Bringing Certainty to Shale
Data Science & Analytics

Bringing Certainty to Shale

Rosemary Jackson  •  

Q: How does Petro.ai bring certainty to shale?

A: Richard Gaut, CFO of Petro.ai leads us into the discussion, “The biggest problem with shale development historically is over promising and under delivering. People assumed that new wells would perform as well as the old wells, or the next well would perform as well as the previous well. What they didn’t properly account for was parent/child interaction. The drainage of that previous well. So, they ended up way over promising and underdelivering.

That poor ability to forecast is uncertainty. You don’t know what your returns are going to be when you start a new project. What our tech does, it gives you high certainty about what the performance is going to be when you drill a new well.

“And the way Petro.ai does this,” Dr. Troy Ruths, CEO of Petro.ai continues, “is through understanding and characterizing the reservoir, the geology and the geoscience that affects productivity. At Petro.ai we account for the variation in shale and those productivity drivers.

Photo on Unsplash by Ricardo Arce

Certainty has two aspects: accuracy and precision. Accuracy is how close you get to the bullseye and precision is how big the errors are from there. Are you clustering or are you spread out? You could have high accuracy but low precision. It’s easier to just think of accuracy, but we’ve tracked both.

“It’s important because nothing is fully understood in shale, we’re still learning. Without a sense of accuracy, you can’t know how the changes you’re making are affecting the productivity of a well. Accuracy gives you something that you can build trust in because you can compare to analog pads. In the process of comparing to analog pads the high accuracy builds trust. That’s a big part of using the system to optimize a particular DSU or investment opportunity.

“If we look at the definition of certainty, it can mean ‘a firm conviction that something is the case.’ That speaks to correlation versus causality. We take a lot of time to make sure we’re not chasing correlations but are finding causes. That’s why it’s important to take the reservoir and the geology into account.

“Certainty also is, ‘a quality of being reliably true.’ Petro.ai certainty means that we bring a consistent process, a reliable strategy. We follow the same process every single time we build a Petro.ai model. That means that if you trust the steps that are going into it, then the outcomes are reliable. We’ve definitely seen that across every shale basin.

“Let’s look at the status quo that Petro.ai is changing. The status quo before Petro.ai lies in the realm of the uncertain. Uncertain because companies don’t use all the data that they have access to. The fidelity of translated information that Petro.ai is building with all your data and applicable AI means the results are transformational with a deeper, more confident understanding of the subsurface.

Part of the Petro.ai process is the data you feed the system combines with the assumptions you put in. So, an example is spacing degradation factors. The way most customers have a status quo solution for spacing degradation factors is they have a fuzzy coefficient that they change based on what they’re seeing in the field but that could vary widely. It could be tight in certain formations, in certain reservoirs, in certain development plans.

“Customers are responding to our approach because we’re feeding in new data types that have never gone into these kinds of systems before but have always been a part of the engineer’s mental models and their intuition of what’s going on. Petro.ai builds certainty into our models because we’re pulling in all the data, all the information that our clients would use to create their own intuitive model.

“Shale has been built on the intuition of reservoir engineers, geologists, and completion engineers working together. There really hasn’t been a system that can unite all of those models and data types into one place. So, from what I’m seeing I think that drives a lot of confidence in our process.

The other big thing is transferrable learnings. When it comes to certainty, you want to know that you can transfer what you’re learning from one part of your acreage to another. It can’t just be a unique experiment every single time. With Petro.ai, you’re investing in your understanding and characterization of the whole reservoir.

Photo on Unsplash by Hitesh Choudhary

That also gives our clients’ certainty as they’re collecting these data types. They can give it to a system that gets more accurate. As we give the data types to the system, the system does get more accurate. We see accuracies go from 70% up to the high 90’s when clients go through the Petro.ai process and have loaded all the different pieces of information into the system. The accuracy that’s part of the Petro.ai algorithms will be the new standard for the oil and gas industry.

PETRO
TEXAS
114 Main Street Ste. 200
Houston, TX 77002
CANADA
4-5609 Avenue du Parc
Montreal QC, H2V 4S8

© 2022 Petro.ai All Rights Reserved