Mapping Reservoir Properties From Well Measurements Guided By Seismic Attributes

Abstract

ABSTRACT Well measurements and seismic attributes can be used to derive multi-dimensional calibration functions that are applied to seismic attributes to map reservoir properties. Estimation of various properties is guided by multiple seismic attributes simultaneously. The estimates fit the well measurements and are guided by the seismic data away from the wells. Significance estimation and error mapping provide a quantitative measure of confidence. Specific seismic attributes have been found to correlate well with individual reservoir properties, thus improving our ability to map them. For example, nterval velocity and acoustic impedance often correlate well with porosity, depth is useful for water saturation mapping. INTRODUCTION Seismic data are routinely and effectively used to estimate the structure of reservoirs, but are often put aside when the task is to estimate reservoir properties suchas porosity saturation and permeability. Estimation of these properties is often based solely on well measurements, even when there are relevant seismic data. Porosity, permeability, fluid and gas saturation, and other reservoir properties are measured in wells. But well data cover only the immediate environment of the boreholes, a very small fraction of the reservoir volume. Accurate high resolution estimation of reservoir properties near the wells may be wasted if the estimation of these properties away from the wells is poor. Seismic waves propagate everywhere covering the whole reservoir, not limited to wells. The seismic data are processed and the resulting image is used to delineate the geological structure. With the rising importance of reservoir characterization and stratigraphic exploration, seismic data acquisition processing and interpretation have evolved to give more information about the reservoir properties; seismic processing output is no longer a single image of reflectivity. In addition, seismic attributes can be generated. Some attributes are generated post stack by mathematical transformations (Taner1 et al, 1979; Sonneland2 et al, 1990). These include instantaneous phase and frequency, dip and azimuth, reflection heterogeneity and intensity. Other attributes are based on prestack analysis of travel time versus offset (move out), and amplitude versus offset. These include interval velocity estimation based on move out and AVO slope, intercept, and derived attributes. Acoustic impedance estimated by inversion is an important seismic attribute. In Figure 1, we show a cross section on which porosity is apparently related to acoustic impedance much better than it is to the reflectivity image (Non seismic attributes, e.g. gravity and magnetic maps, could also be used.) The problem in using seismic attributes for property estimation is that their relation to rock properties is not obvious. How would one use, for example, AVO to gas saturation with certain coefficients, and under certain conditions that might be indicated by other attributes. Region familiarity is built by comparing seismic and borehole data. There is a need for interpretation methods and tools to help one build region familiarity, quantify its reliability, and subsequently ": use it to estimate properties. In this paper we presentsuch a method.