Shallow Geo-hazard Detection: Implications on Field Development

Abstract This paper outlines a novel method to improve detection rates of potential shallow formation drilling geo-hazards by identifying geologic anomalies using 3D seismic data. The resulting identification of these anomalies can be utilized to reduce drilling costs and associated lost time from lost circulation zones. The results are compared to the previously used shallow geo-hazard detection method, which was part of a de-risking strategy. The drilling geo-hazards we aim to identify, comprises of highly porous carbonates in the form of caverns and sinkholes, also known as karsts. Identifying these geo-hazards from seismic data was accomplished by first flattening the seismic volume using our novel flattening algorithm, which requires no horizon interpretation on which to flatten. From that flattened seismic volume, we then extract an in-house developed edge detection attribute with a Sobel filter, which was capable of outlining the geologic anomalies and discontinuities in the data associated with those geo-hazards. The results from using this method showed a 10% improvement in detecting these geologic anomalies and in greater detail, compared to the previous approach, once correlated with legacy wells that have encountered lost circulation from drilling through these anomalies. Our new method applies a novel flattening algorithm to the input 3D seismic volume. This algorithm is superior to traditional horizon flattening methods, not only because it requires no horizon interpretation, but it also employs the vectorized disorder algorithm (Fu & Al Dossary, 2021) that further enhances the flattening results. The extraction of the edge detection attribute with an enhanced Sobel-based filter displayed a significant improvement over the previously used variance attribute. The variance attribute outputs the variance of the values within a radius from a center point with an optional smoothing operator, whereas the Sobel-based method works with trace correlations instead of values, allowing it to detect only the geological discontinuities of the input seismic data. This new method is a quick and reliable technique that can be used during field appraisal and development well proposals, to identify and avoid occurrences of drilling losses from geo-hazards. The novelty lies in this method’s ability to bypass traditional seismic horizon interpretation, a step that has historically introduced potential errors into geo-hazard detection, and provides a more consistent approach in doing so. The integration of the enhanced edge detection attribute, coupled with the sophisticated and novel flattening algorithm, represents a significant advancement in seismic analysis, offering a novel and robust tool for cutting drilling costs and improving efficiency.