Chuck Yates Needs A Job

0:21 This is embarrassing actually, because I'm on the advisory board of Montrose Lane, we're a shareholder and all. What the hell does Novi Labs do? Yeah, well, great question, great places to

0:34 share. It's like, this is going to be a call, the portfolio of companies that can figure this out. Yeah, so really we're a data analytics company focused on unconventional and lower 48 We have

0:46 two main business lines. One is that we have machine learning based forecasting tools for operators to design their developments. So, you know, how many wells should I drill in this section in

0:59 Martin County? Should I increase my completion design? Like, what's the optimal for rate of return or maximizing your MPV or whatever it happens to be that they're looking at? And then second. Is

1:09 that like Aries on steroids? I mean, dumbing down for the finance guy Yeah, for sure. Yeah, so, you know, Aries. you might drip your wells together, fit some curves to it. That's all done by

1:24 a human reservoir engineer, and they are subject to different biases. They can only hold a few things in their brain at once. They worked in private equity. Their deals were doing well. That's

1:38 how it always came in. A B-factor of three. I don't think that actually exists, but okay, sure. Whereas the machine learning model, it can look across the entire basin for analogs. It can be

1:49 very deliberate in analyzing all these different factors. And at the end of the day, what that means is you get a more accurate forecast and a better decision. So maybe your reservoir engineer

2:03 doing it by hand says, Hey, we need to pump 2, 900 pounds per foot. But in reality, you're over capitalizing your well and you're leaving a million dollars on the table just by putting all that

2:15 down whole. The model can help you identify places to get better economics out of your developments. Because it felt like towards the last, so I get booted, call it April 2020. So I want to say

2:35 maybe as early as 2016-2017, we were really doing a lot of that mechanically by hand of, Okay, let's drill a well, and let's pump two-thirds of the propin, or whatever. We were trying to be very

2:53 methodical in testing every variable individually, and what we always found would happen is, Okay, we all agree. We're just going to test the sand on this one. And inevitably, the ops engineer

3:06 to save some money would also do something else, and we'd be like, Now we don't know what it means Yeah. I mean, that's exactly right. Plus, you know, if you drill six wells on a pad, the four

3:18 interior wells, their production might go up or down 20, just for no reason, even if you do the exact same thing. So with all of that and all of the craziness on, hey, they decided to cut a

3:31 stage or whatever it is, by looking across many different examples, you can get a better answer for whatever you're trying to look at, rather than just putting all your faith on one science pad for

3:43 your cluster design or whatever it is And so, last thing while we're on Novilabs, if I'm a customer, is it a SaaS model? I'm buying software and data from you, how does that work? Yeah, exactly.

3:56 Yeah, it's a SaaS model, usually one or three year deals. We will, you know, depending on the size of the company and how many licenses you need and, you know, things like that will scale the

4:08 price up or down. Yeah. All the data is included along with that. Gotcha. I have one of the greatest data stories of all time. And the person who did this shall remain name was Ryan Rice.

4:24 Supposedly, supposedly, don't know this to be true. Toby marches down the hall at Rice Center G and goes, you know, Ryan, we've got four different data sets.

4:38 Pick one, I'm not spending a million bucks on data, just pick one and let's go with it The next board meeting, Ryan is passing out the board books and Toby notices that Ryan is passing out four

4:51 different books. They're looking at an acquisition or a development program or whatever, four different books.

4:60 Toby goes, hey, what's the deal with the four books? Well, I have four different data sets right now. We're gonna look at each one and I just want you to tell me which data set I need to get rid

5:11 of And by the way, the results vary by 50. across all four data sets. I believe it, yeah. Is that true about data? And how do y'all deal with that? Yes, it's definitely true about data,

5:23 although it does vary depending on which state you're in. Okay. So if you're in North Dakota and you look at a deal and you're using four different data sets, you're not gonna have a 50 discrepancy

5:37 in evaluation. But if you were in Texas or West Virginia or Pennsylvania, you might And the biggest driver of that, especially in the Texas Permian or the Eagleford, is the production data is that

5:53 for the state is reported at the least level rather than at the well level. So let's say you're evaluating a deal

6:02 in Howard County and the operators have been, you know, the operator in the area that you're looking at has been drilling 20 wells a section and you think, Hey, I think there's potential here in

6:14 the world. can't be, that's maybe not really tapped so far, and we can add in an extra four wells to their developments. You're using guesstimated production for what the Wolfcamp D does in that

6:29 area, because all you know is how much the whole lease produces. So companies like Novi will have to come up with an estimation for how much of the whole lease level production belongs to each of

6:40 the individual wells. So we spend a lot of time on that, and we've also, and this is something pretty unique, gone out to minerals companies and operators, and started licensing their well-level

6:51 data directly. So rather than having to deal with like these guestimations, you get, you know, the ground truth, like straight from the source, basically. So when I would sit there and either

7:02 my guys or a portfolio company with CEO would pick up the phone and call me and say we have to buy this one person interest in this area that weren't bullshitting me They were doing it because they

7:13 wanted actual data. Yeah, that's a great way to do it, 'cause if you're getting that data from your check stubs or working interest data share, you actually know how much the wells are producing,

7:24 which is extremely valuable to know. Yeah, we used to do that, and where it really mattered was when you

7:34 could also get the completion card and know exactly what completion recipe they used, and you could start differentiating by wells on that That actually really mattered. Yeah, exactly, and in

7:46 Texas, for the most part, you don't know how many stages

7:51 you don't have any idea about the clusters, and that can have a big impact on the performance as well. And that's another thing we've, and maybe you heard about this from the Montrose guys, but we

8:02 run this Novi Data Network that allows operators to share their stage data and cluster data for you're maybe, know you, So models training operating in Reeves County or.

8:15 Midland County or whatever, and you've only tried two different stage lengths, and you're like, well, what happens if we do shorter stages or longer stages? Like, who knows? But you pull all

8:25 your data with other operators who've tried other things, and now the model can tell you, oh yeah, you can increase production 5 if you shorten your stages by 75 feet or whatever. I've got friends

8:37 that are software engineers. Yeah. That, I mean, periodically call and go, okay, let me get this straight You guys go out and spend a quarter of a billion dollars on two sections of land, and

8:50 drill 24 wells, and you turn it all on, and you don't know if you're gonna make money or not make money, and you do that without calling the guy next door. Oh yeah. Oh yeah. We might give away a

9:03 secret. Yeah,

9:06 exactly. No, I actually love to hear the, it's not open source, but to some degree, open source to it because I mean we. We all have that deal that we made gazillion dollars on 'cause we had

9:20 some information and we've given it back 10 times by not sharing. And it's ridiculous. Yeah, and I think what everyone is realizing is we're not in this, you know, offshore lease sale mode where

9:35 the value of the acreage can swing 100 X based on some well results. And so you need to be secret agents with super crazy protocols with all your data And in fact, by collaborating more and having

9:47 more openness, the rising tide can lift all boats and everyone in the area can have improved performance and returns and all that. So this is a weird juxtaposition, but it just kind of came to me

10:00 on this side. I have a debate. Go for it. Whatever, you're significantly younger than I am. Do you think part of the reason we have such trouble attracting young people, is because we're so

10:14 close-minded and not, I mean, it feels like this is gonna be, this is gonna be a little derogatory, I don't mean it, but if you grew up with participation trophies and you're not sharing, do you

10:25 really wanna go hang out with these guys? Oh man, is there any truth to that, baby? I don't, that's a good question, but honestly, I think that the more of the younger generation, my cohort

10:37 graduating in the mid 2010s, and then folks who were graduating over the next 10 years, I think it's more about the volatility of the job market and just worrying about the future of oil and gas,

10:50 like even if it's a person who isn't immediately like throwing up in their mouth at the thought of working in an oil and gas company, 'cause that is a huge portion of it, and I'll go talk to

11:01 undergrads at UT, you know, geology students and they'll just be like, I'm not working in an oil and gas. My three kids, my three kids live the greatest, my three kids are going to London for

11:09 Christmas, Should we ban oil and gas? Oh yes. It's so evil. Yeah. Yeah, what's fueling the plane on their way over there? You want to row your way? Exactly. It's going to be warm in that

11:19 polyester coat. Yeah, exactly. Exactly. So,

11:24 okay. You just gave a really cool speech about the DJ basin. Yes? 'Cause it in say you did it What.

11:32 was cool. I've read the slides. I don't want to give spoiler alert. I'll let you deliver it, but it was cool. So I was recently at the Society of Petroleum Engineers Denver Chapter, which shout

11:41 out to SPE in a great organization, still going strong.

11:46 But if you look across all the major shale plays here in the United States, and let's focus on the oil plays, Permian, Eagleford, Williston, Anadarco, DJ, Powder River, even throw it in there,

12:03 almost all of those plays have had declining well productivity over the last couple of years. So, a Permian well that was drilled in 2020 or 2021 was 10 more productive than the wells that are being

12:19 drilled over the last 12 months. I've got a theory on that, but

12:27 when we get to the Permian, we'll put a bow on that one. And

12:31 if you look, that's pretty consistent across the different basins with the exception of the DJ, which has had the opposite trend, where it has gone up 10 over the last couple years. And so we,

12:42 obviously this is interesting. It's an anomaly in the data, things are breaking on the breaking trend. And there's a couple of things going on in the DJ that I think are interesting and are maybe

12:53 things to think about for other basins. But one of them is that the basin is consolidating So 75 of production now is oxy chevron and civitas, you know, just three operators, right? but it's by

13:10 far the most concentrated of the major shale plays. So will a basin which has fewer larger operators have different production? Maybe. The other thing is it's a challenging surface environment.

13:22 Getting permits is tough. You're drilling in the suburbs. It's challenging for a lot of reasons. You have the political uncertainty of will the next election remove your ability to drill some wells

13:37 And so what we did is we took our machine learning models and we ran them on the DJ basin to sort of try and see if we could figure out like what is driving this production increase, productivity

13:48 increase, and can we tie that back to any of the circumstances that are in that basin. And really what we found is kind of a three-pronged recipe for how the operators they're increasing the

14:00 performance. And it's not by focusing on better quality rock It's, the biggest driver is actually pumping more. fluid in their completions jobs, so they have gone from pumping an average of 25

14:16 barrels of foot up to an average of 40 barrels of foot in just a couple years, which is a tremendous increase. So they have been stable for years and years and years, and the operators kind of

14:27 figured out that this was a big deal. And our models think that you can still get a lot more oil for a 40 barrel of foot job compared to a 25. And now 40 barrels of foot would be more like a Permian

14:39 size job than where the DJ used to be. So that's like the biggest difference, but they're also drilling longer laterals and they're being more aggressive than the Permian players. They're drilling

14:51 more, you know, three, two and a half to three plus mile laterals, and then they're doing wider spacing. So it's this interesting recipe where you pump a bigger job, you widen out your spacing,

15:03 you drill longer laterals, and that's unlocking this nice productivity game there, which I think like between the pressure on the operators to get the best out of what they have and a little bit of

15:14 the longer term planning focus that some of these bigger operators have. Clearly they're in the play to stay is why we're seeing these gains affected there to compare to other basins. Okay, let's

15:26 break this down. So number one question and it's gonna be, they're gonna be all over the place, sorry about that. How are we defining better rock? What metrics are you using to define rock? Yeah,

15:40 so we are using our machine learning models taking in rock properties like

15:49 TOC, water saturation, thickness, depth, clay content of the rock. And the models can learn from that, how those different factors drive the performance So our models can come up with really

16:06 accurate maps that are done in a data-driven way. So it tends to be very robust and predictive. And if you're not like assuming like, I need to have 35 TOC, 'cause a lot of that stuff breaks down

16:21 and when you actually get onto the ground. So the models basically learn, hey, the rock quality is better here. It's worse here. This area is more gas prone. This area is more water prone, et

16:31 cetera, et cetera. So thinking that one through, and

16:39 I know we're talking about a whole base in multiple formations, so could be depends, could be your answer. Do you find in the DJ basin that the metrics that matter or the variables that matter have

16:56 a tendency to be linear? Because one of the things we got tripped up on is we actually found some plays where it was binary You needed X amount, you needed 25 feet. you needed X amount of TOC,

17:10 whatever. And you had a great, well, if you had less than that, the analogy I always give is you look at offensive lineman in the NFL and it's 280 pounds. There is not a very rare that you find a

17:26 great offensive lineman, less than 280 pounds, but it's not linear. 330 pounds is not better than 290, it's more binary. What does the rock look like in the DJ? Yeah, we see these non-linear

17:37 relationships all the time, and usually we'll find something like, let's say, depth might be linear, like you might get more, like more productivity, the deeper you are, and that might be very

17:50 - More pressure, look at that. Exactly, I know. Ding, ding, ding. Ding, ding, wow. Thank you, machine learning model, exactly. Even the finance guy got that. Yeah, exactly. But other

17:58 things, like a lot of times the clay volume, maybe it doesn't matter at all until you get above 25 clay. Like that'd be a very common thing. 5 or 15, no difference. Going over that threshold,

18:12 then it starts to degrade. Same thing with TOC, a lot of times we'll see a similar, a similar kind of step change or water saturation. Another theory on the DJ,

18:25 because of the regulatory environment, because of a man-made stuff up above the ground, did, and I don't even know how we would measure this, but did we drill more parent-type wells because of

18:43 surface, and that just explains why things are better, 'cause I think a lot of our problem when we get to the Permian and some other bases is we fucked up parent-child. Yeah. I

19:03 mean, we just did. We, as we were discussing earlier, we didn't share notes and we messed that up Yeah, well, we do see that the DJ. compared to the Permian has less impact from parent-child

19:10 relations, like definitely. Okay. And I don't know, I'll have to go back and check the data. I don't know how much that is just like the nature of the play versus operator design choices and how

19:23 they actually develop the units. But it's much less of an impact up in the DJ compared to the Permian. And we see the bigger impact from the recent impact is due to like wider lateral spacing

19:36 between wells and their neighbors, either one-racked above them or in the same sub zone. You know, and I wonder if that's not driven by at least in part by, you know, in Colorado, it felt like

19:53 you could just get 100, 000 acres. And you know, you had these big federal tracks. Yep. And you could go, wow, this is great And I always thought that's why Denver was so gentlemanly. Because

20:04 if you have a hundred thousand acres, I'd like a sure deal show you can have 20, 000 acres Yeah, we're in Oklahoma and parts of the Permian. Oh my god that three acres are mine You know and they

20:15 fight about it and you know you get the concentration of ownership Which led to tighter spacing etc. I wonder if there's any of that going on Yeah, potentially. I mean, I think the The other

20:29 component of that is there's probably more pressure to build out Contiguous blocks just because of how surface constrained it is. So maybe maybe operators are more willing to Trade acreage so they

20:43 can both drill 10, 000 foot wells. This is I'm just arm waving. Yeah, but I think it makes a lot of songs Yeah, and I think up in you know This is also outside of my bailiwick here But up in

20:54 Canada where you have a surficial issues for other like more of environmental regulations and things like that you probably see similar

21:02 Similar things going on This used to always be our deal. It took us a while. If you called for references on anyone in Denver, you always got 10 out of 10, everybody was fine. And then you'd get

21:14 in there and you go, oh my God, this guy's an idiot or whatever, and all this. Where Oklahoma, the best you could do is seven good references out of 10. And when you trace down the three or more

21:26 bad references, it was always 'cause somebody force pooled some money. Yeah, sure, yeah And nothing to do with quality of human, it was just - That's great. It's the game of Thrones they play in

21:38 Oklahoma. Yeah, it's a different world out there, that's for sure. So what else do we know about the DJ that might be explaining this? I mean, this is gonna pain me to say it, but I'll go ahead

21:50 and say it. So Chevron, a major is actually pretty good at drilling. I mean, is that what we're actually saying? Yeah, potentially yes I mean, I think when we have been looking at this, in a

22:04 DJ, but around the country, a lot of the, a lot of the big dogs that are drilling pretty productive wells. And I think they probably suffer more on their cost models, but they spend a lot of time

22:19 and effort trying to figure out the right design for further wells. And I think for, you know, the key question I have

22:33 about whether or not the, you know, DJ Renaissance can repeat in the Permian as Exxon takes it all over or, you know, whoever takes it all over is, is there a, you know, a nickel line on the

22:40 sidewalk in the case of operators under completing their wells, you know, or have collectively the hundreds of operators out there kind of already figured out what you need to do. Yeah. 'Cause

22:52 that's basically you've played into something when we start talking about the Permian that I'll talk about it. The, yeah, and I mean, at the end of the day, I think you really are just going to

23:01 have three or four operators in DAG. because you have to have an oligopoly to deal with the government. I mean, that's really the way it's going to operate. The other thing that we've got to give

23:13 the DJ credit for too is when it comes to carbon emissions, methane emissions, and all that, better than other basins too. Yeah, definitely shout out to them for that. I don't want to see our

23:28 precious North American resource being flared more than anybody else, acknowledging that wells need to be economic, but I'd rather capture that and use it for making plastics. Yeah,

23:43 that's exactly right.

23:46 Part of that speech, as I was flipping through slides, you did talk about machine learning You talked about the AI learnings, anything pop out at the DJ that we haven't talked about? I think we

24:03 hit most of the major things, but the other slightly unique thing going on there is just how the lateral length efficiency there seems to be better than another place. And I'm not quite sure what

24:18 the exact mechanism is that's driving that, but I think that's part of what has enabled the productivity for those long laterals is just for whatever reservoir reason, it works pretty well Maybe

24:31 it's something about the food content or, you know, who knows, but seems to look better than the Permian for, you know, a 15K versus versus 10K a a 5K. Yeah.

24:42 So one last question on the DJ before we leave, anything out there lurking as potentially a step change function up, or this is as good as it's gonna get? I mean, we didn't like just looking

24:59 through the different trends that the model has seen like other than. Finishing up the transition to higher fluid intensities. I don't think there's anything else that's like at least laying around

25:10 in the data from what we have, but that doesn't mean that there's not other things out there that we just don't really have data for, you know, data analytics is fantastic, but it is backwards

25:19 looking by its nature. So, you know, who knows what other things people might discover out there, you know, like for instance, they tend to be, over the past couple of years, operators have

25:29 been drilling more outside of the Wattenberg core, you know, and they've been finding nice little productive pockets in other parts of the play. I will say this, one thing our models, this is a

25:39 fun little side, this is the Chuck Yates needs a job podcast aside, but the rock underneath Denver would be incredibly productive if we could actually drill

25:50 it. It's, yeah, it looks like as good as anything else in the rest of the play. If we can disguise it as a dispensary, we'd be able to drill anywhere. Yeah, that's one of those things. Put it

26:01 in, put it in, put it behind the dispensary. Yeah, the DJ, the Niebera under Denver, the Marcellus under Pittsburgh, there's a few places out there where there's like really productive rock,

26:12 which is just, you know, fails under the surface condition. So, you know, World War III kicks up and we need some extra resource. You know, you can go into a parking lot in Coorsfield and get

26:22 some nice, nice, The aluminum, the Illuminati is underneath the airport there, right? Isn't that, isn't that the story? Yeah, they're near the resource. Yeah, exactly, smart lizards, you

26:33 know? There was a, I'm blanking on the name of the company. The CEO's name was Dick Lewis, great guy. And he told the story of leasing the airport.

26:46 And so they basically go out to bed and he's studying it and it's trouble getting data from the city and it's hard and all and kind of his people have said it finally, you know. 20 minutes for his

27:01 due, he said, we can't figure this out. Let's just bid10 an acre. So bid10 an acre. Three months later, they get the call back, more data, sharpen your pencil. This

27:11 goes on for 18 months. And finally, I think15 an acre, he leases, turns out for 18 months, they were the only ones that had bid. So the government continued through this process. That's great.

27:26 Of all this. Hey, I got that extra5 an acre Yeah, there you go, there you go, there you go. So, all right, jump to the Permian. Let's do it. What's going on there? Well, the, as I

27:38 mentioned for the context, for the DJ productivity in the Permian has been dropping for a couple of years. And this is enough that like the Wall Street Journal runs cover stories on it, you know?

27:50 My mom notices that. Yeah, like when my dad starts texting me about stuff, you know, it's notable, newsworthy, right? but this is. Obviously cost some consternation because the Permian

28:03 produces 5 million barrels a day and the rest of US. shale produces a little over three. You know, it's more than the DJ, the Bakken, the Eagleford, the Anadarko, et cetera, all combined

28:16 together. Right. Right. So it's Permian is king. We all know that. But so if the well's productivity is dropping, that has you worry And you know, what we, when this was happening and people

28:28 were starting to wrap their heads around it, the big concern was inventory exhaustion, you know, like, have we drilled out the good stuff? Right. It was like, is the Permian dead? Yeah. And

28:38 we, we have spent a lot of time looking at this recently. And

28:44 what we are seeing is that the 2022, 2023 wells, they actually look just as good or better than 2018 or 2019. And I think a lot of the drop that we've seen. over the past two years there, 10 or

29:02 so, is because operators really hydrated their inventory during the low price regime of COVID and its aftermath. So they stopped drilling the tier three stuff. It wasn't a question of like running

29:15 out of tier one. It was more, you know, they stopped drilling tier three. And then when they added it back into their portfolio in 2022, when we had that big recovery and into 2023, it looked

29:27 low and behold, production drops back to where it was, which is not to say that the future inventory looks a little worse than what's already been drilled, but it is to say that like this past

29:40 couple of years drop isn't, I would say, a huge red flag, you know, put up the smoke flare, you know, call up the Saudis and ask them to pump more. It's more just like an artifact of the

29:53 operator behavior during COVID time Okay, so here's my. conspiracy theory, tin foil hat on the Permian. So I'll kind of make up all these numbers 'cause I

30:04 can't

30:34 remember, it's been a while. And plus I don't have a job anymore, so I don't have to be right. Yeah. Let's go horizontal drilling. We start in the Permian, call it 12, 13, something like that.

30:37 Boom, every year we're up 40 in terms of well productivity, right? Then we get to call it 17 and 18, and maybe we're up 15 to 20. So in Biden's world, that's inflation decreasing, right? No,

30:38 but

30:40 anyway, and so we still see kind of up to the right, although with slower growth up, and now we're kind of bending back down 10 What I think happened is call it until 2017. You have five guys in a

30:59 rusty pickup truck for the most part, drilling. That's cane-back companies, NGP-back companies, the entrepreneurial companies, generally drilling parent type wells. Generally, the smaller

31:12 companies are a lot better at optimizing production rates. We're gonna flip, right? We want as high production rates as possible. So we have that, the best acreage in the Perman, Exxon and

31:25 Chevron sitting there, and they're not doing diddly squat. So they pick it up in 17 and 18 with the best practices. 'Cause I think if you took out the Exxon and Chevron wells drilled in 17 and 18,

31:40 instead of being up 15 to 20, it's like down 3. So you already saw the breakover, it's masked somewhat by Exxon and Chevron being late to the game. Am I making that up or is there possibly some

31:54 truth to that? I think there might be some truth to that.

31:58 And in addition to just finding that better quality rock as the larger companies start to move more in their positions, we also see that they're drilling longer laterals. Two guys in a truck might

32:13 not have bothered to, hey, they're gonna flip it anyways. Like who needs to lease this thing up to fit a three mile lateral or whatever it is. And then it does seem like the operators have been

32:24 getting smarter about their parent child relationships and our models are showing like less impact from that. Like in the past couple of years compared to before that. Yeah, 'cause we mess that up

32:35 early on. Just the whole thing of, gosh, we can't share information with anyone. But I bet if you go back and run your models and you tag out Exxon and Chevron, go look at that. 'Cause I think

32:48 you're gonna see that it actually flattened out early. And so this decline is to be expected reason it kept going was just better rock best. I'll take a look, yeah. There you go. Good theory. If

33:02 there's nothing to that, don't mention it ever again. That's not even that much of a tinfoil hat. That's just a normal beanie right there. Yeah, that's just a dude who paid attention. Come and

33:12 give me some beauty gate conspiracy theories for the Permian. That's a real conspiracy theory. Yeah, no doubt. We could sit around and say, okay, who's gonna buy Endeavor? Yeah, sure.

33:24 So there we go. So that's interesting 'cause

33:27 if I'm right about my theory of Exxon Chevron getting there late, that actually pretends to a steeper drop possibly than maybe what your models are looking at. Again,

33:45 how do you see the, how is your model showing the

33:49 decline going forward? 'Cause you said it was just a bit of high grading, drilling during COVID, so 10 down, to be expected, are y'all seeing more flatish or are we up against the brick wall? We

34:02 show more flatish. Brick wall would be a little bit later. Yeah, I mean, most of these, I was just looking at the numbers for the Midland earlier today. And most, like most of the operators,

34:16 especially the large ones, have somewhere between like four and six years of their better stuff remaining And then, you know, another six years of tier three, tier four, whatever you want to call

34:29 it. So I think there's like, you know, there's time left to run with stuff kind of holding steady. Yeah. Gotcha, gotcha. So what other learnings and what else do we need to know about the

34:41 Permian?

34:43 Yeah, so the Permian, which, but by the way, Yeah. I mean, Exxon buying pioneer press conference sounded like you know, a diamond-back acquisition from 2017. Longer letters, federal

34:58 completions. We're gonna double

35:01 the amount of reserves. I'm like, What, money is not that bad, come on. Yeah, no, yeah, they're an excellent operator. Yeah, for sure, yeah. Well, I do think like there is remaining

35:12 potential for longer and longer laterals. I mean, especially like the more you can eliminate anything short of 12, 000 feet, you know, there's still like a lot of meat left on the bone there I

35:23 think it remains to be seen on whether or not there's some completion special sauce or EOR or some new method that will like unlock a step change of productivity. Like we'll see whether that happens.

35:36 But I think the more interesting thing right now is the emergence of all these secondary targets, tertiary targets, whatever you wanna call them across the Permian and like operators in like

35:51 operators the Midland are back to drilling 30. plus Wells in a, in a section, you know, like the cube is back, baby. Nice. It's a, and, and the way that they're doing it, Chuck, is by

36:04 spreading that across, um, more zones. So rather than trying to drill, um, yeah, that many Wells, just in your wolf campaign, your lower sprayberry, whatever it is, they're adding in upper

36:19 sprayberry, middle sprayberry, Joe Mill, Dean, Wolf Camp C, Wolf Camp D, you know, they're maybe, maybe staggering some Wells where they can in different parts of the lower sprayberry. So

36:32 operators are like really taking the original like cube gospel to heart in doing these very complex three dimensional developments. And so we're seeing, I think that's part of the reason why parent

36:45 child is having a little bit less effect over the past couple of years But that's also why you're seeing like that significant density. And just to give you some numbers here,

36:57 in the Midland Basin and in the Delaware Basin, if you add up these secondary zones, like Upper Sprayberry, Wolf Camp D, Dean, like those zones produce more than the Powder River Basin. You know,

37:11 they're like 350, 400, 000 barrels a day. You know, it's like the scraps from the King's Table are a great way to make a nice meal. You know, so that's I think the interesting thing about the

37:23 Permian and why I'm a little more bullish on it to, you know, plateau for a long period of time is just, you know, the sheer amount of like additional data that they're not additional, the

37:33 additional zones that are out there, like big Barnett, big Barnett piece and Forbes or whatever earlier today. Operators are out there drilling bar, super deep Barnett wells and finding a lot of

37:45 oil, actually, not just gas in the Permian. Yeah, and you know, I always come back to, I mean, what are we getting? percent of the oil in places. Yeah, 10 is amazing if we get that. So, I

38:00 mean, there's a lot of oil left down there. For sure. Yeah. And I remember, you know, one of some of the more drizzled old geologists at Conoco Phillips back when I was there would always say,

38:13 you know, hey, 20 years ago, we knew the resource was out there for a few different things There was oil in the shales. There was methane hydrates. There was immature hydrocarbons like in, like

38:28 in Utah. And like, now we figured out how to produce oil for shales, but there's still to your point, 90 of that resource is still there, you know, or more. So if, if, ever anybody can figure

38:42 out how to improve that recovery factor, especially,

38:48 you know, secondary tertiary recovery, whatever you want to call it. There's a lot of hydrocarbon left to be produced. One of my largest LPs became my largest LP because we were talking and they

39:04 just said, Hey, you know, I just don't get this. I mean, you drill down and you drain a pool and then it's gone and I go, No, no, no, no, no, no, no. No, no, no, that's not how it

39:15 works. And I went to the whiteboard and I drew an upside down triangle And then I drew oil price up along the side. And I said, The higher the oil price, those two things, one, it gives me an

39:27 incentive to go figure out how to make this stuff work and it gives me the money to invest, to do RD to figure out how these two things work. So I go, We will never, ever run out of oil. Just, I

39:40 can tell you how much oil's out there and how much is economic. You tell me what the price is gonna be. And a light went off and the guy's like, Oh, okay. And that's why he invested with us But

39:51 yeah, I mean -

39:58 It's just a150 oil, we can get a lot more of that oil out of place. Yeah, I mean, there's like a trillion barrels of oil lying on the surface in Utah. You just go mine and you go put it in an

40:01 oven, it'll come out of the rocks, you know? It's just, you need it to be200 to make it that economic. And there's just all sorts of stuff like that out there that with the right price, you can

40:11 make it work, you know? So let's talk real rock now. Okay. Like super rock, vodka, moata? Yes. So I'm excited, we've got our libertarian dude Yeah. I'm bummed out though, because here's

40:24 what's gonna happen. He doesn't have any control of the legislative branch of government there. So he's gonna get tagged with all the failure, even though none of his policies will be put in place.

40:34 Sure. If for some reason his policies get in place, and we have this libertarian, Mecca, tell us about that play, because we talked about this on BD. the other day, largest number of

40:47 unconventional wells outside the United States. In Canada, in Canada, yeah. In Canada, okay. Yes, they're

40:57 the 51st state when it comes to the day. Yeah, 51st state Alberta, 52 BC. Yeah. Well, I'm all about a little optimism bubble for Argentina, like let's do it. The amazing thing for the Vaucom

41:07 where it is, the recent wells in the oil window are around 30 more productive than the Permian You know, the wells are incredible. And on top of that, the wells and the gas window are that much

41:27 more productive than Appalachia. So imagine having a super Marcellus and a super wolf camp in the same basin, you know? It's extraordinary. And this aerial extent of the basin is bigger than the

41:41 Permian

41:43 or the Eagleford. It's extraordinary I mean, it's not quite as, not quite as stacked as the Permian. but it does have many different benches, lots of different tight opportunities and sandstones

41:57 and siltstones in addition to the proper shale place. So I think the basin has almost unlimited potential. It's a question of how scared are operators that they'll get nationalized?

42:13 Which has happened before So my prescription is so bad on my eyes that it's illegal for me to get lysic in the United States. Now my eye doctor who's been my eye doctor for 40 years has said, I've

42:29 seen enough data out of Europe with people with your prescription. I'm comfortable if you wanna get lysic. And I go, really? I gotta fly to Europe and he goes, no, no, no, no. We just go down

42:39 to Mexico and right across the border We have an office and you go get LASIK done. and it'll be my partner. He flies down with you from Houston and we get it done. And he goes, you wanna do it?

42:52 And I said, no way. And he goes, what are you talking about? I go, imagine me starting the sentence. I got my eyes cut on in Mexico and now I can't see. I mean, who doesn't say no shit to that,

43:02 right? Yeah. So, yes, Argentina. So yeah, there's political risk. The little political risk. Costs have been historically higher there, although I think from what I've been seeing in IRX as

43:14 they've been coming down quite a bit, which is good news for obviously the economics of those wells. But then you have issues around off-take, especially for the gas. I've heard the pipeline

43:26 situation in regulatory-wise, it's just a mess and -

43:33 Yeah, but it's one of those things, and if you can figure it out, you can power Buenos Aires, you can power Chile, you can power

43:43 Rio de Janeiro and Sao Paulo clean gas from,

43:49 you know, that Southern cone just coming from the vodka and where to, if you just put in the pipes, like not to mention LNG, you know. Cool, yeah. Yeah, no, again, it's one of those things

43:58 that go raise a big huge fund. We're gonna invest there. We lose all our money when it gets nationalized. Yeah. Should have thought it, but that'll be interesting to watch 'cause I mean, he's

44:08 talking a good game. He's talking about, we're gonna sell, you know, we're gonna sell this off and, you know, we're gonna free up growing, so that'll be interesting. Yeah, I really hope, I

44:20 really hope he's able to be successful with all that 'cause of the potentials there. And I mean, that's the key basin, but there's other interesting basins there. There's, you know, offshore is

44:32 also interesting there after all this stuff in Namibia coming up. You know, it's, I don't know if you've been following that, but there's billion barrel discoveries going on in Namibia and, you

44:40 know, it's right across the conjugate margin, what used to be connected to Namibia's. offshore Argentina, offshore Uruguay, like it's all that, like that whole trend, like the South of Southern

44:53 Africa, going from South Africa up to Namibia, it used to be connected to that Eastern part of Southern South America. So, I mean - I love our analogy is literally in a different hemisphere. Is

45:07 that a different, other side of the world? Yeah. Here's our analogy. Hey, this is thinking of like a geologist, you know? I used to be scared two offsets away. That's how old I am in my career.

45:18 Yeah, we can't move over there. Yeah, well, you know, Guiana used to be offshore, or like, you know, used to be the same geology as like offshore Ghana and those, where you had all those

45:28 initial discoveries and that kind of, that coast of Africa, western Nigeria. So, that was a lot of the reason for them to go explore and Guiana was knowing that

45:39 200 million years ago, they were in the same place Nice.

45:44 Crazy there's something going on in the uenta. I haven't heard about this. Yeah. What's going on? I love the uenta. It's a it's an incredible basin. So is that the waxy? It's waxy, dude. Yeah,

45:55 um, so Here's I'll give you the good news and the bad news. So good news. The wells are as productive as the Permian Um, and if you get the crude to market because it's waxy and can be more easily

46:08 turned into lubricants and other Like higher value added petrochemicals. You you actually get a positive differential relative to um, you know wdi um, so And it has um in addition to just really

46:18 productive wells. It has stacked pay So it's it's you know,

46:27 not just one zone. There's a number of zones to target there Um, it has a pretty small footprint. So Like it's not going to be another you know, it's not it's never going to produce five million

46:36 barrels a day But it's been like going up like crazy. I think the unconventional is there now through a hundred thousand barrels a day And it's not like, it's not, you know, there's this kind of

46:48 common misconception that I think has maybe delayed or hindered capital getting into the basin, which is that you can't get the crude out of the basin, but in reality, you can truck it out to rail

47:01 terminals. It's just you had a truck it farther than you would like. So, and they're starting to, they're getting approvals for building more rail terminals and rail outlets that'll improve the

47:12 transportation costs even more. But it's really fantastic. You've got companies like XCO and Eventive and the Javelin team out there like drone really nice wells. Yeah, no, I'm blanking, but

47:26 there used to be, I think, two publicly great petroglyph, maybe, and something else. And we used to, it seems like we would look at it and pitch it as an acquisition target to somebody and

47:38 they'd spend a lot of time that ultimately got scared because there were the two refiners in Salt Lake City. Yeah. They'll refine it. Oh, they're about full. No, it's too bad. Yeah, but no,

47:49 that's wild to hear. That's wild to hear. All right, give me a really, really cool story about working with a client, and then I'm gonna ask you the flip side of it too, just when you're

48:03 celebrating a little too much, I'll ask you the

48:08 horror story. Okay, well, I'll give you a cool story, and I'm gonna anonymize this, 'cause I don't have permission to share the full thing, but we had a client in the Permian who was using our

48:19 models, and our models were saying that there was a primary zone that they were targeting, and our models were saying, hey, the geology of the zone above really matters to

48:23 your primary zone, and they're like, that doesn't make sense.

48:34 It's all filled with water. Like, why would it matter? And it just kept showing up in the model results, And then they went and looked at some more of the data, you know, hey, actually our

48:45 wells that are drilled near the top of our primary zone are the best wells. Like that doesn't make sense. Frax normally grew up and they're growing up in a more water mystery, right? Well, they

48:54 went back and looked at their data and went back all the way to core and they found, oh lo and behold, it's not full of water, it's full of oil. And the models had kind of learned that just by

49:05 looking at the data and seeing these relationships. And so after finding that out, this company added in, started targeting that zone specifically and it became some of their most productive

49:17 inventory. So like huge uplift to their nav from, I don't know, bumping their locations 10 or 15 by finding this new zone, like thanks to the models, finding this unexpected kind of insight, if

49:30 you will. You know, it's interesting that you say is when you look at our track record over funds four, five, six, and seven. Generally speaking, we got better over that time and why we got

49:52 better was not because our best deals got better. They kind of stayed the same. It's that our worst deals we spent less money on. So one, we learned to cut bait earlier. And two, it was exactly

50:09 what you were saying, being more just agnostic to whatever and just looking at data and figuring out because the thing you forget about the shale revolution, it was the, I mean, it shouldn't have

50:21 even worked if all the tradition, you know, the traditional beliefs. I mean, it was literally, let's go down there. Let's go. I mean, who the hell had the balls to do the first fine mesh

50:32 sandcrack, right? Now we're just going to come up the reservoir. Why are we doing this? Lumba, look at how it works. I like to think about that. The tools that we have for producing are over a

50:45 hundred years old. And I think about that a lot for geological evaluations. We have very well-defined processes and ways of estimating reserves and risk and all that type of stuff that are super

50:60 well-defined for conventional. And then you get to unconventional and we struggle even just building a physical model that will replicate more than a single well at a time. And so it's like how much

51:12 more is out there that we don't know yet So I think that humility and just, all right, let's just start from the ground up with an open mind is - Only change one variable. Yeah. Only change one

51:23 variable, yeah. Yeah, that was great. Now give us a horror story. Man, so we

51:30 have, you call this a recurring nightmare. Okay. Maybe is what you wanna call it. So, you know, we've been in business now for eight or nine years, I guess. I've been there for four and so

51:43 we've worked you know, operators and every major jail play. And two things, two things we hear over and over again. I'll give you some specific examples for each, but like one of them is we'll be

51:55 in the pre-sales process and they're like super enthusiastic and they are gung-ho and ready to get started and start using this new ML stuff. You know, the board is supportive. Let's go, dude.

52:09 And we're like, great. So once you give us your data, it'll take four to six weeks to get stuff

52:17 up and running or whatever it is. And this is faster now, but we'd say something like that. It'd be like, great. Our data

52:30 is in excellent shape. You'll have it on day two, like fantastic. I mean, like six months later, still don't have any data. And this happens over and over again, and this is just so much

52:35 potential there for operators to clean up their act where they just, You know, the completions data is. is stuck in some database and they don't know where their cluster data is and their

52:48 production data. Oh, we only have monthlies. We don't know where the dailies are. We lost the data from that company that we acquired. And like next thing you know, we have to end up using our

52:58 public data because even a big sophisticated operator just like doesn't have their data anywhere. So that's part of the reason we have this data offering in the first place. Second horror story.

53:08 And yeah, no, I totally believe that And that it's stunning to me that you walk in and there's like a paper ledger with production data. You're like, Holy cow. Yeah, yeah. They're like, I

53:22 don't know, Tammy's the only one who knew where that is. She's gone now. No, that's exactly right. I've heard that a million times. The other horror story and this happens again all the time is,

53:37 you know, they'll say, Well, how accurate are your models the forecasting production, Ted. You know, usually, you know, once we get them fine-tuned and everything, you know, for most wells,

53:47 they'll be, you know, 15 to 20, like, or less, you know, off from what the well production is. And they'll go, Ah, that's terrible. Like, how can it be so inaccurate? We go, Well, hey,

53:58 why don't we do a little test here? And we'll, let's go find, you guys did some AFEs back in 2021, a budget cycle, like, Why don't we take your reservoirengineering pre-drilled type curvesand

54:09 just put it to the test? And, you know, we had an operator once who over-forecasted their whole base in production by 50 for

54:21 a two-year period. You know, how do

54:25 you think that impacted their IRRs and so many wells that didn't end up paying out 'cause they had over-forecasted that badly? And we've seen the opposite too, by the way, companies under-forecast

54:35 all their wells by 30 And that's because maybe they had the opposite cultural. pressure, where every Welsh should be exceeding your forecast. So nobody wants to say that a well is going to produce,

54:46 you know, whatever,

54:48 700k barrels, and it comes in at 600, right? Yeah. Yeah. No, that's, that's not surprising. So let's close out on this, because it is all the buzz these days. And, you know, this is wild.

55:03 We actually at Digital Wildcatters, and we built this internally for our own use. We have an AI-driven search model for all of our podcasts, all our presentations at Energy Tech Night, and, you

55:20 know, index stuff, you can literally type in, you know, Colin's face, and it'll pull up everywhere. Colin's face is you,

55:27 you can, you know, type. So anyway, I have gone in the last nine months from saying Google is the greatest website ever invented. It sucks. I've read these articles and all this stuff.

55:41 Give me AI stuff about, and you've kind of through all this talked about what it's done today. So maybe you can summarize that if you want, but I also want to hear where it's going. Yeah. Well,

55:56 the biggest developments recently are in these large language models that do really good jobs at handling text, or there are similar type models that do really great with imagery or other kind of

56:12 visual data. And I think there's a couple pieces of real potential there for the oil and gas space.

56:23 Those new chat GPT and all those things honestly actually doesn't affect, but we do that much. If you're just forecasting how much oil is going to produce, those aren't really applicable. But

56:34 there is tremendous potential for digitizing old data.

56:44 back to our previous conversation. These models are great at image recognition and you got some paper logs that nobody ever scanned that's in the bottom of a filing cabinet that smells no problem.

56:51 Like just run it, you can do much better job of digitizing all that. So this is great. Just a side note 'cause it's a funny story 'cause you're so young. I feel as the old guy in the room, I need

57:01 to share. Back in call it late '90s, there was a company called A2D, analog to digital And basically, they would digitize paper well logs. And literally, by hand, scanning it to do you know

57:19 where the guy found his best employees? No, where's that? Nail salons. The ladies that did nails were the best at tracing the well logs. And had the most accurate - Excellent motor control.

57:33 Excellent motor control. Find attention detail. Isn't that crazy? That's how we used to do it back in the late '90s. Hey, I believe it. Yeah, that's great. Yeah, and then I think the other,

57:44 you know, Google's becoming obsolete to you. I'm finding myself going to chat GPT for a lot of my research questions now at this point as well. And I think you'll see companies and a lot of them

57:55 already going after this, kind of like you guys are doing for all your content, do the same thing for their, everything that they have across their enterprise, you know, PowerPoint presentations,

58:07 GIS documents, images, PDFs, all that type of stuff. And, you know, I remember back in when I was getting started in the industry, there was this whole great crew change, great crew change,

58:21 you know, as this whole meme that all the boomers are gonna retire and we're gonna have to hand the keys over to Gen X and the millennials and like, what are we gonna do with all that knowledge

58:29 being lost? And I think we kind of forgot about that just because we had bigger fish to fry, like how do we stay in business? Yeah, you know, mine is to37 oil.

58:40 Yeah, yeah, we were happy to do a great crew change because we you know, and then we had fewer employees or whatever, but That's still a big deal Especially as we have to return to the offshore

58:51 return to conventionals for you know The next 20 million barrels of production growth globally, right? And I think these models will play a huge role in that. Yeah Now that's that's interesting.

59:01 It's been fun to watch an amazing to watch and Feels like I know it's you know the overnight discovery that's been worked on for 20 some odd years. Yeah, still It's been amazing. So Ted, how do

59:14 people reach you if they want to chat? Yes a couple different ways You can find me on LinkedIn Ted Cross you can find me on Twitter at Ted Cross or Shoot me a message T cross at novielabscom for an

59:30 email. Oh It said you were cool to come on. Yeah, thanks Chuck. Really appreciate it Hey, one thing we didn't talk about, yeah, I don't know. this even matters is one of the bigger than I

59:44 expected variables. Maybe that's the way to put it was actually figuring out artificial lift after the fact, because it felt like in the Bakken, fuck,

59:58 really? The jet pump burned out again? What are we going to do now? We'll go to a rod pump. Oh, shit. It just went down 30 barrels a day. What's going to be flatter? Do you see any of that in

1:00:09 your data? That's something that we really don't do too much with, because the data quality for that stuff is terrible. It's really terrible.

1:00:21 It's hard to find that good quality data publicly, and then it's hard to find it as a good across the well life. Companies are like, yeah, we want to do an artificial lift study. We're like,

1:00:30 great. Let's start next week, and then they give us the data, and it's in terrible, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know,

1:00:38 you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know,

1:00:38 you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know,

1:00:38 you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know,

1:00:38 you know, you know, you know, you know, you know, you know, you know, you know, you know who work in

1:00:41 that space, but that's one of the big unknowns to our models that like one day, if we had all the data, it would be fantastic. And you can probably optimize a lot more production there. Yeah, no,

1:00:49 and we had a couple of companies that were really, really good at watching that and documenting it and,

1:01:00 in effect, using human intelligence to kind of watch it. But even those guys would get surprised Man, it just looks like it, 18 months, it runs out and we gotta go to a rod pump or whatever in it.

1:01:13 I mean, we wound up having,

1:01:16 we drilled a lot of, I don't even think the guys would mind if I mentioned their names 'cause they did a great job, the cracking guys in the bokeh. I mean, they've drilled as many wells as anybody

1:01:22 out

1:01:24 there and they're really smart, they're really good. I mean, we wound up having multiple type curves over time and it was all artificial lift driven. It wasn't rock, it wasn't completion, it

1:01:35 wasn't any of that, it was literally just figuring out. Yeah, where do you change out the pumps? Yeah, it's a huge deal, definitely. Ted, you were cool to come on. Yeah, my pleasure