The influence of overburden on time-lapse saturation interpretation

This paper studies the effect of overburden channels on saturation changes estimated from 4D seismic data. It considers the impact of source and receiver positioning in combination with a realistic overburden and reservoir structure on the ability to interpret production induced changes. The study is based on data from the Nelson field where Quaternary and Miocene age channels above the reservoir are known to cause problems when interpreting the 4D signal. Synthetic pre-stack seismic gathers are generated and processed to limited offset stacks. Interpretation of the saturation changes estimated using these post-stack data reveals contributions from spatially incoherent noise and an anomalous seismic response introduced by the presence of the channels. Most of the errors in the saturation change estimates are confined to areas directly below the channel location. For seismic attributes affected by amplitude changes alone, localised errors of as much as 30% occur below the steep dipping channel margins, which should be compared to the maximum saturation change of 45%. The absolute error in the saturation change predictions increases to 68% if the attributes are influenced by amplitude and time shift changes. The errors revealed cannot be readily predicted by considering only acquisition based repeatability metrics.