SPE Drilling & Completion
Volume 23, Number 4, December 2008, pp. 394-408

SPE-102781-PA

Wellbore-Stability Study for the SAFOD Borehole Through the San Andreas Fault

View full textPDF ( 7,219 KB )

DOI  More information 10.2118/102781-PA http://dx.doi.org/10.2118/102781-PA

Citation

  • Paul, P. and Zoback, M. 2008. Wellbore-Stability Study for the SAFOD Borehole Through the San Andreas Fault. SPE Drill & Compl23 (4): 394-408. SPE-102781-PA.

Summary

This paper presents a wellbore-stability study of the San Andreas Fault Observatory at Depth (SAFOD) research borehole located near Parkfield, California, USA. In the summer of 2005, the SAFOD borehole was drilled successfully through the active trace of the San Andreas Fault (SAF) in an area characterized by fault creep and frequent microearthquakes. In this study, we report how the analysis of wellbore failures in the upper part of the hole, geophysical logs, and a model for stress gradients in the vicinity of the fault were used to estimate the mud weights required to drill through the fault successfully. Because logging-while-drilling (LWD) acoustic caliper data and real-time hole-volume calculations both showed that relatively little failure occurred while drilling through the SAF, the predicted mud weight was successful in drilling a stable borehole. However, a six-arm caliper log, run after drilling was completed, indicates that there was deterioration of the borehole with time, which appears to be caused by fluid penetration around the borehole. The LWD-resistivity measurements show that essentially no fluid penetration occurred as the hole was being drilled. Because of this, the mud weight used was capable of maintaining a stable wellbore. However, the resistivity data obtained after drilling show appreciable fluid penetration with time, thus negating the effectiveness of the mud weight and leading to time-dependent wellbore failure. Using finite-element modeling (FEM), we show that mud penetration into the fractured medium around the borehole causes failure with time.

Introduction

Drilling perturbs the stress state around a well, and wellbore-stability problems can occur when the near-wellbore stresses substantially exceed the strength of the rock. Excessive instability around a wellbore can be suppressed by choosing an optimally stable borehole orientation and sufficiently high mud weight. Some types of wellbore-wall failure, such as key seating, usually do not cause instability in the borehole but can exacerbate failure in an already unstable borehole.

As described below in the paper, drilling through the SAF in the SAFOD project was conducted in various phases. In this paper, we discuss an analysis of wellbore failures after the first phase of drilling in order to predict the mud weight required to drill successfully through the SAF during the second phase of the project. The main challenges during the SAFOD drilling were the unknown stress field and rock strength along the planned drilling trajectory. Hence, to estimate the mud weight to be used for drilling the SAF zone, we first calibrated a theoretical-stress model (Chéry et al. 2004) for the SAF with the observed borehole failures and minifrac-test data of the first phase of drilling. Then, we estimated and calibrated a uniaxial-compressive-strength (UCS) profile for the rocks encountered by modeling the severity of borehole failures, allowing for a wide range of expected rock strengths because of the possibility of extensive damage to the rocks caused by earthquakes on the SAF.

As will be shown, LWD-caliper data during the second phase of drilling show successful drilling through the SAF using the estimated mud weights inferred from the geomechanical model. However, wireline calipers were enlarged at the top section of the borehole, indicating severe failure with time. We use FEM to show the borehole failure with time. Block theory (Goodman 1989; Goodman and Shi 1985), used for the stability of underground openings, suggests that only the removable blocks of the top section fail because of gravity, but the lower section remains in gauge despite having removable blocks. This explains why the time-dependent enlargement of the SAFOD borehole is restricted to only the top section of the hole.

View full textPDF ( 7,219 KB )

History

  • Original manuscript received: 22 June 2006
  • Meeting paper published: 24 September 2006
  • Revised manuscript received: 20 May 2008
  • Manuscript approved: 15 June 2008
  • Version of record: 10 December 2008