System and method for joint inversion of bed boundaries and petrophysical properties from borehole logs
Systems and methods for optimizing bed boundary locations and prediction of associated petrophysical property values of a wellbore in a reservoir formation are disclosed. To optimize bed boundary locations and predict petrophysical property values, a process can be used in which a high resolution wellbore log is used to get an initial estimate of boundary positions. The initial estimate can then be used to constrain a joint optimization of bed positions and associated petrophysical property values. Such a process automates both the generation of a single set of boundaries that are common to all measurements and the prediction of petrophysical properties via joint inversion. The same inversion process can be used to derive quality indicators on the corrected bed boundary locations and the petrophysical property values.
1. A method for evaluating a bedded formation for extraction of a commodity therefrom, the bedded formation having one or more bed boundary locations intersected by a wellbore, the method comprising:
measuring image data for at least one wellbore log measured over the bedded formation having the one or more bed boundary locations intersected by the wellbore by logging the bedded formation with a wellbore logging tool in the wellbore;
calculating at least one axial response function (ARF) for each of the at least one wellbore log;
estimating one or more initial measurements for the one or more bed boundary locations;
calculating a maximum bed boundary correction margin for the one or more initial measurements;
calculating an optimized bed boundary location for each of the one or more initial measurements and calculating one or more petrophysical property values for the bedded formation by solving a constrained optimization problem based on the measured image data and the maximum bed boundary correction margin;
correcting the image data of the bedded formation for the wellbore with the one or more calculated petrophysical property values relative to the one or more optimized bed boundary locations; and
quantifying the commodity in the bedded formation using the corrected image data.
2. The method of claim 1 , wherein one of the one or more petrophysical property values comprises a bed density value.
3. The method of claim 1 , wherein one of the one or more petrophysical property values comprises at least one of bed porosity, gamma ray, spectral gamma ray, photoelectric absorption factor (PEF), elemental composition, neutron capture cross section, and acoustic velocity.
4. The method of claim 1 , wherein one of the at least one wellbore logs comprises at least one of a density log, a porosity log, a gamma ray log, a spectral gamma ray log, a photoelectric absorption factor (PEF) log, an elemental composition log, a neutron capture cross section log, and an acoustic velocity log.
5. The method of claim 1 , one of the at least one wellbore logs is an image log.
6. The method of claim 1 , wherein the constrained optimization problem to compute the optimized bed boundary locations and one or more petrophysical property values is a nonlinear problem that is solved using an iterative alternating optimization between boundary corrections and inversion of petrophysical property values starting from the estimated initial measurements.
7. The method of claim 1 , wherein the constrained optimization problem to compute the optimized bed boundary locations and petrophysical property values is reduced to a bilinear problem.
8. The method of claim 7 , wherein the bilinear problem is decomposed into multiple subproblems when the constrained optimization problem is of a large scale.
9. The method of claim 7 , wherein the bilinear problem is solved using a sequential linear programming procedure.
10. The method of claim 9 , wherein the sequential linear programming procedure is coupled with a cutting plane procedure.
11. The method of claim 7 , wherein the bilinear problem is solved by using a branch-and-bound-algorithm.
12. The method of claim 11 , wherein the branch-and-bound algorithm for solving the bilinear problem is used to derive quality indicators for any of the calculated optimized bed boundary locations or petrophysical property values.
13. The method of claim 12 , wherein the quality indicators represent at least one of one or more upper and lower bounds on one or more calculated petrophysical property values or an aggregation thereof.
14. The method of claim 12 , wherein the quality indicators represent a maximum variability for any combination of the calculated petrophysical property values.
15. The method of claim 12 , wherein the quality indicators represent a statistic of interest on one or more of the calculated petrophysical property values or a combination thereof.
16. The method of claim 7 , further comprising casting the bilinear problem to a linear problem by employing convex and concave relaxations.
17. The method of claim 1 , wherein calculating at least one ARF comprises:
performing a simulation to compute a discretely sampled ARF; and
modeling the discretely sampled ARF via a continuous model.
18. The method of claim 1 , further comprising calculating one or more thresholds on a maximum error in the received data from the at least one wellbore log.
19. The method of claim 18 , wherein the calculation of the optimized bed boundary locations and petrophysical property values is done by sampling from one or more bounded bipolyhedral spaces based on the received data, the correction margin, and the one or more thresholds.
20. A non-transitory program storage device, readable by a processor and comprising instructions stored thereon to cause one or more processors to perform a method for bed boundary measurement correction and petrophysical property prediction according to claim 1 .
21. The non-transitory program storage device of claim 20 , wherein the constrained optimization problem to calculate the optimized bed boundary locations and petrophysical property values is a nonlinear problem that is solved using an iterative alternating optimization between boundary corrections and inversion of petrophysical property values starting from the estimated initial measurements.
22. The non-transitory program storage device of claim 20 , wherein the constrained optimization problem to compute the optimized bed boundary locations and petrophysical property values is reduced to a bilinear problem.
23. The non-transitory program storage device of claim 22 , wherein the bilinear problem is decomposed into multiple subproblems when the constrained optimization problem is of a large scale.
24. The non-transitory program storage device of claim 22 , wherein the bilinear problem is solved using a sequential linear programming procedure.
25. The non-transitory program storage device of claim 24 , wherein the sequential linear programming procedure is coupled with a cutting plane procedure.
26. The non-transitory program storage device of claim 22 , wherein the bilinear problem is cast to a linear problem by employing convex and concave relaxations.
27. The non-transitory program storage device of claim 22 , wherein the bilinear problem is solved by using a branch-and-bound algorithm.
28. The non-transitory program storage device of claim 27 , wherein the branch-and-bound algorithm for solving the bilinear problem is used to derive quality indicators for any of the calculated optimized bed boundary locations or petrophysical property values.
29. The non-transitory program storage device of claim 20 , wherein the instructions stored thereon further cause the one or more processors to calculate one or more thresholds on a maximum error in the received data from the at least one wellbore log.
30. The non-transitory program storage device of claim 29 , wherein the calculation of the optimized bed boundary locations and petrophysical property values is done by sampling from one or more bounded bipolyhedral spaces based on the received data, the correction margin, and the one or more thresholds.
31. The non-transitory program storage device of claim 20 , wherein the instructions stored thereon cause the one or more processors to calculate at least one ARF by:
performing a simulation to compute a discretely sampled ARF; and
modeling the discretely sampled ARF via a continuous model.
32. A system for evaluating a bedded formation for extraction of a commodity therefrom, the bedded formation having one or more bed boundary locations intersected by a wellbore, the system comprising:
a wellbore logging tool logging the bedded formation having the one or more bed boundary locations intersected by the wellbore and measuring image data for at least one wellbore log;
a memory for storing the measured image data for the at least one wellbore log;
a display device; and
a processor operatively coupled to the memory and the display device and adapted to execute program code stored in the memory to:
receive the measured image data from the at least one wellbore log measured over the bedded formation;
calculate at least one axial response function (ARF) for each of the at least one wellbore logs;
estimate one or more initial measurements for one or more bed boundary locations;
calculate a maximum bed boundary correction margin for the one or more initial measurements;
calculate an optimized bed boundary location for each of the one or more initial measurements and calculate one or more petrophysical property values for the bedded formation by solving a constrained optimization problem based on the received data and the maximum bed boundary correction margin;
correct the image data of the bedded formation for the wellbore with the one or more calculated petrophysical property values relative to the one or more optimized bed boundary locations; and
quantify the commodity in the bedded formation using the corrected image data.
33. The system of claim 32 , wherein the constrained optimization problem to compute the optimized bed boundary locations and one or more petrophysical property values is a nonlinear problem that is solved using an iterative alternating optimization between boundary corrections and inversion of petrophysical property values starting from the estimated initial measurements.
34. The system of claim 32 , wherein the constrained optimization problem to compute the optimized bed boundary locations and petrophysical property values is reduced to a bilinear problem.
35. The system of claim 34 , wherein the bilinear problem is decomposed into multiple subproblems when the constrained optimization problem is of a large scale.
36. The system of claim 34 , wherein the bilinear problem is solved using a sequential linear programming procedure.
37. The system of claim 36 , wherein the sequential linear programming procedure is used with a cutting place procedure.
38. The system of claim 34 , wherein the bilinear problem is solved using a branch-and-bound algorithm.
39. The system of claim 38 , wherein the branch-and-bound algorithm for solving the bilinear problem is used to derive quality indicators for any of the calculated optimized bed boundary locations or petrophysical property values.
40. The system of claim 34 , wherein the bilinear problem is cast to a linear problem by employing convex and concave relaxations.
41. The system of claim 32 , wherein calculating at least one ARF comprises:
performing a simulation to compute a discretely sampled ARF; and
modeling the discretely sampled ARF via a continuous model.
42. The system of claim 32 , wherein the processor is further adapted to execute program code stored in the memory to calculate one or more thresholds on a maximum error in the received data from the at least one wellbore log.
43. The system of claim 42 , wherein the calculation of the optimized bed boundary locations and petrophysical property values is done by sampling from one or more bounded bipolyhedral spaces based on the received data, the correction margin, and the one or more thresholds.