Highly Detailed Reservoir Imaging by Using Sparse Layer Inversion in a Complex N.Sea Turbidite

Abstract

SUMMARYSparse layer inversion (SLI), like sparse spike inversion (SSI), invokes a sparse reflectivity solution for the reconstruction of noisy seismic traces in the presence of a known, band-limited wavelet. However, solving for layers, i.e. dipoles, rather than individual interfaces, holds the potential for achieving increased detail and lateral stability over that usually achieved with SSI. In this paper, we present the method and show the application of SLI to a complex turbidite reservoir in the UK Central North Sea. IntroductionWhen inverting the seismic convolution model to estimate broadband reflectivity, a sparse representation of the number of interfaces is a common choice. Sparseness leads to increased temporal detail and sharp interface definition. However, given the performance of such sparse spike inversion (SSI) methods with regard to noise and layer continuity, an alternative means for achieving bandwidth extension is worth considering. In this paper we discuss sparse layer inversion (SLI) (Zhang and Castagna, 2011;Zhang et al., 2013), which aims to resolve subtle seismic events below tuning thickness. This method was applied to seismic data from a Paleocene discovery in the UK Central North Sea, focusing on the complex thin bed interference pattern affecting the Mey sandstone reservoir. The objective was to 'deblur' the seismic image sufficiently to enable improved stratigraphic interpretation. An additional challenge was the variable frequency-dependent amplitude loss over the centre of the structure due to overburden gas. This was addressed prior to SLI. Results were compared to a conventional SSI reflectivity inversion. In addition, the SLI results were tested for their suitability as input to acoustic impedance inversion.