Dynamic well construction model
A method for dynamically constructing a well includes receiving inputs that specify characteristics of a proposed well. An indicator is generated for construction of the proposed well using the inputs and a machine learning model trained using data from actual wells, and the indicator is transmitted.
1. A method comprising:
receiving first inputs, from a user via a user interface of a computing device or from a user device via a network interface of the computing device, wherein the first inputs specify at least a geographic location of a proposed well;
displaying input fields on the user device or on a graphical user interface (GUI) of the user interface, wherein the input fields are associated with a machine learning model, of a group of machine learning models associated with a plurality of geographic locations, and wherein the machine learning model is associated with the specified geographic location of the proposed well;
receiving second inputs to the input fields, from the user via the user interface or via the network interface of the computing device, wherein the second inputs specify characteristics of the proposed well, wherein the second inputs include at least a rate of penetration input, a depth input, and a planned drilling distance input, and wherein the second inputs are received as part of an application programming interface call;
transmitting, by the computing device to the user device or to the GUI of the user interface, an indicator with feedback for construction of the proposed well using the second inputs and a machine learning model trained using data from actual wells, wherein the indicator indicates a percentage chance that the construction of the proposed well, based on the specified characteristics of the proposed well, will meet at least one run objective for the proposed well, wherein the indicator further indicates at least one of the second inputs that is identified as causing a reduction of the percentage chance, and wherein transmitting the indicator occurs automatically as a response to the application programming interface call;
receiving revised inputs, in response to the indicator, from the user via the user interface or from the user device via the network interface of the computing device, wherein the revised inputs specify at least one adjusted characteristic of the proposed well, and wherein the revised inputs include the at least one of the second inputs identified as causing the reduction of the percentage chance; and
constructing the proposed well in accordance with the adjusted characteristics.
2. The method of claim 1 , wherein the actual wells comprise wells in the geographic location.
3. The method of claim 1 , wherein the second inputs further include a dogleg severity input.
4. The method of claim 1 , wherein at least one of the second inputs is range restricted.
5. The method of claim 4 , wherein the at least one of the second inputs is range restricted based on the geographic location of the proposed well.
6. The method of claim 1 , further comprising generating a revised indicator based on the revised inputs.
7. The method of claim 1 , wherein the machine learning model is trained using at least one of a decision tree model, gradient boosting, or a gradient boosted decision tree model.
8. The method of claim 1 , wherein the indicator includes contributions of a plurality of the second inputs to a calculation of an overall percentage chance of success for the construction of the proposed well.
9. A system comprising:
a processor;
storage media accessible to the processor;
a user interface or a network interface; and
processor-executable instructions stored in the storage media and executable by the processor to instruct the system to:
receive first inputs, from a user via the user interface or from a user device via the network interface, wherein the first inputs specify at least a geographic location of a proposed well;
display input fields on the user device or a graphical user interface (GUI) of the user interface, wherein the input fields are associated with a machine learning model, of a group of machine learning models associated with a plurality of geographic locations, and wherein the machine learning model is associated with the specified geographic location of the proposed well;
receive second inputs to the input fields, from the user via the user interface or via the network interface, wherein the second inputs specify characteristics of the proposed well, wherein the second inputs include at least a rate of penetration input, a depth input, and a planned drilling distance input, and wherein the second inputs are received as part of an application programming interface call;
transmit, to the user device or to the GUI of the user interface, an indicator for construction of the proposed well using the second inputs and a machine learning model trained using data from actual wells, wherein the indicator indicates a percentage chance that the construction of the proposed well, based on the specified characteristics of the proposed well, will meet at least one run objective for the proposed well, wherein the indicator further indicates at least one of the second inputs that is identified as causing a reduction of the percentage chance, and wherein transmitting the indicator occurs automatically as a response to the application programming interface call;
and
receive revised inputs, in response to the indicator, from the user via the user interface or the network interface, wherein the revised inputs specify at least one adjusted characteristic of the proposed well, and wherein the revised inputs include the at least one of the second inputs identified as causing the reduction of the percentage chance.
10. The system of claim 9 , wherein the actual wells comprise wells in the geographic location.
11. The system of claim 9 , wherein the second inputs further include a dogleg severity input.
12. At least one non-transitory computer-readable storage medium having stored thereon computer-executable instructions that, when executed, instruct a computing system to:
receive first inputs, from a user via a user interface of the computing system or from a user device via a network interface of the computing system, wherein the first inputs specify at least a geographic location of a proposed well;
display input fields on the user device or a graphical user interface (GUI) of the computing system, wherein the input fields are associated with a machine learning model, of a group of machine learning models associated with a plurality of geographic locations, and wherein the machine learning model is associated with the specified geographic location of the proposed well;
receive second inputs to the input fields, from the user via the user interface or via the network interface of the computing system, wherein the second inputs specify characteristics of the proposed well, wherein the second inputs include at least a rate of penetration input, a depth input, and a planned drilling distance input, and wherein the second inputs are received as part of an application programming interface call;
transmit, by the computing system to the user device or to the GUI of the user interface, an indicator with feedback for construction of the proposed well using the second inputs and a machine learning model trained using data from actual wells, wherein the indicator indicates a percentage chance that the construction of the proposed well, based on the specified characteristics of the proposed well, will meet at least one run objective for the proposed well, wherein the indicator further indicates at least one of the second inputs that is identified as causing a reduction of the percentage chance, and wherein transmitting the indicator occurs automatically as a response to the application programming interface call;
and
receive revised inputs, in response to the indicator, from the user via the user interface or from the user device via the network interface of the computing system, wherein the revised inputs specify at least one adjusted characteristic of the proposed well, and wherein the revised inputs include the at least one of the second inputs identified as causing the reduction of the percentage chance.
13. The at least one non-transitory computer-readable storage medium of claim 12 , wherein the actual wells comprise wells in the geographic location.
14. The at least one non-transitory computer-readable storage medium of claim 12 , wherein the second inputs further include a dogleg severity input.