Keele Research Repository

With deep-water basins becoming mature exploration areas, and the number of undrilled structural traps diminishing, stratigraphic traps are growing increasingly important. Onlapping turbidite sands are a common stratigraphic trap type in confined basins; however, onlap architecture can assume a number of different geometrical and facies characteristics as a result of interactions between the turbidity currents and the onlap surface. These variations have important implications for the accumulation and trapping of hydrocarbon reserves, in that the variety affects the quality and recognition of good onlap seal. To investigate stratigraphic traps in turbidite systems, the spectacularly exposed, seismic-scale Montagne de Chalufy onlap section in the Grès d’Annot turbidite system of the French Alps is used to generate various forward synthetic seismic 2D sections. Three parameters are varied independently: input model detail, seismic impedance contrast, and dominant wavelet frequency. Seismic velocity and rock density values are taken from producing reservoirs to represent the observed lithologies in the Grès d’Annot. The three petroleum reservoir scenarios investigated are a Plio-Pleistocene Gulf of Mexico reservoir and Tertiary and Jurassic North Sea reservoirs. Two levels of 2D input model detail are investigated: 1) a detailed model having observed structural and sedimentary complexities to test vertical and lateral seismic resolution, and 2) a simple onlap model without the complexities. Two Ricker wavelet dominant frequencies are chosen, 26 and 50 Hz. Synthetic seismic results show outcrop analogues of turbidite sandstone reservoirs can be usefully converted to forward seismic sections. 2D seismic sections generated from high frequency wavelets allow subsequent geological interpretations that almost replicate detailed input models, although onlap termination or drape was still difficult to resolve. Gross architectures of massive, onlapping, turbidite sandstones may still be resolved in low frequency seismic datasets.