Borehole pressure management methods and systems with adaptive learning

1. A method of managing pressure in a borehole, the method comprising:

obtaining a model-based pressure prediction and sensor-based pressure measurements during operations in the borehole;

determining real-time learning calculations based at least in part on a difference between the model-based pressure prediction and real-time feedback from at least some of the sensor-based pressure measurements;

updating formation pore pressure, fracture pressure, shear failure pressure, and in-situ stress predictions based at least in part on the real-time learning calculations and on compaction trend data filtered by an adaptive learning module in real-time; and

controlling flow rate, choke pressure, and drill rates in real-time for the operations based on the predictions.

2. The method of claim 1 , wherein determining the real-time learning calculations comprises:

comparing a model-based pore pressure curve prediction with a sensor-based pore pressure measurement curve; and

comparing a model-based fracture pressure curve prediction with a sensor-based fracture pressure measurement curve.

3. The method of claim 1 , further comprising filtering the obtained sensor-based pressure measurements using the adaptive learning module and determining the real-time learning calculations based on a filtered set of sensor-based pressure measurements.

4. The method of claim 3 , wherein the adaptive learning module includes a neural network.

5. The method of claim 3 , wherein the filtered set of sensor-based pressure measurements includes look-ahead log data and look-around log data.

6. The method of claim 3 , wherein the filtered set of sensor-based pressure measurements includes log data selected from the group consisting of: resistivity log data, acoustic log data, density log data, neutron density log data, gamma ray log data, and seismic log data.

7. The method of claim 1 , further comprising generating a log that displays pore pressure and at least one logging tool measurement as a function of depth, wherein an operator updates the predictions based at least in part on the generated log.

8. The method of claim 1 , wherein the updated formation pore pressure, fracture pressure, shear failure pressure, and in-situ stress predictions are based at least in part on historical pressure test data filtered by the adaptive learning module.

9. The method of claim 1 , wherein the updated formation pore pressure, fracture pressure, shear failure pressure, and in-situ stress predictions are based at least in part on real-time pressure test data filtered by the adaptive learning module.

10. The method of claim 1 , wherein the updated formation pore pressure, fracture pressure, shear failure pressure, and in-situ stress predictions are based at least in part on event signature data filtered by the adaptive learning module.

11. The method of claim 1 , further comprising repeating said obtaining steps, said determining step, said updating step, and said controlling step, during the borehole operations.

12. A system for managing pressure in a borehole, comprising:

a controllable choke unit to adjust pressure in the borehole; and

a processor that:

determines real-time learning calculations based at least in part on a difference between a model-based pressure prediction and real-time feedback from sensor-based pressure measurements obtained during borehole operations;

updates formation pore pressure, fracture pressure, shear pressure, and in-situ stress predictions based at least in part on the real-time learning calculations and on compaction trend data filtered by an adaptive learning module in real-time;

determines a pressure setpoint using the updated predictions in real time; and

provides the pressure setpoint to the controllable choke unit which controls a choke in real-time.

13. The system of claim 12 , wherein the processor determines the real-time learning calculations by comparing a model-based pore pressure curve prediction with a sensor-based pore pressure measurement curve, and by comparing a model-based fracture pressure curve prediction with a sensor-based fracture pressure measurement curve.

14. The system of claim 12 , wherein the processor filters sensor-based pressure measurements using the adaptive learning module, and determines the real-time learning calculations based on a filtered set of sensor-based pressure measurements.

15. The system of claim 12 , wherein the sensor-based pressure measurements correspond to look-ahead look-around log data filtered by the adaptive learning module.

16. The system of claim 12 , wherein the sensor-based pressure measurements correspond to log data filtered by the adaptive learning module, the log data selected from the group consisting of: resistivity log data, acoustic log data, density log data, neutron density log data, gamma ray log data, and seismic log data.

17. The system of claim 12 , wherein the processor generates a log that displays pore pressure in real-time and at least one logging tool measurement as a function of depth, and updates the predictions in real-time based at least in part on operator input in response to the generated log.

18. The system of claim 12 , wherein the updated formation pore pressure, fracture pressure, sheer failure pressure, and in-situ stress predictions are based at least in part on data filtered by the adaptive learning module, the data selected from the list consisting of: historical pressure test data, real-time pressure test data, and event signature data.