Real-Time Completion Logic from Same-Well Measurements
Treating Pressure Does Not Reveal Downhole Distribution
Treating pressure is the primary signal engineers watch during a frac job, but it is a convoluted response of pipe friction and perforation friction.
Pressure changes imply that something changed downhole, but not whether clusters are distributed evenly, perforations are dropping out, or near-wellbore conditions are degrading during the stage.
Fiber optics have proven the value of real-time subsurface feedback, but cost and complexity limit broad deployment.
How SAFA Enables Real-Time Completion Logic
SAFA deployment requires minimal additional hardware at the wellsite: a high-frequency pressure transducer on a surface treating line and a data acquisition unit tied into the frac data stream.
The software continuously monitors pressure and rate and automatically triggers analysis when it detects a rate-change event.
Planned analysis points are introduced by briefly changing the pumping rate early in the stage, mid-stage, and near stage end. Each event updates the live metrics within seconds and logs a traceable snapshot of near-wellbore performance.
Those snapshots are connected to a configurable decision tree. The workflow starts with an initial efficiency check, re-evaluates trends mid-stage (especially as slurry conditions change), and governs end-of-stage actions (continue, optimize, stabilize, or exit).
The ultimate objective is repeatability: consistent decision-making criteria leading to actions grounded in same-well measurements.
Operational Examples During Pumping
Flow-velocity tuning: Each formation exhibits a flow-velocity window where cluster contribution and uniformity are maximized. Because metrics are calculated live, rate can be tuned to stay in the optimal band. One example increased rate from 90 to 99 bpm (flow velocity ~32 to ~38 bpm/in²) and improved perforation efficiency from 67% to 76%.
Adaptive proppant ramps: Across a pad dataset, perforation efficiency declined as proppant concentration increased, with stable performance above 90% up to ~0.8 ppg and wider variability once concentrations exceeded ~1.0 ppg. Live monitoring enables an immediate response: pause or smooth a steep ramp when efficiency collapses, then resume once distribution stabilizes.
Volume-neutral reallocation: When UI falls below a defined threshold, continuing to pump can concentrate proppant into runaway fractures. With live monitoring, low-performing stages can be curtailed intentionally, with remaining slurry redeployed to later stages that demonstrate stronger distribution. Total fluid and proppant for the well are unchanged, and capital is deployed more effectively across the lateral.
