High integrity wellbore surveys: methods for eliminating gross errors
An earlier paper by the same authors, SPE 1037341, pointed out the potential safety and commercial costs of unreliable directional survey data. It described how a significant degree of reliability can be achieved with the application of quality control checks internal to the directional data, but it also identified the fact that such checks fall short of providing comprehensive reliability assurance.
This paper documents weaknesses in conventional directional survey quality control (QC) procedures through theoretical considerations, statistical analyses of real survey data, and real examples of failed surveys that have made it through conventional QC procedures without detection. The paper defines principles for survey programme design and implementation to eliminate these weaknesses. It proposes a new set of minimum requirements for survey validation, which, in general, incorporate the need for an overlapping verification survey or other independent observation. This requirement is not generally acknowledged in most of today’s directional surveying QC procedures, which may be too weak to ensure the validity of a given error model and thus the integrity of the final well survey.
This paper also discusses how analysis of a sufficiently large quantity of data, with the recommended QC methods applied, allows validation and refinement of the error model.
This paper is the product of a collaborative work in the SPE Wellbore Positioning Technical Section (WPTS).
One important conclusion of SPE 1037341 was that, in addition to the commonly applied internal QC checks, “Supplementary data and additional quality tests are required for high integrity wellbore positions??. This paper demonstrates that, although it cannot identify all potential errors, a second independent survey is the most powerful external data QC check. The desirability of independent redundant surveys has been recognized within the Industry for many years, was acknowledged in print certainly by 19892 and adopted as standard practice by some operating companies. However, not all operating companies see such practices as being cost effective, taking the view that two competing survey positions cause confusion rather than increased reliability. This view might be justified if the performance of neither system is well defined or the data are not subjected to valid internal QC tests. However, companies who advocate survey redundancy generally recognise the need to validate the individual surveys.
It is more difficult and expensive to acquire independent redundant surveys in extended reach and horizontal wells, and even those operators who advocate the policy have looked for more cost effective alternatives. Their answer has generally been multi-station analysis of the Measurement While Drilling (MWD) surveys3,4, with or without additional support from in-field referencing of the magnetic reference field (IFR). However, although it provides some benefits not available from any other QC method, it will be shown that multi-station analysis does not achieve as high a level of overall reliability as can be achieved through comparison with a fully independent redundant survey, not even when integrated with IFR. Proponents of multi-station analysis recognise this and usually insist on independent supervision of the overall survey quality assurance (QA) process as a means of increasing reliability. The independent cross checking of all aspects of data acquisition and processing is certainly good practice and aids identification of gross errors, but it does not provide the numerical quantification of agreement between data and error model prediction that independent redundant surveys do. The highest level of reliability is achieved by the application of all three QC techniques; internal checks, external checks and independent cross checking.