Interview with Geologist Eli DenBesten
Season 3
- Back to all episodes
- Listen to all episodes on
Episode Transcription
Welcome to the official Erdos Miller podcast, where we spend our non-production time talking about drilling tech and get the latest insights from the industry leaders on our show. I'm Ken Miller.
And I'm David Erdos.
And today our guest is Eli DenBesten. Eli is the operation geology manager at Tap Rock Resources.
That is correct, yeah.
Perfect. All right, so David, I'm podcasting from my new messy home office, you're podcasting from your closet. That's why we can't see you today, right?
That's right. Yes, and it wouldn't be a very good sight.
Do what now?
It would be very dark, not a very good video shot.
My question though is, is it tidy? Because I've been to your house before, and your house is exceptionally and impressively tidy. So I have to know is the closet tidy too, or do you just hide all the mess in the closet?
Some of the closets are tidy, some aren't.
Okay. So there are some parts of your home that are not tidy?
There are, unfortunately.
Okay, well I'm still impressed. Most people are like me and my own office isn't even tidy right now. So another COVID special podcast, broadcasting from our homes, so I was hoping to say we could get out of this mode, but it looks like cases are going up, and that's interesting. So anyways, Eli, tell us about your journey in oil and gas so far.
It's been kind of, I would say, from the geology side, pretty unorthodox. The path for a lot of geos, it seems to be, especially before the 2014 crash, was go to undergrad, go to grad school, work in the office directly. And I got out of undergrad and really was pretty sick of not making any money and being really poor, and moved out west and just had the four year degree and kind of worked from the ground up to make my way through the field and then eventually in the office. I think starting off in the field, I started off in the DJ Basin geosteering and mud logging there right off the bat.
I think getting that opportunity to start geosteering in the field initially was pretty huge and a pretty big jump for me to be able to go from that point, have the rig experience, talk with DDs, talk with MWDs, and work on the rig for a while, to have that hands on experience to then transfer into an office setting and bring a lot of value there, even without having a master's degree. Now some companies really want you to have a master's degree as a geologist. It's pretty stiff competition generally to get jobs in the oil industry in geology. But more and more, I've been working out of the office now for about four years, and it seems like the guys that we hire on and the people that seem to bring value, at least to the operations side, and this seems obvious to probably everybody else, but are the ones that have spent time in the field, a significant portion of time in the field really working hands on.
Especially in fast basins, where they're making target changes on the fly, they're talking with DDs, you have to watch real time constantly as you're steering. Those are the guys that I see bring a lot more value to an operations program, which that may be just me tooting my own horn. That was my path, obviously, but putting in those years in the field and really getting to know how the rig works and just how to drill a well I think has been really valuable. It's, I guess like I said, not common, but the operations geology, specifically the role, it's really changed and it continues to change as we're really just doing this underground mining and really trying to drill faster wells. It's really starting to really become heavily data focused. So that's an aspect that software has really enabled over the last few years. I guess that's kind of a longer answer to your question, but it's my path at least.
No, that's fantastic. It's funny, talking to David, David I would say your approach was very traditional with schooling, right? Mine was very non-traditional, so I think my path was much-
That's right.
... closer to Eli's. I think that both can be valuable. I think, personally, I think that especially with the tuition rates, college has become way overvalued. It's just getting out of control. I don't even know what to save up for my children to go to college, who are one and four respectively right now, because I don't know what that cost is going to look like when he's 18 in 14 years. Linearly extrapolate that and we're in trouble. But, no, I mean, there's just so much to be said for practical experience, and one of the things that we try to do at Erdos Miller is even software developers, anybody in the office who's building anything for directional, we try and get them to the rig at least once, so they can see what it's like and how tough it is and how much is going on, and just the weight of these things, like what's a million pounds? A million pounds is a lot. So, a couple things really stood out to me there I really want to ask you about, so how do you just start geosteering, because you said you started off as a mud logger, right?
Yep.
And then you got into the ... Can I just say that you don't fit the stereotype of a mud logger?
Why's that?
Oh, I don't know, you seem very normal.
Yeah, I think the fact that ... I guess I don't know how to answer that necessarily, but-
Well, forgive me, but my experience in the industry was the first rig I went on, the rig breaks, the mud logger's like, "Hey, you guys want to go do some LSD and jump in the river?"
Yeah.
"No, we don't. No, we're okay."
I think I have a good answer for that, I mean, I'm still a rock nerd. I guess, maybe kind of a little bit of a hippie as far as I just love to be outdoors. I live in Colorado, I love being in the mountains. That aspect I really love being on the rig for that, but I think really it's kind of the cream has risen to the top. When I started in 2014, it was kind of similar to that, with the LSD and the drugs you're talking about.
It was just, it was mostly a joke.
I know, but drug testing, even that, operators got more serious about it, because it's like, "Okay, we have to drop rigs." So I think it's really allowed the more professional guys and the more professional mug logging companies in general to rise to the top. So since I've been in it, I guess it's not that traditional a view, but, I mean, the guys that I have working on my rigs now, they're PHDs some of them, working in the field, which is you're like, "I think you should have my job." But it's really changed. I think that that job in general has changed as we've dropped down to just under 300 rigs now in the U.S.
Actually we're just over 300.
Did we just hit over 300? Okay, because we were down-
Yes, so it's-
... like the 260s. But yeah.
It was a little bit of a depressing thought to realize that the industry is about 10% of what it was when I stared.
Yeah, right. And since I started, it's just continually gone downhill. Everything's gotten more efficient, costs have gone down and that's really allowed ... I think that's really made the good true geologists to be on those rigs. Now to be a mud logger, you have to have a geology degree, pretty much. You don't get these guys that are just dead beats trying to just make a couple of hundred bucks a day working on the rig and just calling shale versus limestone. It's almost always college educated guys, and if not, really highly experienced that have been in the industry for a long time if they're still out there.
But yeah, starting out in the DJ, really, I think getting the opportunity to geosteer and mud log right away, just like you asked just starting, how did I get started? I got hired on by a small company, basically as a contractor and they plopped me in front of the computer with SES and they went over it with me and for a week kind of trained, worked side-by-side with a guy and then they just let me go and figure it out. It was a huge learning curve and I think that was a huge opportunity for me, because a lot of times you have to start out mud logging and then you have to go to a geosteering position and then you eventually maybe get to work in the office. Working for a smaller operator or a smaller company for an operator was how I was able to do that. But yeah, definitely they put a lot of weight on my without really much experience, which is, they probably shouldn't have done in hindsight.
Well, I had a similar experience. The best thing about small companies is, this can be a blessing or a curse, but they'll almost always give you enough rope to hang yourself. So I have to imagine it's very difficult and just takes a lot of learning and thinking and trial and error to build up a mental model that you really trust of what that underground geology looks like, because that's what you have to do. These holes are not big enough for anybody to go down there and just take a look or whatever else. Imagine that, that would be really cool if you could. But the GEOVATOR thing at the Houston Museum of Natural Sciences is actually really. I definitely suggest everybody check that out if you haven't seen the Wiess Energy Hall there. They may not during COVID, I'm not sure. I think they're open. I think my wife just took her kids there a few weeks ago.
But anyways, so talk to me about that process of becoming confident enough, of understanding the underground geology enough as to advise the rig where to go. Because that's where we get from ... I mean, mud logging in my mind, and correctly me if I'm wrong, is more about categorization and observing and just classifying what we're seeing and then when you jump in the geo steering, it's when you're now comfortable with that middle model to now advise the rig as to where to go, or is that too primitive of understanding?
No. Well, I think they go hand in hand. It really depends on the base and so, a place like the DJ Basin and everything is pretty, in geology, we'll call it layer cake. But it's really pretty consistent beds of deposition. You can almost steer a well using samples in a sense where you start drilling out of the [inaudible 00:11:16] into maybe more of a [inaudible 00:11:18]. You'll see that in the samples. If you've come down you can tell them to come back up. But DJ has a lot of faults which is also something that that's where bringing in the mud logging data really helps you. And they really go hand in hand, that's how I steer in general. If you hit a fault and you drill into maybe a lower section, that section of rock probably has some sort of characteristics you can see in the mud log. Even if it's the same lithology, it maybe has certain fossils that weren't maybe up in this higher layer, that can tell you, "Oh, we actually faulted into this lower block. We need to go down to get back into our pay zone."
It's really not one or the other. I think looking at geosteering from only loading gamma ray into a software like SES or StarSteer, is really just a very short sighted way to look at the operation side in general, especially when you start looking at basins like I've experience in the Anadarko Basin, which is we would say is heterolithic, not consistent. Then when you start going over to the Delaware Basin as well, really inconsistent rock. It's really taking all that data, looking at all your offset logs and really coming up with an idea of what the depositional environment originally was. What is feasible geologically? What feasible from a depositional standpoint?
And then, what's feasible structurally in making that mental model in your mind using all of the data that you have from offset verticals and from the current while you're drilling offset laterals in order to steer those wells. And constantly being in communication with your mud loggers on site to be able to identify those zones. Okay, we came up into a limestone. Does that limestone have the same characteristics as the line that we saw as we were landing the curve, or does it look different to you? Does it have specific attributes that you would say, "It is not the same limestone?" Those are the kind of things that I think these geosteering centers that a lot of mud logging companies have. I'm not saying they're bad, but they really don't have the full picture of what's happening in order to make the decisions, at least in really technical basins that they really need to, in my opinion.
What's missing?
Really what's missing is all the geologic modeling software and all the data that we have working for an operator. They are given, for the most part, they get the MWD data, the surveys from the rig and-
Oh, so you're talking about like a geosteering ... sorry, office that ... a remote geosteering office that maybe a directional driller has set up?
Yeah.
Okay, so they might not have access to the original, what do you call it? Geofoam data and all this kind of stuff, right?
Yeah, if there's seismic, if everything's planned off of older vertical offsets, usually, which we have logs run through, and that's where we can design out targets and look at the depositional environment. I think in a really continuous basin, I think those service companies do a good job, where we have that heterogenous layer cake model like I was talking. The DJ, I think is a good one, that you can do that in. Probably the Bakken. But the more complex basins, you have to bring in, I think, more of a wider perspective on it and really understand the geology of what's going on.
It sounds like they're-
That's really interesting.
Oh, go ahead David, sorry.
I was going to say, I've never heard of geosteering and mud logging working so closely together. Obviously I'm on the drilling technology side, so I'm pretty far removed from the geosteering and actual rig operations, but how similar are these different offset wells. If you're in a given basin, how much can trust other wells that were maybe a couple of miles away to reference and base your geologic model off of?
It's always an analysis of the data quality first of all, checking KBs, looking over the directional surveys and then also sometimes looking at the mud logging company, if we do have mud log data for that offset to see, "Okay, who logged this well? Do we trust what's going on here?" Because there is some human error factor there. But you also have things like XRD, XRF, which they don't have ... It's a machine that tells you ... So XRF tells you elemental composition of the rock and the XRD tells you what the actual lithology is. If you have that data, those are good hard points to be able to tell you, this si what the rocks are looking like over here. A lot of times with offset wells, the verticals, those are good, those are our best data points. We use those gamma ray logs and any other logs if we have triple combo to correlate across the basin or even just across in our local area in the sections. And we use those to see what's continuous or maybe what's pinching out. And that's what we build our structure grids off of.
But using offset laterals really give us an idea. We can re-steer those, look at where they landed them. Look at the gamma that they say, see if it changed and put that into our model before we drill our wells to see, does our model hold up or does it not? Really that's where I'm saying, looking at all that offset data. Looking at any piece of data that you have is huge, even down to how did they steer this well? They went down and they went back up. Okay, our structure grids aren't showing that. Is that feasible that structure's doing that, or were they lost? Those can be really good data points to tell you if they were cutting down section, if they were cutting up section. It's a lot of guessing. I would say geosteering is kind of being ... To me, these decisions are always a coin flip a lot of times. It's kind of like a 51% this way and a 49% the other way. Usually you want it to be higher than that. But sometimes, it's really, sometimes you can get really lost in these basins and you just have to really be able to pull in all that data and say, "The most feasible option is that we are high and we need to go down." Or low, meaning we need to go up. Sometimes that's all that you do.
But relying on the mud log, is just it's one piece of that, but it's especially in heterolithic basins, it's really huge to be able to know what the rocks actually look like down there and be able to cross correlate that to what we're seeing in gamma, because they're related. I mean, shales you're going to have hot gamma. Limestone the same, you're going to have cold gamma. If those aren't matching up, then you have something very wrong. A 20 API difference you can see that sometimes in a shale, the difference in that shale, the mud logger can be able to tell, or you can't. That's where the mud logging comes in and having those gamma ray re-logs, making sure that your tool is calibrated exactly the same is huge. I know nobody likes to re-log, but that's why geologists always want to re-log. It's like if we drill back through the shale and it has an API that's 30 or 40 API hotter, than we really don't have a way to correlated that. So then, we're guessing the interpretation for the rest of that run.
Hey, Dave, do you want to tell him or should I tell him, that the gamma instruments are all really poorly calibrated?
I think you should tell him.
Yeah, I'm sure there's not a ton of faith in it, but-
Well, I mean, no, I just, I don't want to betray my little section of the industry here, but I mean, directional instruments, magnetometers and accelerometers are extremely well calibrated and checked all the time. [inaudible 00:19:33] and all this kind of stuff. Gamma ray just kind of is what it is. I mean, calibration is ... I mean, people certainly do it, but it could certainly be done better. It's not done over temperature. It's one co-efficient for the entire well.
Directional, we actually apply a curve for the accelerometers and magnetometers, we actually look at how the accel mag changes over its temperature and come up with a polynomial fit for that calibration, versus gamma ray, it's one scaler, just apply an API scaler to it and then we just assume that that's the same thing the whole time, and obviously, it's not when your instruments perform differently over temperature. Now, is that actually going to make a huge amount of difference to go in and calibrate them? Probably not, but it's always been like a head scratcher for me as to why we calibrate the directional instruments so well and gamma instruments done so haphazardly, right?
Yeah, it's a problem with our data sets for sure. I mean, per well, you want it to be close. If you have that scale factor the same for you well. For one well, that's really what matters. For our basin wide data sets, yeah, the calibrations are all over the place. And that's where you have to bring in a petrophysicist to run normalizations on all those logs and really try to get all of the logs into one database and then be able to get them to where they're reading same API values. But really, when we're correlating a lot of this stuff, it's looking at the trends in the gamma ray, versus the actual API values. Sometimes it can be actual API values if you look at all your logs and you say, "Okay, these all look pretty similar, the normalizations all look pretty good." But, a lot of times you're looking at a finding upward sequence or coursing an upward sequence to tell you where those depositional, those pro-grading and retrograding sequences are happening to be able to tie those across multiple sections to figure out what your structure is doing.
That's fantastic and I mean, it's really cool to hear that you guys normalize all that gamma data. And with gamma, that's what it really comes down to is the trend, right?
Yeah.
I mean, you scale the trend up, scale it down, all this kind of stuff, but I mean, the signature is the signature. That's probably why the absolute calibration of the instrument isn't that big of a deal, is because you're primarily looking for the trend. Then it sounds like you guys have developed techniques to normalize the data anyway, right?
Yeah, we have. It's not the easiest thing in the world. I think being able to have all these larger data sets live on ... Just the computing power now, makes us able to be able to really normalize thousands and thousands of wells. That's not my specialty per se, but a good petrophysicist really is able to run those algorithms and write those algorithms to be able to do that.
We have a similar problem with shock and vibe measurements, because nobody's tool is going to give you the exact same number of Gs as anybody else's tool. Because the shock and vibe data from the bit, if you think about it goes to the bit, it goes to the motor, it goes to the MWD, it goes to accelerometer at the MWD and it's actually observed at the accelerometer on the MWD. And everybody mounts those accelerometers differently in the board, the boards are mounted chassis differently, the chassis is mounted to the MWD differently, the MWD is mounted to the collar differently. There's shock absorb designs, there's lock down designs. I mean, trying to get everybody to register a 50 G shock in the same way is a family mission. But, I've never seen two shock sensors from the same tool that didn't match the trends identically. So we rely heavily on the trends there. One of the things that I picked up on is that you said that there's definitely basins that are ... What did you say heteral-normal?
So heterolithic or-
Heterolithic, right?
... homogeneous. Yeah.
Kind of how you describe it.
Heterolithic just means just the deposition is not consistent. It's that you maybe have a debris flow coming in and a shale deposition and then sands are coming in this way and then another debris flow. You get this weird not perfectly ... I think a lot of people think of rocks, maybe not this way, but just perfect like you're stacking books on top of on another. Like, "Okay, here's the layer, we're trying to drill right in that one."
Well, that's how it gets explained every time. Nobody ever [crosstalk 00:23:55], dips and the faults and whatever else-
That's how I understand it.
... that's always after thought.
I think the first looking at if from a book standpoint or a layer cake is what they call it, and then you start thinking, "Okay, what's happening structurally? Which way is it dipping." Then, "Okay, do we have faults running through that?" But then, you start looking at what I like to call is depositional structure. So, you would have maybe multiple sand lobes that are coming off into a basin that are covered in these shales, so it's hard to not have a diagram to show it, but basically your gamma signature is not going to be consistent, because you're going to have lobes just depositing on top of one another. Then another shale, maybe a flood service will come on top of that.
If you're trying to read the actual dips of those sand beds, you're going to get fooled every time because they're not actually deposited in this perfect continuous package. In the Delaware Basin, if you're close to the central basin platform, you have a lot of shale in the middle of the basin, and then those layers, as you get closer to the edge, you start having limestones come in. So you have shale and sand and then your sand actually starts turning in, these debris flows start replacing your sands and you could be drilling a while over there and your good sand or your good shale is all of a sudden replaced by a large giant blocks of limestone that have caved off the central basin platform and been deposited there.
That sounds extremely fun.
It's extremely fun. It's extremely difficult and that's what makes the heterolithic basins just way more difficult to geosteer in. And being able to like, before I was talking about, looking at those offsets, looking at stuff if we're on the east side of the basin looking at offsets to the east of our well is going to tell us, "Okay, was there a debris flow here?" You can't always pick it up on the logs but you can see increase in limestone in specific logs and then you can see it decrease as you go to the west. Seismic is huge for that. If you have seismic, you can not necessarily pin point exactly where those debris flows are, but you can start to see where those trends are. So that's another big aspect that gets more into the exploration side. But yeah, like I said, it's bringing all the data possible that we have to try to figure out the puzzle and figure out if we drilled into a giant block of limestone that was deposited there or if we faulted into it.
Well, that kind of reminds me a bit of failure analysis with MWD. So, I mean, every extra bit of data you get when analyzing a failure, just makes the possible range of conclusions and the likelihood that you're going to come to the correct conclusion, so much better. Because if you just give me the memory, well, I mean, I can tell so much from the memory. It's really good. I'll probably come to the correct conclusion. But, I mean, you can have just a summary, you can have memory, you can have pictures of the tool, you can have access to the tool, all these kind of things, but if you have just a little bit of the data, you can come to a totally wrong conclusion that looks right based upon just having a little bit of data.
For sure, yeah.
So I mean, what you're telling me is that there are basins that are super easy, like the DJ Basin, right? Some not so crazy, but relatively straight forward to geosteering, right?
The deposition, yes. Yeah, yeah, yeah.
So do you have basins ... You said which basin was really tough?
Anadarko Basin generally.
Anadarko.
Drilling some of the sands there is pretty difficult. Then the Delaware Basin I would say. That's most of my experience. But yeah, yeah, yeah.
I know a group of guys that are really trying to automate this entire process and they're trying to throw AI and neural networks at geosteering and have those things make the decisions, and I can see the look on your face right now.
Yeah. Yeah.
But it sounds like if they were going to be successful, they'd probably want to start off just focusing on the DJ Basin, and focusing on trying to automate the more straightforward basins to geosteering and the stay the hell away from the Delaware and Anadarko, right?
Yeah, I think so. For sure, you want to go somewhere, where you know the gamma ray signature is going to be consistent throughout your entire section. And the DJ has a lot of structural complexity, but if you're riding up and down section, you can find a tie in. So you might hit a fault, but that's still, you're going to be able to know essentially once you get a little bit more data, exactly where you are. There are assisted automated geosteering. StarSteer has a version of there, I think. There's a couple other companies like you said, trying to do it. And they do a good job in those basins that have consistent gamma ray packages, where your deposition is extremely consistent, you know what you're going to look like. But yeah, when you start getting into more complex basins, how does that algorithm deal with, "Well, now our gamma ray log looks completely different, because we drilled into a debris flow or the sand that we were drilling in originally is actually not even here anymore," that can happen. So yeah, that's where you still need to have a ... There's a reason that we spend all our time mapping and trying to correlate these things, is you still need to have a geologist to say what's feasible and understand the depositional environment in those cases.
Perfect.
Well, I was going to say, I think we're running a little short on time, but I do want to squeeze in one more question. We talked a little bit about the importance of all this data, what additional measurements or more dense measurements would make your life as a geosteer easier?
Hmm, that is a good question. I think azimuthal gamma ray is a good tool. I haven't really used it very much. On the directional side, really, I think one of the biggest tools honestly is being able to use rotary steerable assemblies to a certain extent, and be able to point that BHA where you want it to go. A lot of times, you spend a lot of time undulating between survey points with the conventional assembly and you really, I mean, you guys know, this is the farther you get out into your lateral, the more difficult it is to really get a true CVD reading on what that CVD actually is. That is a huge error bar in our interpretations. And then, if we're using those interpretations to make maps for future wells, that makes things much more difficult. Survey accuracy, I would honestly say, just in general, I think with the rotary steerable, you get a better idea of where you are the entire time. And I think you probably have a better ... That error gets smaller and smaller.
But that's for me, just looking at it from a mapping perspective and trying to use the data to continue to model the basin. We have so much horizontal data than we did before and that can all be used, but we can't run a gyro down each one of these laterals, so being more accurate with that is, I think, a huge part of it. Right now, I have an algorithm that I run that basically uses continuous inclination, which I think is massive. It wasn't when I first started working out in the DJ, they didn't have it. And then they started implementing it a little later, but running a continuous end calculation between survey points to tell you where you are and see those deviations and see how far away you've actually trended from where you think the [inaudible 00:31:48] is, is really huge. I think that's a part that a lot of geologists and geosteerers overlook and they don't really think about the actual accuracy of that data.
That's one recent well, actually, we had continuous inclination included with the survey corrections, versus without, there was a 28' TVD difference.
At TD?
Yeah.
Yeah, yeah. So I mean, that's a huge difference. I mean, if you're trying to model a basin, and it's pretty flat and you're actually 28' off, I mean you go drill a well next to it and your zone's only 28' thick, you might be out of zone with that model.
I can see in my head a specification for a cost effective simplistic rugged rotary steerable system, that has continuous inclination and azi gamma sensors built right into. I think that's the tool that just takes the market, honestly.
That would, yeah.
I mean, and it's all very possible, like the azi gamma packages are small enough and rugged enough, you can put them in there. They can be short hopped back to the MWD for transmission or you can build the MWD directly in. But I mean, I feel like of all the possible measurements, all the possible things that you could ask for, that's like the AK47 drilling tool, right?
For sure.
Your getter done basic, this is just when I need to know more kind of specification, right?
Yeah, I would 100% agree.
So David, I've got a couple of more questions, I've just got to ask.
Yeah, go for it.
I've always been told, we operate in the U.S. a lot and then a little bit internationally and internationally, these guys can't live without at least [inaudible 00:33:35], if not triple combo. So I've heard numerous explanations over the years as to why that is and why the U.S. just mostly predominantly relies on gamma ray and all this kind of stuff. But I'd like to hear your perspective on that.
I think offshore stuff, they're probably dealing a lot more with conventional wells, if I'm thinking about this right. Or not offshore, I should say, but oversees. They're drilling conventional reservoirs where they're poking holes and anticlinal features and trying to look at the quality of the reservoir, and most of that is based off of seismic. So you at least want to have sonic logs to be able to tie your seismic too, to be able to tell what those TBDs are going to be. In the U.S., I think really it comes from being cost effective a lot of times and not really ... I mean, when we started geosteering, we were doing it with Excel. I mean, there wasn't these fancy programs like StarSteer or even SCS was a huge jump in technology to be able to really model that and look at it. I think it's really a cost thing.
I think that we can get by without needing it and I think that's ultimately why, especially just drilling these horizontal wells, your gamma ray can still tell you a lot. And really, if you have a triple combo ... We still do triple combos, we still cut core. That's still in our science, budgets to those things. It tells us about the quality of the reservoir and we can tie those high tier quality wells across areas if we have them. Then we can correlate with our gamma ray between those. Having resistivity is huge for some of those correlations depending on the basin, but if it's a simplistic basin, you don't necessarily need the resistivity even. So, I think it really comes out of cutting cost and really it's not always necessary if you are going to be drawing that horizontal, either way, you can pick you pay zones off of your good triple combo logs and then extrapolate from there, so yeah, that's kind of yeah.
Talk to me about big data with geosteering right now. You alluded to that earlier. What's geosteering like today with big data?
Right now, there's a big push to be able to connect really programs like Petra is traditionally what's been used to map and model. These are projects that have an entire basin built on them with all the wells, all the logs, all the casing points, everything. And that's where all the geologic mapping is done and where all the prospects are looked at. The steering software never really talked with that data, so you pull the data out of that large data set, put some of it in StarSteer, SCS, steer your wells and then you'd have a screen shot of that. Then you go, "Okay, we kind of know what we're looking at. That's okay." Then you were able to start exporting control points in the backend of Petra to be able to do that mapping.
It's gotten better and better now to where programs like StarSteer, you can pull all of your data from that large database into StarSteer and actually map in that program. So you can use all those control points along a well bore. When I say control points, it would be the formation tops that we saw when we were drilling that horizontal and then looking at the average dip and using that dip to create a new map. So that's kind of the first aspect of it. Then, now we're able to actually export all the data from our steering projects for an entire basin, put that into a server where we can look at all of that data from the gamma ray and all the geologic stuff. All the mud logging data, the XRD, the XRF, like I mentioned, maybe mass spec data, gas data. But also, bring in all of the drilling data that we get.
So I'm constantly running MSC in my projects. I calculate my own MSC with a Python script, looking at collar RPM, looking at weight on bit, comparing all that to our ROP, exporting all that data from StarSteer, put it into Spotfire and looking at all that data and really analyzing how are these rocks drilling and where can we optimize our pay zone or in our verticals where are we burning up bits. Look, we slowed down maybe our gallons per minute, lowered our RPMs through this specific rock and we actually saved a little bit of bit life there, we made it a little bit farther. Maybe we need to look into that more. That's one aspect of it.
Then, you can also bring in your completions data. And you can link your production data to that. So now, it's really becoming an operations geologist has really, you're not just geosteering anymore really. You're really running a lot of this analytic. And going back to where we started at the beginning of the conversation, having experience in the field and understanding what's happening and really understanding what drilling is doing and what these parameters mean and being able to look at them, is extremely helpful to be able to bring a drilling engineer in and show them a plot and say, "This is what we're seeing on the geologic side." And show them maybe ROP heat maps or something that we can optimize our drilling. We can drill these faster. So much of what I'm doing is trying to figure out where can we drill faster, but still close to our pay zone, or within our pay zone. And rotary steerable tool-
[crosstalk 00:39:32].
Yeah.
Awesome. That is all so awesome.
Like a rotary steerable tool like you said, is the mecca of that, just due to the ability to make those quick corrections and maybe keep it within a smaller pay zone. Even if our pay zone is 20 or 30' thick, if you find a good shale in the middle of it that drills an extra 50' per hour, and you have azimuthal gamma ray with that rotary steerable, you can see where you're getting close to the bottom and just continually just work it up or work it down, depending on where you're at in that zone, so making those tiny corrections.
You don't have these over-shoots anymore with like a conventional assembly, right?
Exactly, yeah.
Which if you're a control systems, looks suspiciously like PID over-shoot, right?
Yeah.
But I really picked up on the fact that you do a lot of Python, so do you actually write the Python code yourself for all the analysis and stuff like that?
Some of them I do, some of them I don't. These programs have built out scripts already, a lot of them. But I do modifications and write other ones to calculate different log parameters that maybe I want to see. I can basically run calculations on any logs that I have, that I can put into StarSteer. Anything from Pason, any of the parameters from the rig and anything that mud loggers have given me I can put them in and basically run calculations right in the program in Python, so yeah, I think, it's a mix of both. I'm definitely not a Python expert, but-
Well, that's the wonderful thing about Python, is it is just such a fantastically easy to use programing language. You don't have to be an expert to actually achieve something in the real world of Python, right?
Exactly.
I really can't recommend learning Python enough to people. I mean, it is such a great way to get introduced into programming. And look, a lot of people are having a tough time and their career right now, between the oil price crash and COVID and everything. A lot of people looking for what they're going to do next, [inaudible 00:41:25]. You could add so much value to what you can do in your career if you learn a little bit of programing and Python is a fantastic place to start.
It really is.
I highly recommend it.
Yeah, and we just hired on a couple more operations geologists and the biggest thing for us really is, do you at least understand these language and how are you with basically being able to manipulate this data and do a little bit of programming, that's where this operations geologist role has gone from really nerding out on the geology to you still have that, but we have massive data sets, tons of data that we can look at, and in a lot of companies just sitting there. So finding a way to really incorporate all that data from the drilling site all the way through to production, it's becoming more and more the role of the operations geologist and that's why I think it's realLY changing. And hiring on new guys, it's like, what type of skills, especially in programming or maybe sequel server, or even being proficient in Spotfire, what sort of tools can you bring to the table that are going to help us optimize our ability to really digest all the data that we currently have and take it to the next step.
And the data is in new formats. It's not longer on paper printouts. It's in databases and Excel files and CSP files and SQLite files and whatever else, and there's 100,000 times more of it than there used to be. And just I mean, there's so many positions anymore. I mean, even as CEO of a tech company, I've been sitting here writing script in Python, to work with our receiver data systems. There's just so much of anyone's job anymore that's just pushing around data and analyzing it, that typically wouldn't, it's just a critical skill to have.
So I've got one last question, and then we'll wrap her up. MWD, friend or foe?
Am I a friend of MWDs?
Well, are MWDs a friend or foe to you? Help getting your job done.
I like to be friendly with MWDs personally. I start to get ... Yeah, when I worked in the field, I was friends with most of those guys that I worked with. It all depends on how much bad data you send me. If I tell you to correct something four times, and every time you send it back the same, that's when I start getting mad. But no, I mean, I like MWDs. Definitely not a foe.
I'm just teasing. I'm just curious as to whether or not that's typically an at distance relationship or a very close relationship. So that's kind of what I'm getting at.
Yeah, not as close, being in the office now. Definitely not as close. When I worked in the field, yeah, I was talking to the MWD a lot, because half of the time the DV's up sliding or it's in the doghouse, so yeah. I spent a lot of time, obviously, hanging out with MWDs and yeah. I'm not a fan.
There was a funny statement from the MWD perspective I heard one time, which is, "Why the hell should I care about where we are if the geologist can tell me where the hell we're supposed to be?"
That's fair. That's very fair.
So, very good. Well, Eli, this was fantastic. I really enjoyed our time here today.
Yeah, thanks so much for having me on, this was really enjoyable.
Thank you.
Yeah, thanks.
All right, well that's all the time we have for today. Thanks again, Eli for joining us and for sharing and giving us some insight into the geology in geosteering. Be sure to checkout our podcast on iTunes, Spotify and YouTube. Thanks for tuning into today's episode.