Abstract Stochastic maps of reservoir lithology are extremely useful for planning during field development, production and management. Moreover, lithology types may have significant controls over porosity and permeability distributions within the reservoir. Therefore, the stochastic mapping of lithology may be considered to he the first step in geostatistical modeling of porosity and permeability distributions during reservoir characterization.
Several stochastic mapping (i.e. conditional simulation) techniques have been developed in recent years to generate equally-probable, multiple realizations of reservoir lithology. Bach realization constitutes a high resolution stochastic image, or map, displaying a realistic level of heterogeneity. It also reproduces the histogram and the spatial pattern of the actual data besides honoring these data at all sample locations.
In this study, the Sequential Indicator Simulation (SIS) technique was employed to generate stochastic maps in a carbonate reservoir consisting of three lithology types (dolomite, limestone and anhydrite). The basic data consisted of lithology indicators representing these lithology types and were derived using several techniques. The simulation provided a total of twenty realizations indicating the presence or absence of one of the three lithologies within cells each having 300m × 300m × 0.5ft dimensions. The results indicate an excellent match between the histogram of the input data and that of the simulated values. Comparison of lithology maps derived from simulation with those of the corresponding well-log porosity fields indicates that porosity does not correlate well with lithology except locally with anhydrite lenses. This is explained by the gradual change of well-log porosity values across lithologic boundaries observed in the input data.
Introduction Lithologic maps and/or cross-sectional plots of lithology are extremely useful tools for planning purposes during the entire lifetime of oil field projects, particularly for field development, production and management. Classical deterministic or geological maps are generally smooth and do not reveal a realistic level of heterogeneity. Therefore, stochastic mapping or conditional simulation techniques have been employed in recent years to provide stochastic maps of lithology. Stochastic mapping of lithology has also been proposed as the first step in modeling petrophysical properties such as porosity and permeability whenever the lithology types have significant control on distribution of these variables.
Stochastic mapping techniques provide equally-probable, multiple realizations of lithology distribution within the reservoir. P. 183^