Optimization of Coalbed Methane Multi-lateral Drilling in the San Juan Basin Via Wellbore Stability Modeling and Data Analytics

摘要

Remaining in target and ensuring wellbore stability are among the many challenges of horizontal drilling. We introduce an integrated borehole geomechanics and data analytics workflow for improved drilling of horizontal wells. A Normalized Rate of Penetration (NROP) is generated by deconvolving ROP measurements from drilling input parameters to yield a real-time rock strength indicator. The integrated workflow involves a two-step process to assess mud weight requirements while drilling. First, real-time NROP is transformed into apparent rock strength (defined herein by apparent Uniaxial Compressive Strength, UCS). Second, a training data set, built via a commercial geomechanics software, is invoked to translate apparent UCS into mud weight. The practicality and reliability of this approach is validated with a case study utilizing a dataset from the Fruitland Coalbed Methane (CBM) in the San Juan Basin, although it is also applicable to other drilling targets. The dominant rock types contacted during CBM multi-lateral drilling are coal, shale, sandstone, and bentonitic volcanic ash. For the given lithological rock strength, the corresponding required apparent mud weight is accurately predicted from an analytical UCS-NROP exponential function and subsequent machine learning regression. Findings from this work support horizontal drilling operations in two ways, 1) by qualitatively informing real-time geo-steering decisions to reduce out of target drilling and 2) by quantitatively informing mud weight requirements in real-time to avoid wellbore instability issues. Viability of applying this methodology to other horizontal plays and basins is currently being progressed.