Achieving Zero Stuck Pipe Incidents in Horizontal Wells Through AI-Based Real-Time Risk Mitigation and ROP Optimization: A Story From BHA Graveyard in Offshore East Java Indonesia (SPE/IADC Asia Pacific Drilling 2026)

A field-validated case study with PETRONAS Carigali Ketapang II: two horizontal extended-reach infill wells in the JT field, offshore East Java — where half of the previous sixteen wells had experienced significant stuck pipe incidents — were delivered with zero stuck pipe events, 13 days ahead of the planned combined duration.

SPE/IADC-233133-MS — SPE/IADC Asia Pacific Drilling Technology Conference and Exhibition, Bali, Indonesia, August 5–6, 2026

Y. F. Yusicha, M. I. Akbar, J. Tobing, F. Ardhiansyah, T. Setiawan (Petronas Carigali Ketapang II Ltd); H. P. Setiono (SKK MIGAS); N. Hanafiah, M. Regan (Exebenus)

DOI: 10.2118/233133-MS
‍

Abstract

PETRONAS has previously drilled 16 high-inclination and horizontal extended-reach development (ERD) wells in the JT field, offshore East Java, between 2015 and 2023. Historically, 8 of these wells representing a 50% risk probability, experienced significant stuck pipe incidents, which were driven by a combination of complex heterogeneous formations, variable formation pressure regimes inducing differential sticking tendencies, partial to severe fluid losses, and wellbore instability associated with planes of weakness when drilling at inclinations exceeding 60 degrees.

In 2025, a two-well horizontal infill drilling campaign was completed in the same field. The first well built inclination from 40° to 61° through a 12¼″×14¾″ under-reamed section and built from 61° to 74° through 10-5/8″ × 12¼″ under-reamed section before landing a 3,200 ft RSS-geosteered 8½″ horizontal lateral. The second well extended a 4,800 ft tangent at 68° inclination before landing a 1,500 ft 8½″ RSS-geosteered horizontal lateral. Both well profiles presented significant drilling challenges amplified by the demanding geological environment, where conventional real-time monitoring, inherently reactive and reliant on manual surface parameter interpretation, was insufficient to pre-empt emerging risks in a timely manner.

This paper describes the successful field application of an integrated, real-time AI-driven advisory solution for predicting and identifying stuck pipe risk and supporting ROP optimization along with vibration monitoring, deployed without any requirement for additional model training from offset well data or prior well-specific calibration. The collective solution, comprising multiple interoperating real-time machine learning agents, facilitated flawless execution of both challenging infill wells, achieving zero stuck pipe incidents and completing operations 13 days ahead of the planned combined duration. Optimal drilling efficiency was sustained over key intervals, with overall average ROP improvement of 34% and 11% recorded for the first and second wells respectively over constrained sections. The substantial material savings realized in time, logistics, consumables, HSE exposure, emissions, and cost demonstrate the significant value of real-time, predictive AI advisory in complex horizontal well campaigns.