Risk is inherent in the outcome of all reservoir characterization oil volumetries. This risk comes from the very small sample of information relative to the total reservoir volume and the error in data measurement. To incorporate the risk into reservoir characterization, some qualitative aspects are best modeled deterministically, whereas other data can be modeled statistically. Probability distributions were generated for original oil in place (OIP), original mobile OIP, original residual OIP, and remaining mobile OIP by combining geologic characteristics modeled deterministically using a 3 -D computer technique with statistical probability functions of petrophysical attributes.

The risk-assessment model applied in this study integrates wireline-derived petrophysical data with results from 3-D geocellular reservoir computer modeling. Model development is based on deterministically setting the net-pay volume derived from geologic flow-unit geometry and risk assessing the variability of the petrophysical characteristics within the flow unit. The net-pay volume is a result of the sum of the cell volumes meeting the pay cutoff criteria within the gross volume of a given flow unit. Wireline petrophysical properties have been analyzed within the flow-unit structure. Petrophysical properties were statistically analyzed to determine the best probability function that described their variability. The Weibull probability function was found to best describe the frequency distribution of porosity, whereas the gamma function best models initial water saturation and bypassed oil saturation. Monte Carlo stochastic simulation was then used to combine the deterministically derived net-pay volume with the petrophysical probability functions to generate oil-in-place density distributions. The resultant skewed nature of the OOIP frequency distribution is best modeled by a gamma function.

Risk assessment verifies that the 3-D geocellular model OOIP results are a good representation of proved and probable (2P) volumes. Geocellular model results are virtually identical to the mean value derived from stochastic simulation. Because of the skewness of the OOIP distribution, the mean OOIP is a conservative value; therefore, the 3-D geocellular model OOIP can also be considered conservative. The mode is nearly 10 percent larger than the mean, displaying a strong probability of larger volumes and quantifying the upside of the 3-D model.

You can access this article if you purchase or spend a download.