Determination of hydrocarbon production rates for an unconventional hydrocarbon reservoir

1. A method comprising:

receiving, by a computer system, data from a hydrocarbon reservoir, the data comprising reservoir characterization data, well log data, and hydraulic fracturing data of the hydrocarbon reservoir, the computer system comprising a physics-constrained machine learning model;

generating, by the physics-constrained machine learning model, a predicted hydrocarbon production rate for the hydrocarbon reservoir as a function of time, the physics-constrained machine learning model comprising an artificial neural network and a hydrocarbon fluid flow model of the hydrocarbon reservoir, the generating of the predicted hydrocarbon production rate comprising:

generating, by the artificial neural network, a plurality of parameters of the hydrocarbon fluid flow model based on the data from the hydrocarbon reservoir, the hydrocarbon fluid flow model communicably coupled to the artificial neural network, wherein the plurality of parameters of the hydrocarbon fluid flow model comprise a total hydraulic fracture area of the hydrocarbon reservoir, an average fracture spacing of the hydrocarbon reservoir, a permeability and a porosity of the hydrocarbon reservoir, a hydrocarbon in place (HIP) metric of the hydrocarbon reservoir, a pressure drawdown of the hydrocarbon reservoir, and pressure-volume-temperature (PVT) data for producing fluids in the hydrocarbon reservoir; and

providing, by the hydrocarbon fluid flow model, the predicted hydrocarbon production rate as a function of time based on the plurality of parameters; and

presenting, using a display device of the computer system, the predicted hydrocarbon production rate for the hydrocarbon reservoir as a function of time.

2. The method of claim 1 , wherein the data from the hydrocarbon reservoir comprises at least one of:

a bottomhole pressure of the hydrocarbon reservoir;

a wellhead pressure of the hydrocarbon reservoir;

reservoir fluid pressure-volume-temperature (PVT) data of the hydrocarbon reservoir;

a permeability and a porosity of the hydrocarbon reservoir;

an initial reservoir pore pressure of the hydrocarbon reservoir; or

formation layer data of the hydrocarbon reservoir.

3. The method of claim 1 , wherein the data from the hydrocarbon reservoir comprises at least one of:

a hydraulic fracture geometry of the hydrocarbon reservoir;

an injection rate of the hydrocarbon reservoir; or

a mineral composition and an in situ stress of the hydrocarbon reservoir.

4. The method of claim 1 , wherein the hydrocarbon fluid flow model of the hydrocarbon reservoir represents a one-dimensional hydrocarbon fluid flow from the hydrocarbon reservoir to one or more hydraulic fractures.

5. The method of claim 1 , further comprising:

generating, by the computer system, training data for the physics-constrained machine learning model, the training data comprising:

measured reservoir characterization data and measured hydraulic fracturing data of a plurality of hydrocarbon reservoirs; and

measured hydrocarbon production rates and measured flowback rates of the plurality of hydrocarbon reservoirs; and

training, by the computer system, the physics-constrained machine learning model to generate predicted hydrocarbon production rates as a function of time for the plurality of hydrocarbon reservoirs based on the training data.

6. The method of claim 5 , wherein the training of the physics-constrained machine learning model comprises:

updating, by the computer system, a plurality of weights of the artificial neural network to minimize a difference between the predicted hydrocarbon production for the plurality of hydrocarbon reservoirs and the measured hydrocarbon production rates.

7. The method of claim 1 , further comprising:

determining, using the computer system, whether to drill a production well at a particular geographical location corresponding to the hydrocarbon reservoir, the determining based on the predicted hydrocarbon production rate for the hydrocarbon reservoir as a function of time.

8. A non-transitory computer-readable storage medium storing instructions executable by a computer system, the instructions when executed by the computer system cause the computer system to:

receive data from a hydrocarbon reservoir, the data comprising reservoir characterization data, well log data, and hydraulic fracturing data of the hydrocarbon reservoir, the computer system comprising a physics-constrained machine learning model;

generate, by the physics-constrained machine learning model, a predicted hydrocarbon production rate for the hydrocarbon reservoir as a function of time, the physics-constrained machine learning model comprising an artificial neural network and a hydrocarbon fluid flow model of the hydrocarbon reservoir, the generating of the predicted hydrocarbon production rate comprising:

generating, by the artificial neural network, a plurality of parameters of the hydrocarbon fluid flow model based on the data from the hydrocarbon reservoir, the hydrocarbon fluid flow model communicably coupled to the artificial neural network, wherein the plurality of parameters of the hydrocarbon fluid flow model comprise a total hydraulic fracture area of the hydrocarbon reservoir, an average fracture spacing of the hydrocarbon reservoir, a permeability and a porosity of the hydrocarbon reservoir, a hydrocarbon in place (HIP) metric of the hydrocarbon reservoir, a pressure drawdown of the hydrocarbon reservoir, and pressure-volume-temperature (PVT) data for producing fluids in the hydrocarbon reservoir; and

providing, by the hydrocarbon fluid flow model, the predicted hydrocarbon production rate as a function of time based on the plurality of parameters; and

present, using a display device of the computer system, the predicted hydrocarbon production rate for the hydrocarbon reservoir as a function of time.

9. The non-transitory computer-readable storage medium of claim 8 , wherein the data from the hydrocarbon reservoir comprises at least one of:

a bottomhole pressure of the hydrocarbon reservoir;

a wellhead pressure of the hydrocarbon reservoir;

reservoir fluid pressure-volume-temperature (PVT) data of the hydrocarbon reservoir;

an initial reservoir pore pressure of the hydrocarbon reservoir; or

formation layer data of the hydrocarbon reservoir.

10. The non-transitory computer-readable storage medium of claim 8 , wherein the data from the hydrocarbon reservoir comprises at least one of:

a hydraulic fracture geometry of the hydrocarbon reservoir;

an injection rate of the hydrocarbon reservoir; or

a mineral composition and an in situ stress of the hydrocarbon reservoir.

11. The non-transitory computer-readable storage medium of claim 8 , wherein the hydrocarbon fluid flow model of the hydrocarbon reservoir represents a one-dimensional hydrocarbon fluid flow from the hydrocarbon reservoir to one or more hydraulic fractures.

12. The non-transitory computer-readable storage medium of claim 8 , wherein the instructions further cause the computer system to:

generate training data for the physics-constrained machine learning model, the training data comprising:

measured reservoir characterization data and measured hydraulic fracturing data of a plurality of hydrocarbon reservoirs; and

measured hydrocarbon production rates and measured flowback rates of the plurality of hydrocarbon reservoirs; and

train the physics-constrained machine learning model to generate predicted hydrocarbon production rates as a function of time for the plurality of hydrocarbon reservoirs based on the training data.

13. The non-transitory computer-readable storage medium of claim 12 , wherein the training of the physics-constrained machine learning model comprises:

updating a plurality of weights of the artificial neural network to minimize a difference between the predicted hydrocarbon production for the plurality of hydrocarbon reservoirs and the measured hydrocarbon production rates.

14. The non-transitory computer-readable storage medium of claim 8 , wherein the instructions further cause the computer system to:

determine whether to drill a production well at a particular geographical location corresponding to the hydrocarbon reservoir, the determining based on the predicted hydrocarbon production rate for the hydrocarbon reservoir as a function of time.

15. A computer system comprising:

one or more computer processors; and

a non-transitory computer-readable storage medium storing instructions executable by the one or more computer processors, the instructions when executed by the one or more computer processors cause the one or more computer processors to:

receive data from a hydrocarbon reservoir, the data comprising reservoir characterization data, well log data, and hydraulic fracturing data of the hydrocarbon reservoir, the computer system comprising a physics-constrained machine learning model;

generate, by the physics-constrained machine learning model, a predicted hydrocarbon production rate for the hydrocarbon reservoir as a function of time, the physics-constrained machine learning model comprising an artificial neural network and a hydrocarbon fluid flow model of the hydrocarbon reservoir, the generating of the predicted hydrocarbon production rate comprising:

generating, by the artificial neural network, a plurality of parameters of the hydrocarbon fluid flow model based on the data from the hydrocarbon reservoir, the hydrocarbon fluid flow model communicably coupled to the artificial neural network, wherein the plurality of parameters of the hydrocarbon fluid flow model comprise a total hydraulic fracture area of the hydrocarbon reservoir, an average fracture spacing of the hydrocarbon reservoir, a permeability and a porosity of the hydrocarbon reservoir, a hydrocarbon in place (HIP) metric of the hydrocarbon reservoir, a pressure drawdown of the hydrocarbon reservoir, and pressure-volume-temperature (PVT) data for producing fluids in the hydrocarbon reservoir; and

providing, by the hydrocarbon fluid flow model, the predicted hydrocarbon production rate as a function of time based on the plurality of parameters; and

present, using a display device of the computer system, the predicted hydrocarbon production rate for the hydrocarbon reservoir as a function of time.

16. The computer system of claim 15 , wherein the data from the hydrocarbon reservoir comprises at least one of:

a bottomhole pressure of the hydrocarbon reservoir;

a wellhead pressure of the hydrocarbon reservoir;

reservoir fluid pressure-volume-temperature (PVT) data of the hydrocarbon reservoir;

an initial reservoir pore pressure of the hydrocarbon reservoir; or

formation layer data of the hydrocarbon reservoir.

17. The computer system of claim 15 , wherein the data from the hydrocarbon reservoir comprises at least one of:

a hydraulic fracture geometry of the hydrocarbon reservoir;

an injection rate of the hydrocarbon reservoir; or

a mineral composition and an in situ stress of the hydrocarbon reservoir.

18. The computer system of claim 15 , wherein the hydrocarbon fluid flow model of the hydrocarbon reservoir represents a one-dimensional hydrocarbon fluid flow from the hydrocarbon reservoir to one or more hydraulic fractures.