Operating intelligent completion valves in a hydrocarbon well using machine learning

Abstract

A system for operating intelligent completion valves (ICVs) used in a hydrocarbon well operation is disclosed. The ICVs can be positionable downhole in a wellbore. A trained machine-learning model can be generated by training a machine-learning model on training data comprising acoustic signal data generated by a pressurized injection fluid flowing through the ICVs at various flow rates and different ICV positions. When new acoustic sensing system sensor data associated with an ICV of multiple ICVs in the wellbore is applied to the trained machine-learning model, the trained machine-learning model can generate a result determining that the amplitude spike is attributable to a change in the position of the intelligent completion valve. The system may also determine a magnitude of the change in the position of the intelligent completion valve. The result may be output and used to control the ICV.

What is claimed is: 
     
         1 . A system comprising:
 a plurality of intelligent completion valves positionable downhole in a wellbore in a formation;   a distributed acoustic sensing system positionable in the wellbore;   a processor; and   a memory communicatively coupled to the processor, the memory including instructions that are executable by the processor to cause the processor to perform operations comprising:
 storing a trained machine-learning model using training data comprising acoustic signal data generated by a pressurized injection fluid flowing outward through the plurality of intelligent completion valves at various flow rates and at different intelligent completion valve positions; 
 receiving, from the distributed acoustic sensing system and in response to an amplitude spike in an acoustic signal associated with an intelligent completion valve of the plurality of intelligent completion valves, sensor data associated with the intelligent completion valve; 
 applying the sensor data to the trained machine-learning model to generate a result determining that the amplitude spike is attributable to a change in the position of the intelligent completion valve; and 
 outputting the result that is useable to control the intelligent completion valve. 
   
     
     
         2 . The system of  claim 1 , wherein the distributed acoustic sensing system includes an optical fiber cable located in the wellbore. 
     
     
         3 . The system of  claim 1 , wherein the change in position of the intelligent completion valve is a change in a choke setting of the intelligent completion valve. 
     
     
         4 . The system of  claim 1 , wherein the training data further comprises acoustic signal data generated by the pressurized injection fluid flowing outward through the plurality of intelligent completion valves at different pressurized injection fluid pressures. 
     
     
         5 . The system of  claim 1 , wherein the training data is buildable by:
 operating an injection logging tool in the wellbore to record flow rates of the pressurized injection fluid at locations near each of the intelligent completion valves;   concurrently with operating the injection logging tool, operating the distributed acoustic sensing system to record acoustic signals generated at each of the intelligent completion valves; and   while operating the injection logging tool and the distributed acoustic sensing system, controlling the intelligent completion valves such that:
 the position of the one of the plurality of intelligent completion valves through which the pressurized injection fluid flows is changed in a step-wise fashion. 
   
     
     
         6 . The system of  claim 1 , wherein training the machine-learning model comprises training the machine-learning model by supervised learning using a labeled dataset including:
 the acoustic signal data;   pressurized injection fluid flow rate data; and   engineered features selected from the group consisting of moveout-corrected stack semblance, ratio of short time average to long time average, ratio of a root-mean-square amplitude of the acoustic signal before an amplitude spike to a root-mean-square amplitude of the acoustic signal after an amplitude spike, and combinations thereof.   
     
     
         7 . The system of  claim 1 , wherein in the acoustic signal data of the training data:
 a general increase in the amplitude of the acoustic signal over time following the amplitude spike, indicates an intelligent completion valve opening event; and   a general decrease in the amplitude of acoustic signal over time following the amplitude spike indicates an intelligent completion valve closing event;   wherein the operations further comprise, in response to detecting the general increase or the general decrease in the amplitude of the acoustic signal over time:
 determining that a rate of change in the amplitude of the acoustic signal has become stable at a time within a predefined length of time following the amplitude spike in the acoustic signal; and 
 in response to determining that the amplitude of the acoustic signal has become stable within the predefined length of time, assigning the flow rate of the pressurized injection fluid at the time the rate of change in the amplitude of the acoustic signal became stable as the flow rate of the pressurized injection fluid through the intelligent completion valve at a current position of the intelligent completion valve. 
   
     
     
         8 . A method, comprising:
 receiving, by a processor of a computing device, from a distributed acoustic sensing system positionable in a wellbore in a formation, sensor data associated with an intelligent completion valve of a plurality of intelligent completion valves positionable in the wellbore, the sensor data generated in response to an amplitude spike in an acoustic signal associated with the intelligent completion valve;   applying the sensor data to a trained machine-learning model using training data comprising acoustic signal data generated by a pressurized injection fluid flowing outward through the plurality of intelligent completion valves at various flow rates and at different intelligent completion valve positions;   generating a result determining that the amplitude spike is attributable to a change in the position of the intelligent completion valve; and   outputting the result that is used to control the intelligent completion valve.   
     
     
         9 . The method of  claim 8 , wherein:
 the distributed acoustic sensing system includes an optical fiber cable located in the wellbore; and   the change in position of the intelligent completion valve is a change in a choke setting of the intelligent completion valve.   
     
     
         10 . The method of  claim 8 , wherein the training data further comprises acoustic signal data generated by the pressurized injection fluid flowing outward through the plurality of intelligent completion valves at different pressurized injection fluid pressures. 
     
     
         11 . The method of  claim 8 , wherein the training data is built by:
 operating an injection logging tool in the wellbore to record flow rates of the pressurized injection fluid at locations near each of the intelligent completion valves;   concurrently with operating the injection logging tool, operating the distributed acoustic sensing system to record acoustic signals generated at each of the intelligent completion valves; and   while operating the injection logging tool and the distributed acoustic sensing system, controlling the intelligent completion valves such that:
 the position of the one of the plurality of intelligent completion valves through which the pressurized injection fluid flows is changed in a step-wise fashion. 
   
     
     
         12 . The method of  claim 8 , wherein the machine-learning model is trained by supervised learning using a labeled dataset including:
 the acoustic signal data;   pressurized injection fluid flow rate data; and   engineered features selected from the group consisting of moveout-corrected stack semblance, ratio of short time average to long time average, ratio of a root-mean-square amplitude of the acoustic signal before an amplitude spike to a root-mean-square amplitude of the acoustic signal after an amplitude spike, and combinations thereof.   
     
     
         13 . The method of  claim 8 , wherein in the acoustic signal data of the training data:
 a general increase in the amplitude of the acoustic signal over time following the amplitude spike, indicates an intelligent completion valve opening event;   a general decrease in the amplitude of acoustic signal over time following the amplitude spike indicates an intelligent completion valve closing event; and   further comprising, in response to detecting the general increase or the general decrease in the amplitude of the acoustic signal over time:
 determining that a rate of change in the amplitude of the acoustic signal has become stable at a time within a predefined length of time following the amplitude spike in the acoustic signal; and 
 in response to determining that the amplitude of the acoustic signal has become stable within the predefined length of time, assigning the flow rate of the pressurized injection fluid at the time the rate of change in the amplitude of the acoustic signal became stable as the flow rate of the pressurized injection fluid through the intelligent completion valve at a current position of the intelligent completion valve. 
   
     
     
         14 . The method of  claim 8 , further comprising determining a magnitude of the change in the position of the intelligent completion valve. 
     
     
         15 . A non-transitory computer-readable medium comprising instructions that are executable by a processor of a computing device, for causing the processor to perform operations comprising:
 receiving, by a processor of a computing device, from a distributed acoustic sensing system positionable in a wellbore in a formation, sensor data associated with an intelligent completion valve of a plurality of intelligent completion valves positionable in the wellbore, the sensor data generated in response to an amplitude spike in an acoustic signal associated with the intelligent completion valve;   applying the sensor data to a trained machine-learning model using training data comprising acoustic signal data generated by a pressurized injection fluid flowing outward through the plurality of intelligent completion valves at various flow rates and at different intelligent completion valve positions;   generating a result determining that the amplitude spike is attributable to a change in the position of the intelligent completion valve; and   outputting the result that is useable to control the intelligent completion valve.   
     
     
         16 . The non-transitory computer-readable medium of  claim 15 , wherein:
 the distributed acoustic sensing system includes an optical fiber cable located in the wellbore; and   the change in position of the intelligent completion valve is a change in a choke setting of the intelligent completion valve.   
     
     
         17 . The non-transitory computer-readable medium of  claim 15 , wherein the training data further comprises acoustic signal data generated by the pressurized injection fluid flowing outward through the plurality of intelligent completion valves at different pressurized injection fluid pressures. 
     
     
         18 . The non-transitory computer-readable medium of  claim 15 , wherein the training data is buildable by:
 operating an injection logging tool in the wellbore to record flow rates of the pressurized injection fluid at locations near each of the intelligent completion valves;   concurrently with operating the injection logging tool, operating the distributed acoustic sensing system to record acoustic signals generated at each of the intelligent completion valves; and   while operating the injection logging tool and the distributed acoustic sensing system, controlling the intelligent completion valves such that:
 the position of the one of the plurality of intelligent completion valves through which the pressurized injection fluid flows is changed in a step-wise fashion. 
   
     
     
         19 . The non-transitory computer-readable medium of  claim 15 , wherein training the machine-learning model comprises training the machine-learning model by supervised learning using a labeled dataset including:
 the acoustic signal data;   pressurized injection fluid flow rate data; and   engineered features selected from the group consisting of moveout-corrected stack semblance, ratio of short time average to long time average, ratio of a root-mean-square amplitude of the acoustic signal before an amplitude spike to a root-mean-square amplitude of the acoustic signal after an amplitude spike, and combinations thereof.   
     
     
         20 . The non-transitory computer-readable medium of  claim 15 , wherein in the acoustic signal data of the training data:
 a general increase in the amplitude of the acoustic signal over time following the amplitude spike indicates an intelligent completion valve opening event; and   a general decrease in the amplitude of acoustic signal over time following the amplitude spike indicates an intelligent completion valve closing event;   wherein the operations further comprise, in response to detecting the general increase or the general decrease in the amplitude of the acoustic signal over time:
 determining that a rate of change in the amplitude of the acoustic signal has become stable at a time within a predefined length of time following the amplitude spike in the acoustic signal; and 
 in response to determining that the amplitude of the acoustic signal has become stable within the predefined length of time, assigning the flow rate of the pressurized injection fluid at the time the rate of change in the amplitude of the acoustic signal became stable as the flow rate of the pressurized injection fluid through the intelligent completion valve at a current position of the intelligent completion valve.