Wellbore Breakout Analysis in the Southeast Moran Field , Papua New Guinea

Condensed from SPE 88607, presented by A. Khaskar, GeoMechanics International, A.H. Warrington, Oil Search Ltd, M.E. Magee, GeoMechanics International, K.L. Burgdorff, GeoMechanics International, and D.A. Castillo, GeoMechanics International at the SPE Asia Pacific Oil and Gas Conference and Exhibition held in Perth , Australia , 18-20 October 2004.

Abstract

Abnormally high pore pressure regimes and severe wellbore instabilities in the Ieru Formation and its underlying shales in the Moran Field in the actively deforming Papua New Guinea (PNG) Fold Belt region cause long non-productive times. Many of the Moran wells encountered severe wellbore instabilities that resulted in stuck pipe, excessive cuttings, pack-off, lost bottom hole assemblies, and well control problems. Mechanical sidetracks were often required to reach target. Mechanical rock failure along the borehole wall induced by inappropriate mud weights was often the primary factor that led to these drilling problems and cost overruns.

To characterize the physical state of the reservoir and over burden formations and constrain the pore pressure and stress coupling for the Southeast Moran area, we built a geomechanical model based on the analysis of wireline logs, downhole measurements, and drilling experiences gathered from Moran-3X and two sidetrack wells. We then used this model to design optimal mud weights for maintaining wellbore stability in the proposed SE Moran-1 well. After the SE Moran-1X well was drilled and logged, our analysis confirmed the accuracy of the geomechanical model by reproducing borehole conditions as seen in image logs.

Introduction

The geomechanical model consists of the magnitude and orientation of the three principal stresses, the pore pressure, and the uniaxial compressive rock strength. We reviewed existing work and regional data, and then analyzed data acquired in existing wells to characterize and refine the state of stress specific to the Moran Field. Our analysis of available electrical borehole image data collected in Moran-3X and its two sidetrack wells included mapping stress-induced wellbore breakouts and drilling-induced tensile fractures. In addition, we estimated the least principal stress values from leak-off tests, calculated the vertical stress from density logs, obtained pore pressure information from direct measurements, and estimated rock strength from empirical log-derived relationships. The maximum horizontal stress is constrained by modeling the stress and pressure conditions with the rock strengths that are consistent with the observed wellbore failures. To validate the geomechanical model we compared predictions of wellbore breakout development to breakouts observed in image logs and from the drilling experience in the Moran-3X sidetrack wells. We then determined safe mud weights for the proposed SE Moran-1X well using this geomechanical model.

Stress-induced wellbore breakouts occur when the compressive stress concentration around the borehole wall exceeds the rock strength. The presence, orientation, and severity of failure are a function of the in situ stress field, the wellbore orientation, and the rock strength (Refs. 1-2). In a vertical well in a region where overburden stress is a principal stress, breakouts may form on opposite sides of the wellbore at the azimuth of the minimum horizontal far-field compression, as this is where the compressive hoop stress is greatest. If a well is inclined to the principal stresses, the location of the breakouts is a complex function of the orientation of the wellbore and the orientations and magnitudes of the in situ stresses (Refs. 3-4). Drilling-induced tensile fractures occur in the borehole wall where the circumferential hoop stress is negative and exceeds the tensile strength of the rock. These fine-scale features occur only in the wall of the borehole, due to the localized stress concentration, and do not propagate away from the hole. The fractures form either parallel to the borehole axis in vertical wells, or in an en echelon pattern that is inclined with respect to the borehole axis in deviated wellbores (Refs. 3-4). Tensile fractures will remain axial in a deviated wellbore if the well is drilled approximately parallel to the maximum horizontal stress.

Analysis

From the analysis of over 2000 meters of good quality Formation Micro-Imager (FMI) data acquired in the two Moran-3X sidetracks, we found pervasive breakouts, primarily in the shale-rich lithologies, with a generally consistent breakout orientation. The average log-derived rock strength in the shale and siltstone intervals varies from 3200 psi to 5600 psi. The average log-derived rock strength in reservoir sandstone intervals varies from 9,300 psi to 19700 psi. The lower rock strength for the shale intervals is consistent with the abundant borehole breakouts observed in FMI data in the shale sections, as compared to a generally in-gauge hole condition in the sandstone intervals.

We determined a vertical stress gradient of approximately 20.5 ppg at the reservoir level from over 3500 meters of bulk density (RHOB) log data from one of the Moran-3X sidetracks. The minimum horizontal stress (Shmin) ranges from ~13 ppg in the overlying shales to ~17 ppg at reservoir depths as estimated from leak off test ( LOT ) data. Direct measurements in reservoir sandstone intervals indicate the pore pressure gradient below 2200m TVD ranges between 13.9 and 14.5 ppg.

In the absence of direct pore pressure indicators in the overburden, we used wellbore breakout observations from FMI data and drilling experiences to help constrain a pore pressure profile. The pore pressure appears to be sub-hydrostatic to hydrostatic below the water table (~700m TVD). We proposed a pore pressure gradient that increases gradually from hydrostatic to ~12 ppg between ~1670m and ~2050m MD. We assumed the pore pressure linearly increases from 12 ppg at the base of the FMI interval to 14.5 ppg at the top of the reservoir. This pore pressure profile is consistent with the stress and pressure conditions needed to explain wellbore breakout observations from FMI data and drilling experiences.

We constrained the orientation and magnitude of SHmax through a forward model that uses values for Sv, Pp, Shmin, UCS, rock properties, the width and position of wellbore failure, and the recorded mud weight at the depth of the observed failure. Observations of compressive wellbore failure in the two deviated sidetracks allowed us to determine the azimuth and magnitude of SHmax as a function of compressive rock strength at a given depth, assuming the vertical stress is a principal stress (Ref. 4). Widths and orientations of breakouts detected in FMI data along with accurate knowledge of pore pressure constrained by repeat formation tester (RFT) data in sandstone intervals reduced uncertainty in modeling SHmax magnitude at these depths. The modeling results indicate that the average stress orientation at the Moran-3X well location is N21ºE and the maximum horizontal principal stress magnitude is greater than the vertical stress, which implies a strike-slip faulting stress regime.

Verification of Geomechanical Model

To validate the geomechanical model we compared the predicted wellbore stability based on our geomechanical model against actual breakout observations from FMI data and drilling experiences from the Moran-3X sidetrack wells. We predicted breakout width along the sidetrack well trajectories based on the mud pressures recorded during drilling. When the calculated breakout width is close to or exceeds a critical limit of 90°, failure around effectively half of the borehole wall in a vertical well, drilling problems such as tight hole conditions or stuck pipe due to excess cuttings, usually result. In deviated and horizontal wells, excess cuttings from failure of the wellbore wall may increase the risk associated with packing-off. In our models we linearly decrease the critical breakout width from 90º to 30º with increasing hole deviation from vertical to horizontal to allow for the reduced hole cleaning ability in higher angle wells.

After calibrating the geomechanical model in the overburden using wellbore failure observations and drilling experience in one sidetrack well, we compared predicted breakout widths against FMI data and drilling experiences in the reservoir section using observations in the other sidetrack. The pore pressure gradient is reasonably well constrained by direct measurements (RFT's) and drilling data (e.g. kicks) in this interval. We found that predicted breakout width closely match the breakout widths observed from FMI data over the image log interval and drilling experiences are consistent with predicted breakout widths less than the critical limit.

Refining the Geomechanical Mode l after drilling SE Moran 1-X well

Using the verified geomechanical model and adjusting for the proposed formation tops and reservoir pressures along the vertical SE Moran-1X well trajectory, we determined optimal mud weights, safe mud windows, likely casing designs, and probability of success for critical breakout width limits of 90º. Although we verified the geomechanical model against drilling experiences and image logs from the Moran-3X well and side-tracks, the structural complexities and inherent compartmentalization associated with the PNG area suggested that the model might need refinement using drilling experiences and wireline image logs collected in the SE Moran-1X well. We were able to compare the predicted breakout development with breakouts observed in high-resolution EMI image data collected within the 8.5-inch section of the well.

Leak-off tests in the SE Moran-1X well at 750 and 2065 m MD indicated minimum horizontal stress values that were ~6% greater than predicted. Pore pressure measurements in the Toro Formation also indicated a slightly greater value than originally predicted. The higher pore pressure implies that the amount of over-balance was effectively reduced, therefore enabling wellbore breakouts to widen. The increase in breakout width increased the hole size and reduced the circulating carrying capacity of the mud to remove the extra cuttings. The width of the breakouts in the Alene Formation increased to about 70º, which is very close to the predicted breakout width of ~65º. As a result, tight hole and stuck pipe events were common in the Alene section. As shown in the figure, the narrow wellbore breakouts in the Juha and Toro-B Formations were consistent with the predicted breakout widths based on the geomechanical model. The drilling experiences in the SE Moran-1X well confirmed that the geomechanical model developed for the SE Moran Field is an accurate representation of the stresses operating in this sector of the PNG area.

Conclusion

A rigorous approach to identifying the physical state of the reservoir and overburden formations in the Southeast Moran Field in Papua New Guinea was used to construct a well-constrained geomechanical model for the field consistent with rock strength and drilling conditions. By combining detailed analysis of drilling induced tensile fractures and borehole breakouts observations in borehole image data, direct pore pressure measurements in the reservoir, and rock strength estimates to quantify the complete in situ stress tensor, we were able to constrain the pore pressure and stress coupling for the Southeast Moran Field.

Using detailed observations of wellbore breakouts in the overlying Ieru Formation and log based rock strength, along with mud weight data in the Moran-3X vertical well and its two sidetracks, we determined the magnitude of overbalanced drilling needed to re-create the degree of wellbore failure seen in the image logs collected in the Ieru Formation in these wells. The geomechanical model with the resultant pore pressure profile was consistent with the stress and pressure conditions needed to explain wellbore breakout observations from borehole image data and drilling experiences in each of the three holes.

We evaluated wellbore stability for the Ieru and reservoir sections in SE Moran-1X, an exploration well in Southeast Moran Field, using our resultant geomechanical model. After the SE Moran-1X well was drilled and logged, our analysis confirmed the accuracy of the geomechanical model by reproducing borehole conditions seen in the borehole image log in the 8.5-inch hole section, implying the pore pressure-stress coupling is well understood in shales underlying the Ieru Formation and in the reservoir sections. Image and wireline data in the SE Moran-1X well proved to be crucial in verifying the geomechanical model with respect to drilling performance analysis and reservoir characterization.

References

1. Zoback, M. D., Moos, D., Mastin, L., and Anderson, R. N.: “Wellbore breakouts and in-situ stress,” J. Geophys. Res. (1985) 90 , pp. 5,523–5,530.

2. Moos, D. and Zoback, M. D.: “Utilization of observations of well bore failure to constrain the orientation and magnitude of crustal stresses: Application to continental, Deep Sea Drilling Project and ocean drilling program boreholes,” J. Geophys. Res. (1990) 95 , pp. 9,305–9,325.

3. Brudy, M. and Zoback, M. D.: “Compressive and tensile failure of boreholes arbitrarily-inclined to principal stress axes: application to the KTB boreholes,” Germany , Int. J. Rock Mech. Min. Sci. & Geomech. Abstr. (1993) 30 , pp. 1,035–1,038.

4. Peska, P. and Zoback, M. D.: “Compressive and tensile failure of inclined well bores and determination of in situ stress and rock strength,” J. Geophys. Res. (1995) 100 (7), pp. 12,791–12,811.

Figure: Observed breakout widths (~25 ° ) from EMI image data compared to predicted breakout width (23 ° ) in Juha Formation in plot on the left. Observed breakout widths (~30 ° ) from EMI image data compared to predicted breakout width (27 ° ) in Toro B Formation in plot on the right.