Integration of subsurface applications to develop a dynamic stochastic modeling workflow

Building stochastic models in a preproduction phase of field development is crucial for accurate horizontal well positioning in the Sincor field (Orinoco heavy-oil belt, Venezuela), formed by highly permeable unconsolidated fluvial and deltaic sands of the Oficina Formation and containing low-gravity oil (average 8.3degrees API). A multidisciplinary approach to drill the initial approximately 250 horizontal production wells has proven useful to field management. Geomodeling is used to evaluate well pattern development plans; calculate full-field production percentile profiles; and evaluate uncertainties in water production, well production potential, and cluster performance. Data from vertical observation wells (full well-logging suite) and deviated wells (to investigate the stratigraphy away from the vertical well) are used to characterize reservoir architecture. Three-dimensional seismic data, including seismic data inverted into acoustic impedance, are used to construct structure maps and shale probability maps. The integrated and interpreted information is used to position horizontal wells deterministically in a cluster-type (group of wells radiating out from a central point in a 3.2 x 1.6-km [2 x 1-mi] production area) development pattern and serves as input to stochastic reservoir models that are conditioned to vertical well observations and the shale probability maps. Ten realizations are used to estimate the uncertainty range of net-to-gross in the proposed horizontal well trajectories. To ensure a stable model, the average of these 10 realizations is used as a trend to generate one stochastic petrophysical model that is subsequently used for flow simulation. Based on the results, and in combination with the confidence in the interpreted geological data (for example, distribution of areas with a net sand thickness above a minimum), proposed horizontal wells are accepted or rejected.