Artificial intelligence assisted production advisory system and method

A non-transitory computer-readable medium storing instructions that, when executed by at least one processor of a computing system, cause the computing system to perform operations, the operations comprising:

• receiving first data, wherein the first data is not real-time data;
• building a first model of a well based at least partially upon the first data;
• receiving second data;
• building a second model comprising a network of flowlines based at least partially upon the second data, wherein at least one of the flowlines is connected to the well;
• combining the first model and the second model to produce a combined model;
• running a simulation of the combined model to produce simulated results;
• calibrating the combined model to produce a calibrated model based at least in part upon the simulated results;
• updating the calibrated model to produce an updated model;
• receiving real-time data;
• and
• building a machine-learning (ML) model based at least partially upon the real-time data and the updated model via an adaptive learning process comprising weighing one or more model insights and adding one or more failure patterns unrecognizable based on the simulated results, wherein the one or more failure patterns are associated with an electronic submersible pump (ESP) at the well, a gas-oil ratio or a water-to-oil ratio of a fluid flowing out of the well, a well valve, well tubing, well casing, a well pumping module, or any combination thereof;
• and
• generating, via the ML model, a proactive solution in response to predicting one or more potential problems based on one or more data patterns recognized in the real-time data and the one or more failure patterns, wherein the data patterns comprise at least ESP motor temperature, intake pressure, outlet pressure, and motor current amperage (amps), wherein the ML model is configured to provide a probability or confidence of the one or more potential problems or the one or more model insights, and wherein the proactive solution comprises a remedial plan to solve the one or more potential problems.

A computing system comprising:

• one or more processors;
• and
• a memory system comprising one or more non-transitory computer-readable media storing instructions that, when executed by at least one of the one or more processors, cause the computing system to perform operations, the operations comprising:
• receiving first data;
• building a first model of a well based at least partially upon the first data;
• receiving second data;
• building a second model comprising a network of flowlines based at least partially upon the second data;
• combining the first model and the second model to produce a combined model;
• running a simulation of the combined model to produce simulated results;
• calibrating the combined model to produce a calibrated model based at least in part upon the simulated results;
• updating the calibrated model to produce an updated model;
• receiving real-time data;
• building a machine-learning (ML) model based at least partially upon the real-time data and the updated model via an adaptive learning process comprising weighing one or more model insights and adding one or more failure patterns unrecognizable based on the simulated results, wherein the one or more failure patterns are associated with an electronic submersible pump (ESP) at the well, a gas-oil ratio or a water-to-oil ratio of a fluid flowing out of the well, a well valve, well tubing, well casing, a well pumping module, or any combination thereof;
• and
• generating, via the ML model, a proactive solution in response to predicting one or more potential problems based on one or more data patterns recognized in the real-time data and the one or more failure patterns, wherein the data patterns comprise at least ESP motor temperature, intake pressure, outlet pressure, and motor current amperage (amps), wherein the ML model is configured to provide a probability or confidence of the one or more potential problems or the one or more model insights, and wherein the proactive solution comprises a remedial plan to solve the one or more potential problems;
• detecting a first problem using the ML model based upon the one of more potential problems, wherein the first problem is based at least partially upon the real-time data;
• detecting a second problem using the ML model based upon the one of more potential problems, wherein the second problem is based at least partially upon the updated model;
• detecting a third problem using the ML model based upon the one of more potential problems, wherein the third problem is based at least partially upon a pre-defined rule;
• ranking the first problem, the second problem, and the third problem using the ML model;
• identifying a highest-ranking one of the first problem, the second problem, and the third problem;
• and
• determining an adjustment to an operating parameter to address the highest-ranking one of the first problem, the second problem, and the third problem.