Surveying Techniques With a Solid-State Magnetic Multishot Device

Abstract

Summary. An electronic magnetic multishot instrument was introduced in 1985 to overcome accuracy and reliability limitations associated with conventional photomechanical multishot systems. This paper describes the survey system and the instrument's performance capabilities. Our aim is to develop a tool-performance model for general application to the management of surveying operations. Accuracy is considered in two part sensor errors are considered in detail, and external effects on accuracy e.g., axial misalignment, bottom-hole-assembly (BHA) deflection, geo-magnetic influences, and drillstring-induced interferenceare analyzed and quantified. We found that geomagnetic influences and drillstring-induced interference dominate the ultimate performance of all magnetic tools, particularly this electronic multishot instrument. This paper also describes "in-hole referencing," a method of eliminating these errors that involves surveying the first part of an openhole section with a high-accuracy gyroscope and aligning data derived from subsequent magnetic surveys with it. Using this method, we can achieve gyroscope-survey levels of accuracy with magnetic instruments in highly deviated and extended-reach wells without incurring the costs and penalties of running wireline surveys. Introduction When magnetic multishot surveys are taken with conventional photo-mechanical tools, it is standard practice to run two photo-mechanical tools, it is standard practice to run two consecutive instrument stacks to provide a backup in case of misruns and to supply additional information to crosscheck instrument performance. Special quality-control techniques were evolved for analyzing the results of conventional photomechanical surveys run in this way. However, surveys photomechanical surveys run in this way. However, surveys done with these tools exhibit high incidences of erratic data or "sticky compass," mechanical failure of the battery stack or film feed, and unacceptable divergence (greater than 5 × 10(-3) of depth) between stacks. These problems caused losses of rig time and accuracies unacceptable for the majority of deviated wells being drilled. To achieve the required accuracy, the cost and time penalties associated with running gyroscopic surveys at total depth (TD) of the wells had to be accepted. Clearly, an economic incentive existed to develop magnetic instrumentation with improved reliability and performance and surveying procedures to achieve the desired accuracy at minimum cost. In 1980, the British Natl. Oil Corp. encouraged various survey contractors to develop a new generation of magnetic surveying instrumentation based on widely available proven steering-tool technology. This paper describes how one such device was field tested and outlines the mathematical techniques used to model it and to analyze the field data. Development and Testing NL Sperry-Sun developed a solid-state, self-contained directional surveying device that measures tool attitude in relation to the Earth's magnetic and gravitational fields. The gravity-sensor package consists of three orthogonal accelerometers: z along the tool axis, × perpendicular to the z axis and in line with the T-slot at the foot perpendicular to the z axis and in line with the T-slot at the foot of the tool, and y perpendicular to the × and z axes. The magnetic-sensor package consists of three orthogonal flux-gate magnetometers parallel to the accelerometers' axes. The data stored consist of accelerometer output (Gx, Gy, and Gz), magnetometer output (Bx, By, and Bz), probe temperature, and calibration characteristics. The instrument becomes software-configurable by downloading from an IBM-compatible personal computer (PC) to optimize performance for specific applicationse.g., multishot, single-shot, performance for specific applicationse.g., multishot, single-shot, or core-orientation work. Preprogramming of the probe allows shots to be recorded at intervals of between 10 seconds and 1 hour, and the start of recording can be delayed for up to 36 hours after the tool is initialized. There is also a facility to power down between shots when the shot interval is large to conserve battery power. The analog outputs from the accelerometer and magnetometer sensors are converted into digital form before being stored in the data registers. At the end of the survey, the registers can be transferred to an IBM-compatible PC or HP85B computer. Dedicated rig-site software then allows cross matching of those data against surface recorded times and measured depths and such further processing as misalignment and interference corrections, which are processing as misalignment and interference corrections, which are described later. The basic probe is 54 in. [1370 mm] long. When made up inside a pressure casing, a tandem instrument stack is typically 12 ft [3.65 m] long. Fig. 1 shows a schematic of the downhole tool and a typical rig-site configuration. An early version of the survey tool was field tested in a North Sea deviated well in mid-1985. Results indicated that the solid-state electronic multishot device exhibited higher levels of mechanical reliability and data quality than conventional photomechanical systems. The trials provided the necessary justification for routine use of the instrument. The tool's versatility was demonstrated when it was applied to core-orientation operations. Performance Modeling Performance Modeling The sensors in a solid-state survey tool produce output in electrical voltages. These raw values are corrected for bias, scale factor, and temperature effects and are then converted by means of internal calibration data into readings of gravitational acceleration and magnetic field strength. These data, in turn, are converted to borehole inclination, azimuth, and tool face with the output equations given in Appendix A of Ref. 3. When survey instruments are used, the ability to quantify tool performance is as important as the ability to calculate borehole performance is as important as the ability to calculate borehole direction from the results. Terms for sensor error, derived in Ref. 3. can be incorporated into a generalized instrument performance model, such as that proposed in Appendix A of Ref. 5. The complete model must include terms for such external effects as magnetic interference. In this form, it can be used to describe the performance of the whole survey system, and the significance of performance of the whole survey system, and the significance of individual error sources can be assessed. Table 1 lists typical values for the system errors that were derived from tests on a number of survey tools and postanalysis of data obtained from surveys taken during routine operations. Effects of Instrument Rotation Part of the original field-test program involved recording an Part of the original field-test program involved recording an eight-station rotation shot on bottom at the beginning of each survey. Postanalysis of the results showed that the inclination, azimuth, and Postanalysis of the results showed that the inclination, azimuth, and absolute value of measured gravity and magnetic field strengths varied with axial rotation of the tool. One factor that caused the variation in measured field values was traced to the residual biases in each individual sensor's output. We found that the magnitude of these biases could be obtained from the pre- and postsurvey calibration data. By including appropriate terms in the equations for the misalignment correction, we can extract bias values from rotation-shot results taken in the field. SPEDE P. 209