Recommended Data Types for Calibrating Petro.ai Well Designs
Data Science & Analytics

Recommended Data Types for Calibrating Petro.ai Well Designs

Rosemary Jackson  •  

Q: What tests do you recommend doing for the Petro.ai model calibrations?

A: “Petro.ai can integrate any data type into our subsurface modeling,” Kyle LaMotta, VP of Analytics explained, “There are some main ones that come to mind.

“A good place to start is with DFIT which is a diagnostic fracture injection test. If our client has DFITs in the different intervals that they’re targeting, that works as a calibration point for Petro.ai’s stress profile. Most operators will have some of these data points. If they don’t have DFITs, because doing a proper DFIT takes a few days to monitor, they can do a mini-frac.

“A mini-frac is typically done in the toe stage. It’s where clean water is pumped in, the well is shut-in and the water allowed to decay for a few hours. You don’t get as many diagnostics out of a mini-frac but it’s still a good measurement of SHmin. It’s easier to incorporate because an operator can do a mini-frac before each frac job in the toe stage and that gives you a good idea of the frac gradient.

openhole mini-frac test in cap shale formation

Microseismic data helps Petro.ai to understand the shape of the fracs. Microseismic gives a good idea of the half-lengths and the heights, where the fracs are growing. It also helps in understanding depletions, if an operator does microseismic where they’re fracing a well with an offset parent. They can see if the offset parent well has depletion that’s drawing the fracs in that direction. Microseismic is an expensive operation taking a lot of planning and set up so it’s not always possible. You need monitoring wells and a lot of surface equipment. Petro.ai can operate without microseismic and still achieve a high level of accuracy.

Frac simulation output from software that uses physical inputs to calculate the expected frac geometry, can be used to train our frac fingerprint. From their frac simulator we can get the expected frac geometry. Here again, we don’t have to have simulator input to achieve a high level of accuracy.

DFIT

Operators could do a pressure interference test. There are different ways of doing that but essentially, they’re looking for pressure communication between producing wells. So, they could shut in one well and produce another well and see if there’s a pressure response on the one that’s been shut in. Or if there’s any kind of production swapping, they might see that if they shut in one well and the well next to it starts producing more which means there’s some connection between those. That helps determine interference and communication between wells to see if they’re connected or not.

In terms of well logs, companies will take their triple combo which is a variety of different things that they’re measuring for and from that do some processing on those to get some of the petrophysical properties such as total clay content, total organic content, water saturation, oil saturation.Those are really helpful for Petro.ai. Those properties can be used to calculate SHmin. We can use those well logs to calibrate our stress profile and also determine some of the target intervals based on the oil saturation or porosity or permeability. Using specific logs, the operator can do petrophysical calculations on them and Petro.ai can use those interpretations to form our models.

The operators will also have data that Petro.ai can use to generate ISIPs.On every frac stage, someone at the service company will be picking those ISIPs. Petro.ai can use their ISIPs if they trust the picks or we have a tool that will automatically calculate that ISIP based on the treatment pressure.”

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

© 2020 Petro.ai All Rights Reserved