Sea Technology

SEP 2017

The industry's recognized authority for design, engineering and application of equipment and services in the global ocean community

Issue link:

Contents of this Issue


Page 12 of 68

12 st / September 2017 Critical in the successful ap- plication of any seismic inversion procedure is the presence and influence of a skilled, multidisci- plinary integrated study team (IST), whose job is to quality control both the input data being fed into the inversion as well as the out- put physical property models. As with all such processes, seismic inversion is highly sensitive to the quality and spatial density of the input data, and, while smarter op- timization schemes can be used to accommodate some inaccuracies (particularly random noise con- tamination), basic geological, geo- technical and geophysical qual- ity control is imperative. Similarly, the inversion process will always produce an output. While there are automated (often purely statistical) methods for assessing the quality of this output, the truest test of the output is whether they make geological or geotechnical sense. This is still, and will be for the foreseeable future, best assessed by a highly skilled, integrated team of ge- ologists, geotechnical engineers and geophysicists. Why Do We Need Seismic Inversion? Modern site investigations for off- shore infrastructure projects take the form of a multidisciplinary, staged workflow. Over recent years, marine site investigation areas have become larger and more complex as they have moved into deepwater and more ex- treme environments, while simultane- ously tackling areas of more complex geology. The knock-on effects of this evolution are many, but we highlight two particular issues where seismic in- version is perhaps most relevant: Integration vs. Efficiency. With large projects, there is a reasonable temptation to increase efficiency by running the different phases of site investigation data acquisition (geo- physical, geotechnical and geological sampling, and cone penetrometer, or CPT, testing) in parallel, rather than in series. Even for small areas, moving into more environmentally challeng- ing regions (such as Alaska) where ad- verse weather compresses the survey window, such an operational decision is not unwarranted. However, moving from a series to parallel acquisition ap- proach reduces the opportunity for lat- er sampling strategies to be informed by earlier results, particularly with regard to the placing of shallow core, rotary borehole and CPT locations. This can lead to a poor understanding of lateral variations within individual soil units and/or entire soil units being missed by the simultaneous sampling program. Unfortunately, this can often lead to the need for a second sampling campaign to calibrate the geological model ahead of final location-specific engineering design borehole acquisi- tion, thus introducing expensive, un- planned delays and eliminating the expected efficiency. Example seismic inversion results from the eastern Irish Sea, U.K. Panels show the original seismic reflection data (a), inverted acoustic impedance data (b), and pseudo (predicted)-CPT data q t (c) predicted from the seismic inversion results. Panel (d) com- pares the predicted (black) against the measured q t (red) at the CPT location, together with errors (shaded gray region) associated with the inversion results.

Articles in this issue

Links on this page

Archives of this issue

view archives of Sea Technology - SEP 2017