The 3D view shows the reservoir grid modeled inside the 3D model fitting the 11 wells. As the cross-section shows, the wells are primarily interpreted to be inside the point bars.

The granulometry shown for each grid cell follows the river profile and can be used as trend information to interpolate the well's derived porosity and permeability values. 

2) ​​​​Meandering river from Cranfield data (by Ismael Dawuda)

The model fits a tightly spaced set of pseudo wells interpreted along the outcrop, including a local erosion and a very heterogeneous deposition profile across the different wells. As shown in the figure above, thin shale layers are preserved.

1) ​​​Turbidite Lobe (Outcrop data from Sébastien Rohais et al.)

We model a sea-floor fan from 15 wells in a carbonate-siliclastic turbidite environment. The picture above shows that the well's lithofacies data is matched at each location while generating a realistic fan geometry at every deposition step. The generated granulometry trend in the volume provides valuable information to interpolate rock properties information.

5) ​​​​Modeling Meandering Channels With Seismic Control (from V. Koala et al.)


In this example, we incorporate seismic data to refine the geological model. From the seismic interpretation, we add two listric faults and one channel belt into the turbiditic deposition environment. The two listric faults rotate the fault block, creating a mini-deposition basin. The channel belt information constrains the location of the main river. This example shows how our process-based modeling method can use seismic and geological interpretation to build constrained, realistic models.

The 3D reservoir grid is built inside the 3D model shown above. We see how the thin shale layer is preserved as part of the grid geometry or topology.

Given two seismic slices where the same migrating channel is interpreted, we create the intermediate channel geometries (including the point bars). We also create the reservoir grid shown in the picture. In function of the river migration and deposition rates, we can stochastically generate many models whose connectivity will vary greatly, providing an extensive range of uncertainty on the connected volumes.

3) ​​​​Campos Basin Turbidite (from Carlos Bruhn)

4) ​​​​Delaware Basin WolfCamp A Turbidite (from E. Kvale et al.)