Weekly Update
What am I doing; what is coming?
Busy week behind the scenes, and I’ve been a little quiet, but figured I’d give a quick update, just so you know I’m not sitting on my hands over here. Though, to be honest, I’d been probably working 14 hour days since January trying to get the initial launch out prior to my birthday, and then took a couple of days to relax before getting back to the grind. Anyway…
Data
Pushing through a well data update. Aligns with the update cycle for the various states, but my last update was February so working through that currently. Less important than the overall capex data, but I still love having the context, as capex data in isolation feels useless to me without the associated well data. Additionally, if I work up an insight on the data, I’d like to have the most recent data to do so.
I’ll be pulling in recent surveys also, and processing them. I’m working on ways to process surveys for Oklahoma/Louisiana, which I have, but that are quite poor pdf quality and if at all possible, I’d like to avoid using AI-scrapers (which are great but cost money). Still, that will probably be the final answer as it is still cheap compared to hiring overseas labor to do so.
I realized that the post-2020 Oklahoma capex data was not being joined correctly to the main dataset, so I will be adding that in this week. Wanted to begin working up a quick walk-through of the COP divestiture of MRO’s Anadarko acreage and realized that was not in the dataset. That acreage probably ain’t moving that quickly, so I’ll probably push it a week.
If I can fit it in, I’ll process more New Mexico AFE’s that I have in the backlog. I’m looking at using some external labor/AI scrapers on those as well, as the first round was done via regular PDF scrapers. As far as other states (New Mexico/Colorado), I have some AFE data, but it’s not a meaningful enough sample set IMO to be worth adding in quite yet, but we’ll see. You can always forward me some at bd@afeleaks.com if you’re feeling generous.
Insights
Original plan was to push through that MRO workflow, but now I will probably pick up a quick capex/productivity insight on the Waynesville. I’m sorry, I know some Comstock folks followed me, but it’s not like I’m throwing anything out that you haven’t divulged in the last presentation anyway (well, except for F&D/EUR metrics). I will say, the most transparent company in the L48 when it comes to well costs is CRK, and it aligns basically perfectly with our dataset. Additionally plan to work up an insight looking at cost trends in the Delaware for long laterals. Not to get too into the weeds, but essentially what needs to happen is comparison of like-for-like:
Do we have an operator that drilled a well in call it a 5-mile radius in a similar timeframe at a different lateral length?
Was that well in the same formation, or close (ie WC A vs B)?
Were they both drilled as pads, or as standalones?
Just looking at average costs at different lengths in a large area is not very viable, in a similar way that using a single type curve to model your inventory is a garbage approach.
As far as I can tell, that leaves like 20 pads in the Delaware that I can compare against, which is not what I’d call a robust dataset, but will give us some insight into what is happening.
I’m also working up a weekly rig-update insight that will largely be automated, but will be one that I can send out to all subscribers regardless of tier, so hope to have that one set up by end-week. If there’s anything else folks would like to see, just let me know. I’ve toyed with just a daily update that features News and stock price deltas, but haven’t gone through the process and not sure if that’s what you’ll value from this platform.
App
Not to get too into the weeds, but because I set up the business data through Google Workspace, I decided to use Google Cloud SQL to store data. Apparently, the direct connection between apps deployed to the Cloud in Google and their SQL has tons of problems, so I had to work through alternative methods, which did work. That problem solved. Problem 2 was that the app was not recognizing users that had signed up through Substack, so had to revise that workflow. Both issues are fixed now, so good to go.
Upcoming Data
I have a good dataset of supplier-costs, so next week I plan to parse through that a bit more aggressively, and marry that with an insight so that I can also get that dataset into the app. There are some good, largely Permian, stats there as far as things like day-rates for rigs and other various services, so I’m excited to get that one out there.
Finally
I appreciate greatly the initial reaction here. I’ve always felt that truly transparent cost data was a blind spot in our industry, and this is my attempt to rectify that issue. I’ve seen other people put out “should costs” and the like, but to me, actual data has value to everyone. So keep with me and let’s grow this thing together.
Brief story time: For my #EFT followers, AFE Leaks was basically conceived on a random thread within then-Twitter during Covid, and I honestly thought it was a great idea and ran with it. So in a way, this is truly borne from the #EFT community, which is a great resource for what is happening on the ground within our industry.
Get after it folks.
-AFE