No abstract

Low-frequency forced-oscillation methods applied to a reservoir sandstone allowed determination of the 15 Young's modulus and Poisson's ratio (from axial loading), bulk modulus (by oscillation of the confining 16 pressure), and shear modulus (from torsional forced-oscillation), for comparison with conventional 17 ultrasonic data. All tests were performed on a common sandstone core sample from an oil reservoir 18 offshore West Africa. The results show a steady increase in ultrasonic velocities and shear modulus of the 19 dry specimen as functions of pressure, which suggests a progressive closure of the inter-granular contacts. 20 An increase of bulk and Young's moduli and Poisson's ratio is observed on decane-saturation of the 21 sample when tested with a sufficiently small dead volume. This observation, consistent with Gassmann's 22 theory, suggests that such measurements probe undrained (saturated isobaric) conditions. Diminution or 23 absence of such fluid-related stiffening for low-frequency measurements with dead volumes comparable 24 with the pore volume of the specimen indicate partially drained conditions and highlight the critical role 25 of experimental boundary conditions. Directly measured bulk and shear moduli are consistent with those 26 derived from Young's modulus and Poisson's ratio. These results of the inter-laboratory testing using 27 different measurement devices are consistent in terms of the effect of frequency and fluid saturation for 28 the reservoir sandstone specimen. Such broad consistency illustrates the validity of forced-oscillation 29 techniques and constitutes an important benchmarking of laboratory testing of the elastic properties of a 30 porous medium. 31 32 1.0 Introduction 33 Time-lapse (4D) seismic technology is an important reservoir monitoring tool for the oil and gas industry 34

Abstract. The Enfield rock physics model was constructed to enable 4D feasibility studies and interpretation of the 2007 Enfield 4D seismic monitor survey. The rock physics model links reservoir static and dynamic parameters to impedances, using log data from five wells in the field, laboratory core measurements taken from cores on Enfield and neighbouring fields, and theoretical rock models from the literature. The reservoir is modelled by a sand-shale mix: sand properties are described using a modified critical porosity model whereas shale properties are generated from log data averaging. The dynamic properties in the model include saturation and pressure. Saturation is modelled using Gassmann's formula assuming homogeneous mixing. The reservoir sand velocity-pressure relationship is described by an empirical model fitted to dry core plug measurements. An assessment of the effect of uncertainty is included for both the saturation and pressure elements of the model. The resultant rock physics model was used before the acquisition of the seismic monitor survey to assess the likelihood of detecting a 4D seismic signal only 7 months after production start-up. Our modelling results indicate that the strong pressure build-up around the water injectors would result in a detectable 4D seismic signal and this prediction is confirmed by the successful 4D seismic monitor data acquired in 2007. The rock physics model has been validated against the 4D monitor data and is being used to quantify the 4D interpretation, linking the observed 4D response back to predicted pressure and saturations changes in the field.

Multiple state-of-the-art inversion methods have been implemented to integrate 3D seismic amplitude data, well logs, geologic information, and spatial variability to produce models of the subsurface. Amplitude variation with angle (AVA) deterministic, stochastic, and wave-equation-based amplitude variation with offset (WEB-AVO) inversion algorithms are used to describe Intra-Triassic Mungaroo gas reservoirs located in the Carnarvon Basin, Western Australia. The interpretation of inverted elastic properties in terms of lithology- and fluid-sensitive attributes from AVA deterministic inversion provides quantitative information about the geomorphology of fluvio-deltaic sediments as well as the delineation of gas reservoirs. AVA stochastic inversion delivers higher resolution realizations than those obtained from standard deterministic methods and allows for uncertainty analysis. Additionally, the cosimulation of petrophysical parameters from elastic properties provides precise 3D models of reservoir properties, such as volume of shale and water saturation, which can be used as part of the static model building process. Internal multiple scattering, transmission effects, and mode conversion (considered as noise in conventional linear inversion) become useful signals in WEB-AVO inversion. WEB-AVO compressibility shows increased sensitivity to residual/live gas discrimination compared to fluid-sensitive attributes obtained with conventional inversions.

No abstract