Publications

Earth modeler is facing several challenges when building models based on few wells. Two challenges are discussed in the following: how to get a reliable hydrocarbon volume estimation?, and: how to get a reliable estimation of the hydrocarbon volume uncertainty? This study focuses on a discovery with 2 appraisal wells in a fluvial environment which indicates a stacked channel system of circa 500 ft thickness. The structural model is based on 3D seismic interpretation. The value range of the mean sand fraction was estimated from the VSH logs using different cut-offs. The variation of the mean porosity was estimated from the two wells. The water saturation calculation was based on a function of porosity and height above oil-water contact. This function was shifted against the porosity axis to define the low and high case. The hydrocarbon volume uncertainty was captured using Experimental Design. It is important to understand that Experimental Design is based on low and high case values of all relevant modeling parameters. Each model with a combination of low and high case parameters should deliver a reproducible and reliable volume. For instance small reservoirs often show a hydrocarbon distribution with a large variance for one set of parameters when statistical modeling techniques like Gauss simulation are selected. In such a case a reliable strategy needs to be defined for using Experimental Design reliably. Otherwise a different methodology should be considered. Typically Experimental Design is used in two steps: first those parameters are identified that are most influential on the volume distribution. For instance the Plackett-Burman design allows identifying these so-called heavy-hitter parameters efficiently. Then, these identified parameters are used for building a more accurate design like full factorial design. The resultant proxy equation allows deriving the hydrocarbon volume distribution from a couple of thousand simulations.