Skip to content

AI-Driven Predictive Modeling for Real-Time Drilling Optimization in Extended-Reach Wells: Enabling First Trend-Setting Mile-A-Day Performance for a UAE Operator

Nov 2026 · ADIPEC · 0 citations · 2 references

Abstract

Extended-reach wells in onshore fields of Abu Dhabi combine long lateral exposure, thin reservoir targets, and dense interbedded formations, making the selection of drilling parameters a continuous compromise between rate of penetration (ROP), trajectory delivery, and downhole tool reliability. This paper presents the field execution of a machine learning (ML)-driven ROP optimization solution and real-time monitoring workflow deployed in advisory mode on two wells, referred to as Well A and Well B. The objective was to quantify on-bottom ROP gains, drilling time reduction, and economic value under live conditions. ML ROP models were trained on high- and low-frequency historical data from well-specific offset groups. Separate models were developed for rotary steerable system (RSS) and positive displacement motor (PDM) bottomhole assemblies (BHAs) and validated before field use during the proof of concept (PoC) phase. During drilling, the system continuously ingested real-time wellsite information transfer standard markup language (WITSML) data, computed optimum drilling parameter setpoints by depth and formation within limits agreed with the drilling team, and issued targeted adjustments. Across the two wells, more than 59 hours of on-bottom drilling time were saved—a 20% reduction relative to the top-quartile (P75) plan—representing more than USD 208,000 in operational savings. Seven of eight sections exceeded the P75 benchmark, and the 6-in. reservoir section of Well B delivered 5,280 ft in 24 hours, a mile-a-day performance in a thin reservoir target and a first for the operator. The results establish condition-based, data-driven parameter management as a validated bridge from advisory optimization toward end-to-end, closed-loop drilling automation.

View source

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.