Abstract
Directional surveys acquired by Measurement While Drilling (MWD) are subject to many errors that are not easily recognized by traditional Quality Control (QC) procedures. This commonly leads to inaccurate wellbore placement and greater positional uncertainty. Common sources of MWD survey error are inaccurate geomagnetic references, localized distortions in the natural magnetic field, poor instrument calibration, random sensor noise, magnetic mud, and human error. Often times, such errors go unrecognized due to limitations in traditional single-station QC tests. This is a significant problem because wellbore collision avoidance, geological modeling, and reservoir drainage are all greatly affected by wellbore placement accuracy. Fortunately, most sources of MWD error can now easily be identified and corrected through implementation of robust independent survey quality control processes. By using web-based systems to facilitate this process, drillers can benefit from the most powerful quality assurance practices which can be standardized across the industry regardless of service provider or vendor specific technologies.
Well placement by MWD employs the use of orthogonally positioned accelerometers and magnetometers to measure the orientation of the bottomhole assembly (BHA) relative to the Earth’s gravitational and magnetic fields as shown in Figure 1. Taking survey measurements at regular intervals along the well path enables computation of the wellbore trajectory through minimum curvature interpolation.
Standard MWD surveying is subject to numerous error sources which can lead to inaccurate wellbore placement. These sources of error are divided into three categories: gross, random, and systematic. Gross errors occur from human mistakes, instrument failure, or environmental factors that cannot be predicted or estimated. Random and systematic errors occur with some measure of predictability and can therefore be estimated and quantified. The standard approach for estimating positional uncertainty in the wellbore caused by random and systematic survey error is to use instrument performance models called tool codes. Tool codes provide the mathematical framework to compute Ellipsoids of Uncertainty (EOUs) which represent positional uncertainty evaluated at a particular sigma, or confidence level (Grindrod 2016). Figure 2 shows how EOUs form an elliptical tunnel when propagated along the well path which characterizes the statistical distribution of where the actual wellbore could exist. Quantifying positional uncertainty is a critical step in the well planning and drilling processes because it enables drillers to evaluate collision risk and understand wellbore placement.
It is important to note that the MWD tool code used for EOU and anti-collision calculations specifies the permissible magnitudes of the various error terms. Another assumption also made is that surveys are free of gross error, since gross error cannot be predicted or modeled (Torkildsen 1997). To validate EOUs and anti collision scans, it is therefore essential to quality control MWD survey measurements to verify that they are free of gross error and do not contain excessive random or systematic error. If the quality control step is not performed, then there can be very little confidence that the tool code is representative of the actual errors in the wellbore position.
There are three values computed from MWD survey measurements which can be used for quality control purposes. They are B total (strength of the magnetic field), Dip (direction of magnetic field with respect to horizontal plane), and G total (strength of the gravity field) (Ekseth 2010). These measurements are used as metrics for survey quality, because regardless of the orientation of the wellbore and BHA, the measured B total, Dip, and G total should be equal to the values provided by the geomagnetic and gravity reference models. Therefore, any differences between the measured values vs reference values (Δ B total, Δ Dip, and Δ G total) can be attributed to some combination of measurement error and reference error. This concept is the basis for standard single-station MWD survey quality control tests.
It is common in standard MWD surveying practice to rely on these single-station tests as the only metric for survey quality assurance. However, these tests are considerably lacking in their ability to fully validate the assumptions made by the tool code. For instance, typical QC tolerances used by MWD contractors for passing or failing surveys are often arbitrary limits based on legacy practices. A more informative and standardized approach would be to use QC tolerances that are derived from the same tool code used to compute the EOUs. Furthermore, it is not enough to evaluate each survey individually because single-station QC tests are extremely limited in their ability to distinguish different types of error. It is preferable to evaluate single survey points against the entire survey data set in order to identify trends that could indicate what types of errors are occurring and to gain a better understanding of how the various errors may actually impact the wellbore position. Finally, single-station QC tests are not capable of detecting certain types of gross human errors such as applying an incorrect north reference or misreporting the final survey measurement. This makes it critical to independently calculate survey inclination and azimuth from the raw sensor measurements and to independently compute reference values to verify against human mistakes that would otherwise go unnoticed.
