix by Flowco, and the Role of Data in Responsible Energy Production

Watch: Sponsored by: Sigma | Flogistix by Flowco, and the Role of Data in Responsible Energy Production
Talk by Ali Sylvester, Director of Business Solutions, Flogistix (Sponsor Speaker); Danny Burrows, Vice President of Software Engineering, Flowco (Sponsor Speaker)
As global energy demands continue to rise, organizations must boost efficiency while staying environmentally responsible. Flogistix uses Sigma and Databricks to build a unified data architecture for real-time, data-driven decisions in vapor recovery systems. With Sigma on the Databricks Data and AI platform, Flogistix gains precise operational insights and identifies optimization opportunities that reduce emissions, streamline workflows, and meet industry regulations. This empowers everyone, from executives to field mechanics, to drive sustainable resource production. Discover how advanced analytics are transforming energy practices for a more responsible future.
Intelligent analytics for everyone: https://www.databricks.com/product/business-intelligence
BI meets AI: https://www.databricks.com/resources/ebook/business-intelligence-meets-ai
Explore the only unified platform for agent systems: https://www.databricks.com/product/artificial-intelligence
See all the product announcements from Data + AI Summit: https://www.databricks.com/events/dataaisummit-2025-announcements

FAQs

How does Flogistix use data to reduce emissions?

Flogistix uses Sigma analytics on the Databricks Data and AI platform to identify optimization opportunities in vapor recovery systems, capturing otherwise lost materials and preventing atmospheric emissions.

Who benefits from the real-time operational insights?

The unified data architecture empowers everyone from executives making strategic decisions to field mechanics managing equipment operations to drive sustainable and efficient resource production.

What is vapor recovery and why does it matter?

Vapor recovery systems capture vapors that would otherwise escape to the atmosphere during oil and gas production, both protecting the environment and recovering valuable resources for resale.

How does this support regulatory compliance?

By providing precise operational insights and identifying optimization opportunities, the solution helps organizations meet industry environmental regulations while reducing emissions and costs.

Full transcript

All right, everybody. Thanks for joining us today. Uh, we're going to go ahead and get right into it. So, jump into this. Okay, so and we work for a company called Flowistics. We are a brand of an organization called Flowco that focuses on oil and gas production optimization for uh oil and gas production. So we have a really special role in this in that and I'll go over the agenda in just a second, but we have a very special role in this in that we aren't the producers ourselves, but we're supporting the producers in making what they do possible.
So um the agenda, we go back to me real quick. We're going to talk about what we do, how we do it, what drives our business, and then how Sigma helps, which is one of the reasons why we're here today. Okay. So, who are we and what do we do? All right. So, our business um designs, builds, deploys, and manages what we call compressors. Really, the whole foundation of what we do is um optimizing production and vapor recovery.
So, think of it like intake and outtake. Our equipment is on a facility, right? An oil and gas producer digs a hole in the ground and then we're responsible for something in the middle, right? Production optimization is one big part of that. But today, we're going to focus specifically on vapor recovery, which is okay, bear with me for a second. Think about a bottle of soda, right? Like you twist off the top, there's pressure, and as you untwist the top of that cap, there is a phase change of the liquid, and it turns into a gas.
But if you wait long enough, it will turn back into a liquid. All right? So vapor recovery is what our our machinery does. It's sat on that facility and then as you open that soda bottle, we're preventing anything from getting away, right? We're preventing it from going into the atmosphere and we're also pushing it back into the sales line. So it's not just about profits, but it's also about the atmosphere, which we'll talk a little bit more about in just a minute.
We have about 8,000 units that we've deployed in total. Uh we have about 3,000 of those that are rental units and then several other that are customerowned. All right. All right. So I'm going to talk to you today a little bit about how we do what we do. Right. And we're going to talk about kind of three broadscale pieces. We're going to talk about our processes. We'll talk a little bit about platforms. And we'll talk about data because we are here at the data bricks conference and we have to talk to you about data, right?
So our processes are really designed around early detection. uh and detection from the office, right? Uh because we support our units out in the field, we send mechanics out to the field at some point, they actually have to work on our units. So, being able to understand what is going on in the field from the office before we ever roll a truck to the site or put a mechanic on site is really important. It helps us manage our operational efficiency.
So early issue detection, monitoring and understanding the operating conditions of our equipment and our units allows us to continuously give data to the field so they they can fine-tune our units. We're going to talk to you uh kind of in detail about uh our process to fine-tune our units. As far as platforms go, uh I'll talk to you a little bit about uh our telemetry platform and how we collect the data and push it up into the cloud and then how we leverage data bricks to store and manage that data and how we leverage Sigma computing to actually visualize and dive into the data itself.
But it does allow us to target the right units and make sure that we are arriving with real time up-to-date information. Oil and gas traditionally has been a little bit slow to respond to equipment. It's usually a um historical view of what's going on, but we get a real time up-to-date view of what's going on. For us, this helps us uh reduce repeat visits. It helps us be more intentional when we send somebody out to the field, and that saves money for uh us as an organization, but it also empowers our mechanics to be smarter about what they do.
And then lastly, we're going to talk a little bit just about data, right? Everything we do centers around data. It's, you know, compliance and reporting because we work in the oil and gas industry. We have EPA and compliance reporting. We have to worry about uh we want to track and understand the capacity and the utilization of our fleet. Uh this is very important for our salespeople uh to partner with our customers and that can upscale or downscale uh different units and then it allows us to understand how we can unlock value across the entire facility and really the entire oil and gas pipeline.
So I'm going to talk to you uh just real high level. Uh obviously this is a simplified view of what our architecture is. But in the top left we're going to start with our end user. These are our mechanics. It's our operations team. It's the people that are sitting in the office. And what they're doing is they're really working with the units to control and manage the units in the field. So the PLC is a programmable programmable logic computer on the edge.
Uh there's a whole bunch of equipment. We generate and build uh a lot of data points on our smart compressors. So we're very intelligent about how we operate. And because we generate so much data, we can actually use that data to make informed decisions. So we pull that data up into the cloud with our data collection process. We stream it over into data bricks via the streaming pipelines and data bricks into our Delta live tables.
And then we build a medallion architecture inside data bricks that will give us bronze, silver and gold level analytics that we can use and manage. Now collecting the data is one part of it. Visualizing it's another and operationalizing that data requires actionable insights on the visualizations and the data that we collect and this is where sigma comes in and really helps us right. So our mechanics, our operations team, our leads, our supervisors all leverage Sigma to look at the data as it sits up in the cloud and then that they push those decisions or make decisions about how our units are operating in the field in order to optimize and make sure things are running a little bit better.
So that's kind of how it works and it's a circle. As we get better and smarter, we show up to the site, we make a change on the PLC, we collect the data and we just start the circle over and over again. So it's a self-feeding process for us. So Ally, back to you. Yeah. So I can hear myself now. Okay. So I'm product and that's one of the big reasons why I'm here is because I'm one of the primary end users of this data.
As someone who doesn't come from the energy sector, what I have found is the way that I'm able to contribute is to spend a lot of time with our end users that are in the office, specifically the mechanics, and try to understand what can I service and provide to them that makes their lives better. And so we have a quote on here. We usually like to use the one by where it's like everyone has to play until they get punched in the face, but it was overused.
So we threw something else on the screen. Um, which is you can't manage what you don't measure. Um, but more importantly, you can measure or collect a lot of data, but if you don't utilize it, then it doesn't really matter. And so that's what Sigma helps us do. So speaking of how Sigma helps us, there are a couple key capabilities in the Sigma platform that we actually use uh that has made our life a lot easier and has allowed us to empower people like, people like our mechanics that are not technically inclined, that are not data people to really understand and make efficient use of our platform.
So one of the big ones uh that we really liked about Sigma, why we chose them as a partner was their intelligent query. This allows us to run queries and then cache those results for a short period of time. As we're historical in most of the data as as we run it and we're looking over a period of time, this allows us to kind of quickly store and recover information for the things that we have. Balances that performance and freshness, right?
Caching is a hard problem, right? It's one of the last hard problems in computer science that we know about. Um, but for us, this allows us to make sure that we can push insights into the field quickly and efficiently. And the data is cached at the warehouse level. So, we don't actually have to worry about pushing it back into a database. We don't have to do anything like that. Sigma handles that for us. One of the biggest and best features of the Sigma platform that we have found use is the drill anywhere.
And this is really, really important. And we'll have a short example of what this looks like when we talk about how we've used our data to optimize our units. This is drill down, right? data is important but being able to drill down and really understand what the data means to us and really get to the heart of the problems really helpful for uh not just the technical teams but the uh field service teams as well. Um and then kind of one of the big ones there's a direct integration with the unity catalog.
This means that we can control data access and permissions. Um segmentation is really important. We can implement our data governance policies. Um, and you know, Sigma just kind of helps manage all of that for us once we got it all set up inside data bricks. So, we're going to talk a little bit about a very specific example uh where we call we're finding unhappy units. Right now, mentioned that we have compressors. Our compressors are responsible for responding to what's happening in the field.
As pressure goes up and down, the units have to respond to this. Very important. And this kind of up and down process control is what it's called is really important. Now, a happy unit is a unit that's running and keeping everything nice and hot and things are just working the way we expect it to do. Well, what we'll see from time to time is when units are misbehaving, either, you know, there's something faulty with the unit, there's something faulty on the facility or the unit is just set up in a non ideal scenario, a unit will fault.
It will go down, which means it's not processing. And what we have here is a quick view of some of our actual data from uh from our field looking at units that are faulting high. And because of the capabilities here, it's real easy for any mechanic, anybody in our business, pull it up, sort it. It's a spreadsheet. You can copy it. You can play with it. You can really begin to just kind of quickly drill down and understand what's going on.
So, what we see here is at the top we have a unit that's falting a lot. It's going down on something called low suction pressure. it's uh bringing pressure in from gas and uh it's just going down pretty consistently. So we wanted to do some data analysis on this. So this is what we did. Now we pulled up uh our Sigma dashboard. We started to dig into it and we started looking. We noticed a couple of things uh just from the outset.
So this is our carbon emissions uh computation. And this is an internal computation that we run to show how how much carbon CO2 equivalents NGL's and a few other metrics that we're actually capturing and keep it from going into the atmosphere. Right? This is wildly important for a lot of reasons. will dig into that for you in a little bit. But it allows us a quick view of how our units are operating and functioning in the field.
And our major takeaway from this was we are operating at about a third about 30% as efficient as we normally could which means that our unit is not processing what we expect it to do. So we wanted to dig into this right that's kind of a quick high level. So we map our time series data telemetry data right pull this up from the PLC and we take a look at it and what we see are two things. So this top graph is called VFD.
This is our motor that's running uh our compressor, right? Our speed is up and down, up and down, up and down. And we see that we're going down to zero quite a bit. Now, this is an unhappy unit. Our units like to keep the gas hot. They like to keep processing. And we like to know that things are moving appropriately. Now, down here on the bottom graph, we have what's called a suction pressure. That's a target. We try to hold to a target and we control against the target.
There's a lot of math. There's a lot of science that goes into this, but essentially what we're doing is just pulling it up and reading this information. Now, our mechanics are going to look at this and they're going to learn two things very quickly from this data. So, two of the things that we see is that we have a problem with our sensor. Our suction pressure sensor on the way in seems to be miscalibrated, right? We're showing that, and I'll go back to the slide.
We're showing that we're pulling down into a pretty aggressive vacuum pretty consistently, and we know that's not feasible. If we're pulling into a vacuum that aggressively, there would be other problems on the facility. So, we're making decisions very, very quickly based off of the data that we have and we see that our VFD is thrashing. We're up or down. We're up or down. And it's likely that these two things are related, but we're unsure.
But what it does give us is before we ever decide to send a mechanic out to the field, we understand exactly what is happening on the unit, so we can either put the right parts on the vehicle, on the truck, and send them out the door and on their way. So, let's talk about the fix. Yeah. So, this is where I nerd out. Uh, someone who's not from the oil and gas industry originally looking at this data is like I think this is the coolest thing of all time.
Um, so the fix, we looked at that chart and I, like I said, I didn't come from this industry. I didn't know what I was looking at. Uh, but just through looking at these charts, I can make some pretty informed decision on what is going on in an oil and gas production facility with our machine from my desk. I don't even have to look at the unit. I just know some things about it, right? And so, a couple things happen. We equip the mechanics.
We have decided that there's certain metrics that we want to be right in front of their face. We represent that through these visuals that we're putting in front of them. We're sending them emails. We're sending them reports. So the first one that was saying we're going down on faults like 2,000 times in a matter of a very short period of time that gets surfaced up to them and they know, okay, I have a problem unit that I'm responsible for.
So we equip them with the right information based off those visuals and some of those reports and then they make the necessary modifications. The cool thing is is that because our equipment is smart, we actually have the ability to do some of this remotely. So for example, we can't calibrate the sensor remotely. we have to be there to physically calibrate that pressure transmitter. But we can make some changes around set points or targets remotely.
We can essentially connect to that unit from wherever we are. And that's one less deployment to that unit, right? Which is saving money and improving margins. But then we monitor and review those changes, which we're going to look at in just a sec second, which is where I nerd out. But the idea is that we've equipped them, we've encouraged them, we've told them what they need to do, and then they make those changes and either in real time, right, or historically we can evaluate what what's going on.
Were we right with the assumptions that we made and validation? Right. So I was lucky enough to be on this facility when we actually did this. The pressure transmitter did have issues. There was a lot of thrashing in the speed. We were going up. We were bottoming out at zero. So when you see these big blue chunks, that means it's happening so frequently that it's hitting zero that it's right making these blocks. And then you see when we made the change, this is a week before we were on the facility and a week after we were at this facility.
And you'll notice a couple of things is that the speed is stabilized. So, one of the things about our units that we designed them to do is to be responsive, right? They're smart machines, so they're taking in conditions all of the time, and they're responding. So, there are some conditions where we go up and where we go down, but we're not bottoming out, which means the compressor is staying hot, which means we're able to continue moving product the way that we would expect to.
And more exciting on this bottom, um, you can see that set point. We were all over it, right? And now we are controlling with much tighter precision around that set point, which is incredibly exciting. Um I like I just can never get over this. And so one of the things that we can do is remotely whenever we see these units that are oscillating all over a set point, we can actually usually remote into these units and make these modifications and never have to deploy anyone.
Eventually somebody's going to need to crank on the unit. But it's a really cool thing that we can do just from uh looking at these visuals and deciding we need to make a decision. So who cares? Why does it matter that I nerd out about how clean our graphs look? Well, a couple of things. So producers, if they're not moving product, are not making money. And the ability to produce money is by keeping your runtime up, which means you can't be going to zero.
But more importantly than that, we have a really awesome uh responsibility to be good stewards of our environment, right? Oil and gas gets a pretty bad name. Uh there's a lot around energy production that is not pretty, but one of the great things about vapor recovery is it's a win-win. Anything that is essentially getting captured by our vapor recovery unit could have otherwise been vented or flared to the atmosphere.
So, you've driven around probably somewhere, not probably in California, and seen flares, big fire stacks. Those get a pretty bad wrap because you can see them. But worse than that is venting to where you can't see the fire stack and it's just opening a relief hatch, right, for pressure and uh that's going straight to the atmosphere and has a CO2 equivalency, which is a warming effect. This matters because in this change, Is there one more graph,?
Yeah, in this change, what you're seeing is an increase. We're measuring MMBT use here, but we're going to talk about CO2 equivalency for a second. When we look at the metrics on the top, and we summed March, which was before we made the change, and then we summed April, which is after we made the change, we increased our capture of metric tons of CO2 equivalents from 210 in March, to 816 in April, which is like that's like 250% or something crazy.
That's so cool. Um, and the reason why this is really important is because that has direct value for the producer, right? That's more money. We're pushing that through the sales pipeline, which is what they want. Their profits are increasing. But it's also really important for all of us because that means um about 816 metric tons of CO2 equivalency is about the same as 177 cars removed from the road for a full calendar year.
It is the equivalent of about 64 homes powered for the entire calendar year. It is the equivalent of 99. 3 million smartphone charges and like 13,600 trees planted, right? That's incredible. And so this is one unit that was optimized for one month and we have over 8,000 units deployed out there. And if we can optimize all of them through using visuals, we're making a giant impact on energy production, which is absolutely critically necessary, but it's so much more responsible and so much more sustainable.
So, um, like I said, I'm super nerd about this. So is. Uh, and I'll be way less nervous when I'm not on this stage, but I'm super happy to talk to any of you after this. And I think that's what we've got for today. This is how we're using data brick, sigma, and our technology in order to make real business decisions because collecting the data is step one. So, any questions? Thanks, guys. Awesome. All right. Don't forget to uh do your surveys, right?
We're all good. Appreciate you guys. Thank you so much for being here, everyone.

Learn more about the Databricks Data and AI platform.

The information provided herein is for general informational purposes only and may not reflect the most current product capabilities or configurations.