A new technique is presented for a more effective description of the vertical variability of reservoir rock properties that are typically obtained or derived from wireline logs or determined from laboratory measurements of core samples.
Rock properties and other data available for this study were the geological description of cores, XRD analysis of core samples for mineral percentages, SEM and CT-scan data from a few selected wells in two fields, namely Patrick Draw (WY) and Bell Creek (MT), laboratory measured porosity, air permeability, and primary and waterflood oil production data from the wells studied.
In this investigation, first the vertical distribution of reservoir rock properties available as a function of depth is transformed into functions of frequency (cycles per foot) by the application of standard time series analysis techniques. A statistical parameter called the "power spectrum" is calculated which gives the variance of the time series for each frequency. In this way a continuous description of variances is obtained for each frequency into which the original series has been transformed.
Estimation of vertical variability of rock properties was initiated by computation of "power spectra" of core measured permeability, and digitized gamma ray, sonic, density and neutron logs. The power decay patterns with frequency were correlated with reservoir rock properties and oil production, and it was observed that different logs had different decay patterns in sandstones having different distributions of reservoir properties. The power spectra of permeability, sonic, and gamma ray logs for a clean sandstone or for a sandstone with uniform properties (low variance) decayed very rapidly at higher frequencies if there was no thin bed stratification and assumed very low powers at the high frequency end. The sandstone with variable reservoir properties had power spectra that decayed more gently, and the slope of spectral decay was a measure of the vertical distribution of reservoir properties, in the sandstone. For a few wells, the spectra of density logs in kaolinite clay filled sandstones showed very low amplitudes at higher frequencies because the matrix rocks (predominantly quartz) and the clays (dominantly kaolinites) both had nearly the same density (around 2.64 g/cm3).
From minimal geological information, studies of highly automated spectral decay curves of different types of distributions could provide quantitative information on vertical variability of reservoir rock properties in different parts of a field. Under favorable conditions the study of decay spectra of a combination of logs may provide quantitative information on the variations in cement content and the types and volumes of clays in reservoir rocks.
Introduction
Information about the vertical variability of reservoir rock properties such as permeability, porosity, fluid saturation, lithology, and type and volumes of clay, is of critical importance for exploitation of hydrocarbon resources and effective planning of enhanced recovery projects. The vertical sweep efficiency of a waterflood depends primarily upon the distribution of permeability within the reservoir and also upon the mobility ratio. As a consequence of non-uniform permeability, any injected fluid will move through the reservoir as an irregular front. In a miscible slug process, effective design of a slug required to maintain miscibility throughout the displacement process requires knowledge about the degree of stratification of the reservoirs Froning and Treiber reported how the distribution of clays in reservoir pore surfaces may affect micellar design in several ways but primarily by higher adsorption of surfactants in high surface area clays compared to that on silica. They concluded that the amount of clay and the nature of cation present (i.e., Na+ and Ca++) will affect the ionic environment contacted by micellar fluids as it moves through the reservoir.
P. 637^
