Systems and methods for forecasting well productivity
Disclosed are systems and methods for obtaining an input sequence of input data features associated with a well for at least one time stamp during a period of time including well production rates for the well and well operation constraints for the well, dividing the input data features into a training data subset, a validation data subset, and a test data subset, building a well production model for the well using machine learning based on the training data subset, the validation data subset, and the test data subset, and generating a forecast for the well for a future period of time using the well production model, the forecast comprising a future well production rate for the well including at least one of an oil rate, a gas rate, and a water rate.
1. A method comprising:
obtaining, by at least one processor, an input sequence of input data features associated with a well for at least one time stamp during a period of time, the input data features associated with the well for the at least one particular time stamp during the period of time comprising well production rates for the well and well operation constraints for the well;
dividing, by the at least one processor, the input data features into a training data subset, a validation data subset, and a test data subset;
building, by the at least one processor, a well production model for the well using machine learning based on the training data subset, the validation data subset, and the test data subset;
building the well production model for the well using one of long short-term memory (LSTM) and gated recurrent unit (GRU); and
generating, by the at least one processor, a forecast for the well for a future period of time using the well production model, the forecast comprising a future well production rate for the well including at least one of an oil rate, a gas rate, and a water rate.
2. The method of claim 1 , further comprising optimizing well production for the well based on the forecast for the well for the future period of time.
3. The method of claim 1 , wherein the input data features associated with the well for the at least one particular time stamp during the period of time further comprise temporal features associated with the well including an oil rate for the at least one particular time stamp during the period of time, a gas rate for the at least one particular time stamp during the period of time, and a water rate for the at least one particular time stamp during the period of time, spatial features associated with the well, and spatial images associated with the well.
4. The method of claim 1 , further comprising processing the spatial images associated with the well using a convolutional neural network (CNN).
5. The method of claim 1 , wherein the well is a first well, the method further comprising generating a forecast for a future period of time using the well production model for the first well and a second well different from the first well.
6. The method of claim 1 , further comprising deploying the well production model to a server computing device.
7. A system comprising:
at least one processor coupled with at least one computer-readable storage medium having stored therein instructions which, when executed by the at least one processor, causes the system to:
obtain an input sequence of input data features associated with a well for at least one time stamp during a period of time, the input data features associated with the well for the at least one particular time stamp during the period of time comprising well production rates for the well and well operation constraints for the well;
divide the input data features into a training data subset, a validation data subset, and a test data subset;
build a well production model for the well using machine learning based on the training data subset, the validation data subset, and the test data subset;
building the well production model for the well using one of long short-term memory (LSTM) and gated recurrent unit (GRU); and
generate a forecast for the well for a future period of time using the well production model, the forecast comprising a future well production rate for the well including at least one of an oil rate, a gas rate, and a water rate.
8. The system of claim 7 , the at least one processor further to optimize well production for the well based on the forecast for the well for the future period of time.
9. The system of claim 7 , wherein the input data features associated with the well for the at least one particular time stamp during the period of time further comprise temporal features associated with the well including an oil rate for the at least one particular time stamp during the period of time, a gas rate for the at least one particular time stamp during the period of time, and a water rate for the at least one particular time stamp during the period of time, spatial features associated with the well, and spatial images associated with the well.
10. The system of claim 7 , the at least one processor further to process the spatial images associated with the well using a convolutional neural network (CNN).
11. The system of claim 7 , wherein the well is a first well, the at least one processor further to generate a forecast for a future period of time using the well production model for the first well and a second well different from the first well.
12. The system of claim 7 , the at least one processor further to deploy the well production model to a server computing device.
13. The system of claim 7 , the at least one processor further to deploy the well production model to a virtual private cloud.
14. A non-transitory computer-readable medium having instructions stored thereon that, when executed by at least one processor, cause the at least one processor to perform operations comprising:
obtaining an input sequence of input data features associated with a well for at least one time stamp during a period of time, the input data features associated with the well for the at least one particular time stamp during the period of time comprising well production rates for the well and well operation constraints for the well;
dividing the input data features into a training data subset, a validation data subset, and a test data subset;
building a well production model for the well using machine learning based on the training data subset, the validation data subset, and the test data subset;
building the well production model for the well using one of long short-term memory (LSTM) and gated recurrent unit (GRU); and
generating a forecast for the well for a future period of time using the well production model, the forecast comprising a future well production rate for the well including at least one of an oil rate, a gas rate, and a water rate.
15. The non-transitory computer-readable medium of claim 14 , the operations further comprising optimizing well production for the well based on the forecast for the well for the future period of time.
16. The non-transitory computer-readable medium of claim 14 , wherein the input data features associated with the well for the at least one particular time stamp during the period of time further comprise temporal features associated with the well including an oil rate for the at least one particular time stamp during the period of time, a gas rate for the at least one particular time stamp during the period of time, and a water rate for the at least one particular time stamp during the period of time, spatial features associated with the well, and spatial images associated with the well.
17. The non-transitory computer-readable medium of claim 14 , the operations further comprising processing the spatial images associated with the well using a convolutional neural network (CNN).