DeepData provides the option to automatically design your stages to group similar rock. This enables a more uniform frac. For example, if a stage combines both brittle and ductile rock, the ductile rock won’t fracture (without expensive diverter), thereby reducing the efficiency of your stimulation.
THE SOLUTION WE OFFER
WATCH TO SEE HOW YOU CAN MAKE COMPLETION DESIGN
Optimized for Each Well’s Geology Group Similar Rock in Stages Pick Your Perfs by Geology
Geo-Aware Design Optimization
DeepData is designed to enable geologists and engineers to collaborate on completion design, while reducing the design time. Reducing design time is critical, because this is the primary argument against engineered, or geo-aware, completions. The completion engineer has limited time to design each well, and a geometric design–where every stage is the same and you use one pump schedule–is the only option…until now. DeepData customers report that completion design times are actually 50%-75% faster than geometric designs using Excel.
DeepData incorporates your geologic insight in a variety of ways. You start by loading curves (LAS files) for each well. Then you define ColorBands™ turning those busy curves into actionable insights or groupings of rock.
There are a number of ways to optimize your completion designs that will make significant improvements in well economics by both reducing costs and increasing production. All of these optimizations are driven by designing your completion based on the rock along your lateral. Here are just a few examples:
GROUPING SIMILAR ROCK
Chose any curve (.LAS file) and DeepData will shift your clusters to target the optimal locations to hit the best rock. It doesn’t matter whether you use Gamma Ray, Drill bit geomechanics, cutting analysis, anything. You set the constraints and the target and DeepData automes the rest.
DeepData’s patent pending process enables you to define rock quality from any combination of data. Then you can apply physical designs and pump schedules accordingly. For example you might define rocks types as: small, medium and large. When you are out of zone or in poor rock, you might reduce proppant loads by 30%, while boosting it 20% in high-quality rock. This alone can boost production while significantly reducing costs. You can apply allsorts of strategies like this, being as aggressive or conservative as you want to be.
MACHINE LEARNING - COMING SOON
Machine Learning enables us to correlate rock properties, physical design, treatment and results. Running all of this data through machine learning will soon drive continuous optimization.
MACHINE LEARNING - COMING SOON
From Mass Manufacturing to Geo-Aware Mass Customization
The upstream E&P business is evolving along the same trajectory as automobiles. Initially all cars were custom built by hand. Then Henry Ford introduced the era of mass manufacturing, where every car was the exact same. You may have heard Henry Ford’s famous line: “Any customer can have a car painted any color that he wants so long as it is black.” Then the car business transitioned to mass customization, where each car could be tailored to customer demands. The upstream E&P business is currently in the mass manufacturing mode, with geometric completion design. DeepData is a tool that ushers you into the next phase, mass customization, where every well is optimized based on the rock properties of that unique well.
WHAT OUR CUSTOMERS SAY?
“Our neighbors were emulating our completion designs, now each design is unique to that well.”