[First Hit]

International Conference & Exhibition

Datapages, Inc.Print this page

Survey Design Optimization by Data Simulation


Objective The objective of the study is to assess the most suitable acquisition geometry for the forthcoming exploration surveys of KOC which includes the deep Jurassic targets. Procedure Kuwait Oil Company (KOC) conducted a field acquisition and data processing test to evaluate the benefits of wide azimuth acquisition in northern Kuwait. Single sensor acquisition and processing was used to record a full azimuth data set with 6km inline and crossline offsets. From the full offset (6000m) full azimuth data, several decimated subsets were processed to simulate previously used narrow azimuth acquisition geometry (1400m) as well as limited offset (4000m) full azimuth geometry and finally full offset (6000m) full azimuth with sparser source and receiver line spacing (400m). Following are the original and the simulated geometries: P1 (Original Geometry): Line spacing=200m, Max. Inline Offset=Max. X-line Offset=6000m P2 (Simulated Geometry): Line spacing=200m, Max. Inline Offset=6000m, Max. X-line Offset=1400m P3 (Simulated Geometry): Line spacing=200m, Max. Inline Offset=Max. X-line Offset=4000m P4 (Simulated Geometry): Line spacing=400m, Max. Inline Offset=Max. X-line Offset=6000m The data were processed to maintain independence between the data after the decimation stage. Care was taken to apply surface consistent processes to the decimated data. Results The unmigrated stack for the full offset square patch geometry shows S/N improvement from the high fold but the Previous HiteventNext Hit continuity is not dramatically different then the narrow azimuth (P2) data or restricted offset/full azimuth (P3) data. The pre-migration stack for the full azimuth data (P1) is lower in vertical resolution then the P2 and P3 data sets due to far offset stretch & poor long offset moveout correction. On unmigrated data, a higher (vertical) resolution stack could result from a tighter stack mute or by applying a residual moveout correction and possibly stretch compensation. The Previous HiteffectNext Hit of the P1 geometry on multiple attenuation seems marginal at best when compared to the narrow azimuth geometry (P2) on the unmigrated stack data. Conclusions The observations from the data analysis in terms of resolution, bandwidth, Previous HitsignalNext Hit to Previous HitnoiseNext Hit Previous HitratioNext Hit, Previous HiteventTop continuity, and multiple content provided guidance for future survey designs.