An integrated target‐oriented prestack elastic waveform inversion: Sensitivity, calibration, and application

Abstract

An integrated target‐oriented prestack elastic waveform inversion procedure is implemented to estimate lithology, fluid type, and pay thickness of a layered target zone from surface seismic PP reflection data. The integrated inversion procedure consists of three parts: data processing and preparation, iterative seismic inversion, and rock property integration. In the data processing and preparation, the seismic field gathers are processed and transformed into plane‐wave seismograms, and the well‐logs are prepared for establishing a well tie. The well tie is established using synthetics from a layered model obtained from blocking sonic, shear sonic, and density logs. The layered model is later used as the a priori starting model for the iterative seismic inversion. Based on the a priori starting model and using the processed plane‐wave seismograms as input, P‐wave velocity ( [Formula: see text]), S‐wave velocity ([Formula: see text]), density ρ, and thickness (h) of each layer in the layered target zone are estimated at each CDP location in the iterative seismic inversion. Through the rock property integration, the inverted seismic parameters are related to the reservoir properties: the lithology/fluid contents, and the net pay thickness. Synthetic examples of gas‐related bright spot models are used to illustrate the functionality of the procedure. The sensitivity and the vertical resolution of the inversion procedure are investigated. Under favorable conditions, the inversion procedure is able to estimate the net pay thickness down to a tenth of a wavelength. In noisy situations, the inversion tends to break down where the thickness of target is approximately equal to the tuning thickness. The procedure was successfully applied to a gas‐related shallow bright spot prospect offshore Gulf of Mexico. The lithology (sand shale distribution), [Formula: see text] ratio, and the net pay thickness of gas‐sand were derived for the prospect. The estimated net pay thickness of gas‐sand was within acceptable error range when compared with the actual net pay thickness measured from well logs.