Gas channel delineation utilizing a neural network and 3D seismic attributes: Simian Field, offshore Nile Delta, Egypt

Abstract

Depositional architecture and hydrocarbon-bearing reservoir detection are considered the main challenges in the Nile Delta Basin, predominantly in the slope area. Channel fill deposits are very complicated because they are vertically stacked and pass through different depositional stages, forming complex depositional patterns. In the current study, an artificial neural network was used to detect gas and water-bearing reservoirs inside the complex depositional pattern and their spatial dimensions. Modern seismic reservoir characterization attribute analysis provides a good outcome in detecting reservoir fluids by mixing fluid-sensitive attributes with the well outcomes. Artificial Neural Network (ANN) algorithms for rock property prediction are considered to be a new emerging trend. The ANN algorithm has an advantage over other seismic reservoir delineation techniques because it has a good capability for building a nonlinear relationship between seismic data and petrophysical logs.Hence, it can be utilized to spatially anticipate reservoir fluids with a rational degree of accuracy. A back-error propagation algorithm, a supervised neural network method, was implemented on the Simian Field to anticipate the fluid-bearing reservoirs from the seismic-derived amplitudes. The ANN method showed unique results in the case study area by discriminating the gas-bearing intervals from the water-bearing intervals, although combining the assembly of the seismic attributes to supervise fluid classification. The method will help in the development plans of the field and track the volume of water and gas.