Case Study: Automated Stringer Detection Boosts Drilling Efficiency With Less Downtime

Abstract

_ Horizontal drilling has evolved over decades to become a trusted well construction technique in mature oil-producing regions. However, many challenges remain in geologically complex drilling environments. One challenging drilling scenario arises when the drill bit encounters hard stringers interbedded within softer formations like sandstone. The sudden change in rock mechanics in moving from the formation to the stringer often results in deflection of the drill bit and a deviation in the well path known as a dogleg. The rapid directional changes caused by doglegs lower drilling efficiency by moving the drilling assembly off its planned path, and sometimes, out of the desired reservoir zone. The static loads generated in a dogleg can damage the bottomhole assembly (BHA) and lead to premature bit wear, shortened run life, higher maintenance costs, and more trips downhole. The curved path also increases the risk of stuck casing or completions strings, which can prevent the well from being completed to planned target depth (TD). An operator of a mature field in the North Sea faced such challenges while drilling multilateral wells through reservoir sections containing hard calcite stringers interspersed in sand formations with low unconfined compressive strength. The operator aimed to keep its lifting costs low by increasing the gross rate of penetration (ROP) through these sections. But the doglegs created at each stringer often required pulling back to perform a costly, time-consuming reaming operation to smooth out the well path and allow drilling to continue to TD. On average, the operator amassed 2.7 hours of invisible lost time (ILT) on reaming for every 3,300 ft (1000 m) drilled and achieved significantly lower gross ROP—both of which increased the operator’s per-well lifting costs above plan. Working with the operator, Baker Hughes developed an automated stringer detection service that would identify stringers earlier than previously possible to minimize local doglegs, reduce reaming time, and efficiently deliver wells to plan. Developing an Early-Warning Stringer Detection Service With operator input, the service provider developed an automated service that provided reliable, consistent, and early stringer detection to guide fast corrective action. The service includes an automated stringer-detection module embedded in an advanced measurement-while-drilling (MWD) sensor sub. The sub contains several dynamic sensors that collect a range of vibration and load measurements. The detection module uses a physics-based algorithm that combines two MWD measurements—tangential acceleration and dynamic torque at the bit—to calculate a value for high-frequency torsional oscillation (HFTO) in the BHA. HFTOs are torsional vibrations with frequencies between 50 Hz and 450 Hz that only occur during bit-rock interactions in hard formations. The distinct change in HFTO response during the transition from soft formations to hard formations makes it a leading indicator for detecting stringers in real time during drilling.