Building a Bulletproof System for Automatic MWD Survey Processing

Abstract This paper introduces an approach for automatic measurement while drilling (MWD) survey processing that can work fully autonomously and covers the entire workflow: geomagnetic reference calculation, survey data prefiltration, corrections for drillstring interference (DSI), bottom hole assembly (BHA) sag, Stockhausen effect (Stockhausen and Lesso 2003), system health monitoring, survey quality control (QC), and toolcode proposal. The main focus of this study is to formulate principles to realize high reliability of the system, and the reliability of all subalgorithms must be close to 100% to reach the system autonomy requirement. This article highlights four major cornerstones for building such a system: Initialization data validation Static and continuous survey filtration Robustness of the key algorithms QC of output data and system operation. The cornerstones are implemented as follows: Formalization of data schemes and algorithms for internal validation and cross-validation Development of highly robust survey filtration algorithms that are resistant to noise: prefiltration multistation analysis (MSA) algorithm, dogleg severity (DLS) capacity filter, and continuous inclination DLS filter Improvement of accuracy and robustness for MSA, BHA sag, and Stockhausen effect corrections: the calculation of the geomagnetic reference along the trajectory, novel MSA with a priori and steering control information utilization, input smoothing for BHA sag correction, and continuous static survey fusion steady against continuous inclination errors New dynamic acceptance criteria for the total G, total B, and dip angle based on an MWD error model; novel steering acceptance criteria for inclination and azimuth; internal control system for MSA correction, BHA sag correction, and high-definition trajectory calculation; general external control for output trajectory vs. plan; and toolcode assignment based on full covariance analysis Consequently, the automatic system demonstrates complete stability to the noisy data streams of static surveys, continuous inclinations, rotor/slide intervals, capability to work without the no-go zone restriction, correct work within high direction and inclination (D&I) and DSI error ranges, the effective detection of geomagnetic reference issues and system malfunctions, and realistic toolcode assignment. The system significantly reduces human engagement in real-time survey QC and correction. An expert is only involved in case of an issue (D&I module failure, strong external magnetic interference, etc.).