Title: Diffraction imaging: towards clearer images of hydrocarbon reservoirs
Researchers: Brydon Lowney, Dr Ivan Lokmer
When evaluating hydrocarbon exploration prospects, it is essential to understand the network of faults and fractures in the subsurface, in order to better quantify potential leakage and compartmentalisation risk and make an informed decision about the optimal well placement. Hence high-resolution seismic imaging is of great importance to de-risking exploration prospects. However, the standard seismic reflection imaging has its limitations, one of which is relatively low resolution, leading to a failure to identify small subsurface features. A great potential for improving the resolution of subsurface images lies in so-called Diffraction Imaging technique. Coming to an obstacle or irregularity whose size is of the order of the wavelength, seismic waves experience diffraction, the process which creates recognisable hyperbolic patterns on seismic images. These diffraction patterns are used to precisely image very small subsurface features, such as fault edges, pinch-outs, channel edges, or inter-channel features, commonly invisible in standard seismic reflection images. Although the theory of diffraction imaging is well established (e.g., used in nanotechnology), its use for seismic imaging is novel and its benefits and pitfalls are yet to be systematically investigated.
The aim of the project is (i) to systematically address questions about the benefits and drawbacks of the diffraction imaging technique, (ii) to assess the real value of the diffraction imaging for the hydrocarbon exploration and development, and (iii) to apply the technique to the Irish off-shore data. The first two points above have been addressed using synthetic datasets, where the underlying geological models are known The project has started with studying simple synthetic 2D models with isolated features of interest, such as a fault zone or a pinch-out, and later extended to the complex 3D synthetic SEAM dataset (managed by Tullow Oil PLC).The practical limitations and strengths of the method can then be robustly evaluated and the learnings transferred to real data leading to a more complete sub-surface understanding.