With the current oil price downturn, many have started to wonder whether activity in the Bakken will ever return to its previous heights. Using our machine learning models, we estimate that <10% of the remaining inventory is Tier 1, with the real number perhaps <5% due to surface constraints. With lower-tier locations predicted to produce an average of only 40-70% (2-yr cum oil bbl) of a Tier 1 well, that’s challenging news for production outlooks and breakevens.
How does novi define tier one acreage using machine learning?
Novi built a machine learning model to predict oil production using completions header info and well logs from the North Dakota Industrial Commission. We then produced a model artifact to represent rock quality that we call geoSHAP (we’ll have an entire presentation on this at URTeC, shoot us a note if you’d like a copy of our accepted abstract). The distribution of rock quality is trimodal, easily separated into Tier 1-2-3 (see histogram and map above).
The following video goes into more detail on how we built this subsurface model:
Novi video – a workflow to build a subsurface model using logs and tops
How many remaining well locations are “Tier One?”
In order to determine remaining inventory, we started with the wells that were already drilled and assumed any remaining inventory would be 600’+ away. We looked at 660’ spacing (base case) and 330’ spacing (optimistic case) to estimate remaining inventory.
This analysis yields only 1,560 potential (base case) locations of Tier 1 Acreage. We classify 45% of that acreage as “at risk” — it hasn’t been drilled so far due to rugged topography, sensitive environmental issues, or other surface constraints that may mean those wells are never drilled. That means 858 “low-risk” locations.
How much inventory is left under differing rig counts? In 2017 & 2018, the industry drilled a total of 1008 Tier 1 locations — about 500 wells/year. At 660’ spacing, the low-risk Tier 1 inventory could be drilled up in less than two years. Here’s what the remaining inventory looks like at different spacing and rig pace scenarios:
What does this mean for production & economic outlooks?
In the chart below, I am showing the Novi average predicted oil production by tier out to 720 days. Clearly, Tier 1 acreage is expected to produce much more oil than Tiers 2 or 3. To make money in today’s price environment, operators will have to tailor their capital investment decisions to the subsurface conditions present at each well location.
- Lack of high quality Tier One acreage is clearly a headwind for the production outlook for North Dakota. Not only will the same amount of capital produce less oil, but operators are likely to continue directing investment to other plays.
- That said, there is still money to be made outside of Tier One acreage, particularly given depressed valuations. This means Acquisition & Divestiture (A&D) opportunities opening up for well-funded PE-backed firms.
- Both new entrants and veteran operators moving out of the core will have to carefully analyze a broad set of spacing/stacking configurations and completion designs to maximize returns.
Maximizing returns as operators move to lower-tier acreage is the subject of our Novi Masters Series: Core vs. Fringe in the Williston Basin. Check back soon for the next blog post, where I will expand further on strategies to improve development economics in low-price environments.