Type Curves Revisited
Data Science & Analytics

Type Curves Revisited

Rosemary Jackson  •  

Type Curve: Provides a time series so you can do economic time series calculations and back calculate how aggressively you produce.

Pad Designer/Frac Hit: In unconventionals that 12 month cum number is so important because you’ve already produced 65% or sometimes even more than that off the total production you’re going to get out of the well. Which is why being more accurate with Pad Designer/Frac Hit provides a better economic sense of the viability of how much spacing you should have, how to bring in the infill wells.

Well spacing, with a confident, high level of accuracy, is now achievable in the Petro.ai platform.Type curves have been rethought and reapplied within the geomechanical context of unconventional pad development. Type curves used as a baseline combined with the contextual shale understanding of the Frac Hit and Pad Designer apps are an answer O&G has been waiting for.

Petro.ai pad designer for intelligent well spacing

“We need to be careful about disparaging type curves too much,” Kyle LaMotta, VP of Analytics cautions. "They’re used very widely. What we’ve done with the Petro.ai platform is to rethink what a type curve is, because at the end of the day you still want a time series prediction so you can run that kind of cash flow, get full scenario economics and plan out different wells. Rather than doing a copy and paste of 10 wells, maybe you’ve got 10 unique type curves that are based on capturing a lot of the complexity of the spacing, the sequencing of the development, and the geomechanics.”

Dr. Nitin Chaudhary, Senior Data Scientist, adds,

“We’re constraining the type curve for that local geomechanical variation. We’re building on top of it. We’re scaling it based on that. You can still use your type curve and improve it with the Pad Designer and Frac Hit. Then go back and do more of the same type curves because you would have now accounted for the local geomechanics, the local stress state.

I think the biggest problem with how people approach type curves is they’re far too generalized. I’m going to make types for a vintage 2014 type curve or a 2015 type curve or even if you type per formation, it’s usually too general for what is actually happening at the well level.

“Another important issue is when you end up using type curves for well planning. With the type curve you’re never able to capture the geomechanical constraints that you face in shale development. How the fracs grow can have a huge impact on the production. Type curves have no way of knowing that granularity or getting to that level, which is why Petro.ai knows that type curves should be used in conjunction with the Pad Designer approach. You use the Pad Designer to get that 12 month cum number through the geomechanical constraints and then you apply that to the type curve. That’s much better than blindly applying a type curve.

It also averages out the spacing. It’s usually very hard to make type curves for specific benches especially if you don’t have the Frac Hit app. Otherwise, you’d have to manually calculate spacing groups from the wells and then make a type curve and then try to use that as a guide. But even with that process, you have no context of the vertical frac growth and hence you have no context of how geomechanically constrained you are in a specific location.”

Frac hit app in Petro.ai

The type curve has always had mixed reviews. In a 2017 JPT article, the use of the type curve as the go-to answer for what a well produces over time was evaluated as less than adequate for an industry standard. The type curve calculation method, they concluded is “deeply flawed. The negatives have become all the more pronounced by the rise of unconventional oil and gas developments, where the output varies, and is unpredictable from well to well, with decline curves that do not fit conventional norms.”

Dr. Chaudhary continues, “Type curves can fit into our workflows as a baseline comparison because if you think about a type curve you’re looking at an aggregate of a lot of wells. So, if you take the P50 type curve then by definition you’re only going to be right 50% of the time.You know that’s a wrong estimation, but it could be a good baseline to compare all of your wells against and also your predictions against. So, you can compare the predictions against a type curve to see if it’s a reasonable ballpark.”

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

© 2020 Petro.ai All Rights Reserved