Probabilistic area of interest identification for well placement planning under uncertainty

1. A computer-implemented method for well placement planning, the method comprising:

receiving a set of reservoir model realizations covering a subsurface volume, each reservoir model realization including a plurality of cells representing respective subvolumes of the subsurface volume, each cell in each reservoir model realizations including one or more properties representative of the respective subvolume of the subsurface volume;

generating, for each of the set of reservoir model realizations, a corresponding opportunity map including a plurality of entries corresponding to respective cells in the corresponding reservoir model realization, including, for each entry in the corresponding opportunity map, calculating and storing an opportunity value representative of recovery potential based upon at least a portion of the one or more properties in the respective cell in the corresponding reservoir model realization;

generating a combined probability map from the opportunity maps corresponding to the set of reservoir models, the combined probability map including a plurality of entries corresponding to respective entries in each of the opportunity maps, including, for each entry in the combined probability map, calculating and storing a probability value representative of a probability of the opportunity values stored in the respective entries in the opportunity maps meeting an opportunity criterion;

accessing the combined probability map to identify one or more areas of interest in the subsurface volume based upon the probability values stored in the entries of the combined probability map; and

generating a well placement plan that places at least one well in the identified one or more areas of interest.

2. The method of claim 1 , further comprising, prior to generating the corresponding opportunity maps, screening the set of reservoir model realizations to reduce redundancies in the set of reservoir model realizations.

3. The method of claim 2 , wherein screening the set of reservoir model realizations includes executing image analysis on the set of reservoir model realizations to identify one or more groups of redundant reservoir model realizations.

4. The method of claim 3 , wherein screening the set of reservoir model realizations further includes, after executing image analysis:

selecting a representative reservoir model realization for each group of redundant reservoir model realizations; and

removing from the set of reservoir model realizations each non-selected reservoir model realization in each group of redundant reservoir model realizations.

5. The method of claim 3 , wherein executing image analysis is performed using a neural network.

6. The method of claim 3 , wherein executing image analysis includes performing physics-independent pattern recognition on the set of reservoir model realizations.

7. The method of claim 3 , wherein executing image analysis includes performing physics-based pattern recognition on the set of reservoir model realizations.

8. The method of claim 1 , further comprising, prior to generating the corresponding opportunity maps, calibrating each of the plurality of reservoir model realizations.

9. The method of claim 8 , wherein calibrating each of the plurality of reservoir model realizations includes history matching each of the plurality of reservoir model realizations.

10. The method of claim 9 , wherein the history matching comprises ensemble-based history matching.

11. The method of claim 10 , wherein the history matching includes applying a Kalman filter.

12. The method of claim 1 , wherein calculating each opportunity value includes:

calculating a simulation opportunity from at least flow capacity, mobile oil saturation and oil pore volume determined for the respective cell in the corresponding reservoir model realization; and

calculating a reservoir opportunity from at least reservoir quality, pressure gradient and oil pore volume determined for the respective cell in the corresponding reservoir model realization.

13. The method of claim 1 , wherein calculating each opportunity value is performed using a formula:

OV=0.315483*NTG* H *( S OIL −S OR )*( k *ϕ) 0.5 *( P−P initial ) (1/3)

where OV is the opportunity value, NTG is net to gross, H is cell height, SOIL is oil phase saturation, SOR is critical or residual oil saturation, k is permeability, ϕ is porosity, P is actual reservoir pressure and Pinitial is initial reservoir pressure.

14. The method of claim 1 , wherein the opportunity criterion includes a threshold, and wherein the probability value is representative of a probability of the opportunity values stored in the respective entries in the opportunity maps meeting the threshold.

15. The method of claim 14 , further comprising:

generating a plurality of combined probability maps using different thresholds; and

accessing the plurality of combined probability maps to identify one or more areas of interest in the subsurface volume based upon the probability values stored in the entries of the combined probability maps.

16. The method of claim 1 , wherein accessing the combined probability map to identify one or more areas of interest in the subsurface volume based upon the probability values stored in the entries of the combined probability map includes combining entries in the combined probability map based upon potential well trajectories.

17. The method of claim 16 , wherein combining entries in the combined probability map based upon potential well trajectories includes combining entries in vertical columns of the combined probability map.

18. The method of claim 1 , wherein generating the well placement plan includes executing an iterative well placement optimizer restricted to the identified one or more areas of interest to accelerate convergence of the iterative well placement optimizer.

19. The method of claim 1 , wherein generating the well placement plan includes:

executing numerical simulation on the at least one well in the identified one or more areas of interest; and

evaluating performance and risk of the at least one well based upon the numerical simulation.