Flow-after-flow tests in hydrocarbon wells
Disclosed are methods, systems, and computer-readable medium to perform operations including: receiving historical production data associated with a hydrocarbon well; preprocessing the historical production data to remove noise from the historical production data; using one or more machine-learning algorithms and the preprocessed historical production to train a simulation model to simulate a flow-after-flow test for the hydrocarbon well; and testing the simulation model to determine that the simulation model passes predetermined testing criteria.
1. A computer-implemented method, comprising:
receiving, from well sensors, historical production data associated with a hydrocarbon well;
preprocessing the historical production data to remove noise from the historical production data;
using one or more machine-learning algorithms and the preprocessed historical production data to train a machine-learning model to simulate a flow-after-flow test for the hydrocarbon well, the machine-learning model accounting for variance in reservoir pressure data;
testing the machine-learning model, using known data sets, to determine, by comparing to a threshold, a difference between an output of the machine-learning model and a known output, that the machine-learning model passes predetermined testing criteria;
monitoring, using the machine-learning model, an operation of the hydrocarbon well in real-time to detect an operating error in the operation of the hydrocarbon well, the operation of the hydrocarbon well comprising flowing pressure and flowing temperature; and
in response to detecting an operating error in the operation of the hydrocarbon well during the monitoring, performing a corrective action comprising modifying a flow rate of the hydrocarbon well.
2. The computer-implemented method of claim 1 , wherein the historical production data comprises flow rates, flowing pressure, and flowing temperatures.
3. The computer-implemented method of claim 1 , wherein an input to simulate the flow-after-flow test is an input flow rate of the hydrocarbon well and an output of the flow-after-flow test is simulated flowing pressures and simulated flowing temperatures for the hydrocarbon well.
4. The computer-implemented method of claim 1 , wherein the method further comprises:
performing a quality check of preprocessed data by removing noise from the preprocessed data generated by preprocessing the historical production data.
5. The computer-implemented method of claim 1 , wherein the method further comprises:
receiving an input flow rate of the hydrocarbon well; and
using the machine-learning model to simulate the flow-after-flow test for the hydrocarbon well.
6. The computer-implemented method of claim 1 , wherein
the operating error indicates an abnormal well behavior.
7. The computer-implemented method of claim 1 , wherein the corrective action is outputting an alarm indicating the operating error or causing the hydrocarbon well to be removed offline.
8. A non-transitory computer-readable storage media coupled to one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
receiving, from well sensors, historical production data associated with a hydrocarbon well;
preprocessing the historical production data to remove noise from the historical production data;
using one or more machine-learning algorithms and the preprocessed historical production data to train a machine-learning model to simulate a flow-after-flow test for the hydrocarbon well, the machine-learning model accounting for variance in reservoir pressure data;
testing the machine-learning model, using known data sets, to determine, by comparing to a threshold, a difference between an output of the machine-learning model and a known output, that the machine-learning model passes predetermined testing criteria;
monitoring, using the machine-learning model, an operation of the hydrocarbon well in real-time to detect an operating error in the operation of the hydrocarbon well, the operation of the hydrocarbon well comprising flowing pressure and flowing temperature; and
in response to detecting an operating error in the operation of the hydrocarbon well during monitoring, performing a corrective action comprising modifying a flow rate of the hydrocarbon well.
9. The non-transitory computer-readable storage media of claim 8 , wherein the historical production data comprises flow rates, flowing pressure, and flowing temperatures.
10. The non-transitory computer-readable storage media of claim 8 , wherein an input to simulate the flow-after-flow test is an input flow rate of the hydrocarbon well and an output of the flow-after-flow test is simulated flowing pressures and simulated flowing temperatures for the hydrocarbon well.
11. The non-transitory computer-readable storage media of claim 8 , wherein the operations further comprise:
performing a quality check of preprocessed data by removing noise from the preprocessed data generated by preprocessing the historical production data.
12. The non-transitory computer-readable storage media of claim 8 , wherein the operations further comprise:
receiving an input flow rate of the hydrocarbon well; and
using the machine-learning model to simulate the flow-after-flow test for the hydrocarbon well.
13. The non-transitory computer-readable storage media of claim 8 , wherein:
the operating error indicates an abnormal well behavior.
14. The non-transitory computer-readable storage media of claim 8 , wherein the corrective action is outputting an alarm indicating the operating error or causing the hydrocarbon well to be removed offline.
15. A system, comprising:
one or more processors; and
a computer-readable storage device coupled to the one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
receiving, from well sensors, historical production data associated with a hydrocarbon well;
preprocessing the historical production data to remove noise from the historical production data;
using one or more machine-learning algorithms and the preprocessed historical production data to train a machine-learning model to simulate a flow-after-flow test for the hydrocarbon well, the machine-learning model accounting for variance in reservoir pressure data;
testing the machine-learning model, using known data sets, to determine by comparing to a threshold, a difference between an output of the machine-learning model and a known output, that the machine-learning model passes predetermined testing criteria;
monitoring, using the machine-learning model, an operation of the hydrocarbon well in real-time to detect an operating error in the operation of the hydrocarbon well, the operation of the hydrocarbon well comprising flowing pressure and flowing temperature; and
in response to detecting an operating error in the operation of the hydrocarbon well during monitoring, performing a corrective action comprising modifying a flow rate of the hydrocarbon well.
16. The system of claim 15 , wherein the historical production data comprises flow rates, flowing pressure, and flowing temperatures.
17. The system of claim 15 , wherein to simulate the flow-after-flow test is an input flow rate of the hydrocarbon well and an output of the flow-after-flow test is simulated flowing pressures and simulated flowing temperatures for the hydrocarbon well.
18. The system of claim 15 , wherein the operations further comprise:
performing a quality check of preprocessed data by removing noise from the preprocessed data generated by preprocessing the historical production data.
19. The system of claim 15 , wherein the operations further comprise:
receiving an input flow rate of the hydrocarbon well; and
using the machine-learning model to simulate the flow-after-flow test for the hydrocarbon well.
20. The system of claim 15 , wherein:
the operating error indicates an abnormal well behavior.