Aquifer mapping using Pre-stack Seismic Data, WDDM, Nile Delta.

Abstract

Water production is the main production problem in WDDM concession in Nile delta. Water comes in surprisingly so fast with high rates. This comes from the underestimation of the aquifer description. Usually aquifer is estimated "guessed" and modeled as numerical aquifer. Unfortunately stacked seismic reflection data doesn't help so much in mapping aquifers. Water sands reflectivities appear as much dimmer amplitudes. If seismic can give a hand in this it would be priceless knowledge. The study area, Sequoia gas field, is located in West Delta Deep Marine (WDDM) concession. It is about 70 kilometers to the North East offshore Alexandria.

The rock physics model suggests that, the intercept and the gradient of a shale-to-shale interface are close to those of shale-to-water interface. However the shale-to-gas interface has about the same magnitude (but opposite polarity) of shale-to-water however the gradient is much higher. So when the seismic CDPs are stacked, gas sand amplitudes dominates the picture and the water sands lie in the same color range of the shales. In order to overcome this problem, we need to go back to the pre-stack data and invert them for the physical properties, so their effect on the post-stack can be separated, analyzed and used for the seismic facies classification.

AVO simultaneous inversion inverts pre-stack seismic into elastic properties; compressional impedance, shear impedance and density. The products of AVO simultaneous inversion can be coupled with other wire-line logging data via Bayesian classification, so seismic sections can be transformed into geologic cross-sections with meaningful facies codes. Then water sands can be discriminated from other facies existing in the field so they can be delineated, mapped and modeled in the 3D static models. The Bayesian classification outputs gave a reasonable match to the actual facies found in the wells of the study area.