Wellbore Position Errors Caused by Drilling Fluid Contamination

Abstract

Abstract Recent improvements in magnetic survey quality control have revealed the presence of systematic azimuth anomalies, affecting the calculated trajectories of a significant number of wells. Although these anomalies display some features commonly associated with drillstring magnetic interference, they cannot be removed by traditional magnetic correction techniques. New magnetic survey quality control procedures have identified the problem as an attenuation of the transverse component of the earth's field, similar to that caused by magnetically permeable material surrounding the magnetometers. The drill collars housing the sensors were inspected but showed no magnetic susceptibility, which led to suspicion that the problems were caused by magnetic particles suspended in the drilling fluids. Laboratory testing has confirmed the magnetic properties of these fluids, and chemical analysis has verified the presence of iron in significant quantities. Field examples demonstrate the use of a multiple-survey quality control and correction method for identification and removal of these survey anomalies. Introduction Boreholes are commonly surveyed using the earth's magnetic field as a north reference. Survey tools measure three orthogonal components of the local gravity and magnetic fields, and from these data the tool's inclination, azimuth, and toolface orientation can be calculated. It is also possible to calculate the magnitudes of the gravity and magnetic field vectors, and the angle between them, which is the complement of the local magnetic dip angle. These last three parameters may be used for survey quality control purposes. Magnetic interference. Unexpected measured values for the magnitude of the magnetic field vector and/or its dip angle provide an indication that the magnetic azimuth might be out of specification. A common cause of erroneous azimuth is magnetic interference. The earth's magnetic field in the vicinity of the tool's magnetometers can be distorted by magnetized material. Possible sources include: the formation, the casing of adjacent wells, the drilling fluid, and steel drillstring components. The magnetometers are housed within a non-magnetic drill collar, and additional non-magnetic collars can be used to space the sensors longitudinally from the steel components of the drillstring. However, in most circumstances the non-magnetic spacing will be insufficient to isolate the magetometers completely from the steel drillstring components. Magnetized formations and drilling fluids are considered rare, and adjacent wells are avoided, so the usual presumption is that magnetic interference originates in the magnetic drillstring components and is oriented along the axis of the drillstring. Accordingly, a number of correction techniques have been developed to remedy such problems1–4. Single-axis corrections. The technique normally used to correct for magnetic interference assumes that only the axial (z-axis) magnetometer measurement is corrupted by drillstring interference. The local earth field components are determined independently, such as from a geomagnetic model or chart, or from a magnetic site survey. It is then reasonable to assume that the most likely value of z-magnetometer interference is that which results in the minimum vector distance between the post-correction total field components and the reference field. Magnetic corrections of this type1 have been made for the last two decades. Because the correction picks the point where the data best fit the reference total field and dip angle, the residual errors in total field and dip are generally expected to be smaller than for raw data5. Multiple-survey corrections. More recently, multiple-survey magnetic corrections have been developed by several companies3,4. These techniques process data simultaneously from a number of surveys, assuming that all of the surveys are affected by identical disturbances. They can then determine effective bias and scale factor error terms on all three axes, which minimize the overall variance between the corrected measurements and the expected reference field.