Pressure estimation using time-lapse seismic in compacting reservoirs

Abstract

This thesis focuses on developing a new approach to estimate pressure changes from 4D amplitude attributes in compacting reservoirs. The time-lapse seismic signal in these types of reservoirs results from the combination of pressure depletion, rock compaction and stress redistribution within the reservoir and throughout the surrounding rocks. Simulations using iterative coupling are performed to understand the link between geomechanics, fuid fow and the seismic response. The analysis of synthetic data defines a power law equation (eq.1) which relates pressure changes (¢P) to 4D amplitude attributes (¢A). The coefficients (C1 and C2) are a function of initial porosity. The pressure predictions show an agreement with the output from the reservoir modelling. However the results indicate that the rock compaction has considerable effect on the normal average stress, and the 4D seismic response shows a stronger correlation with effective stress than with pore pressure. ¢A = C1 ¤ (¢PC2) (1) The technique is applied to the south east flank of the Valhall Field, Norwegian North Sea. The areas where the initial porosities are above 38 % show a good correlation between the pressure changes predicted from 4D amplitude and the pressure changes from the reservoir model. However, major differences between both outputs occur in areas where no 4D signal is observed; these areas are correlated with low porosity zones where the porosity reduction has not been significant enough to enhance the 4D signal. Furthermore, the pressure predictions from the 4D seismic identify areas where the reservoir has not been properly drained. The impact of geological structure and gas saturation on the technique is assessed. Strong thickness variations within the reservoir interval increase the errors on the pressure prediction. This is mitigated if relative values instead of absolute values are used to estimate pressure changes, i.e. equation 1 becomes equation 2. Furthermore, the presence of gas on the reservoir requires a modification of the i equation 1 in order to accurately predict the pressure changes and account for the presence of free gas on the reservoir (eq. 3). ¢A A1 = C1 ¤ ( ¢P P1 )C2 (2) ¢A = (C3 ¤ ¢P) + C4 (3) Finally, sensitivity analysis suggests that uncertainties in the elastic properties of the overburden (rarely measured with accuracy) have little impact on the reservoir 4D amplitude response. Synthetic models show that variations between 10 % to 15 % in the Poisson's ratio and Young's modulus of the layer immediately above the reservoir causes negligible changes (less than 4 %) in the 4D amplitude response.