Bayesian Gas Lift Optimization for Wellbore Production
Solution Overview
Problem
Current artificial gas lift systems lack real-time optimization capabilities for gas lift parameters, leading to suboptimal oil production efficiency due to reliance on static models and manual adjustments.
Innovation Solution
A system utilizing a computing device with a physics-based or machine learning reservoir model to simulate production data and apply Bayesian optimization for real-time control of gas lift parameters, such as injection rate and choke size, to maximize production efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If static models and manual adjustments are used for gas lift control, then system simplicity is maintained, but production efficiency deteriorates due to lack of real-time optimization
Solution Approach 1:
The patent transitions from static gas lift control models to dynamic real-time optimization using Bayesian optimization algorithms. The system continuously updates gas lift parameters (injection rate, choke size) based on real-time production data and reservoir model updates, enabling adaptive control that responds to changing reservoir conditions while maintaining operational efficiency.
Solution Approach 2:
The patent implements a closed-loop feedback system where production data is continuously monitored, reservoir models are updated in real-time, and gas lift parameters are adjusted based on optimization results. This feedback mechanism enables the system to learn from actual production performance and continuously improve oil recovery efficiency through iterative parameter optimization.
2Productivity
If real-time Bayesian optimization is implemented for gas lift parameters, then production efficiency is improved, but computational complexity increases
Solution Approach 1:
The patent pre-generates a comprehensive training dataset by simulating reservoir behavior across a wide range of gas lift parameters and reservoir conditions before actual field operation. This offline pre-computation creates a lookup table and trained machine learning models that enable rapid real-time optimization without requiring heavy computational resources during production, thus resolving the conflict between optimization accuracy and computational complexity.
Solution Approach 2:
The patent uses machine learning models to create simplified copies or surrogates of the complex reservoir simulation models. These trained models approximate reservoir behavior with much lower computational cost, enabling real-time Bayesian optimization to run efficiently on field computers while maintaining accuracy comparable to full physics-based simulations.
3Productivity
If gas injection rate is increased to maximize oil production, then oil recovery improves, but friction in production tubing increases
Solution Approach 1:
The patent simultaneously optimizes multiple gas lift parameters including injection rate, injection pressure, and choke size using Bayesian optimization. By exploring the multi-dimensional parameter space, the system identifies optimal combinations that maximize oil production while minimizing negative effects such as excessive friction. The optimization balances gas injection benefits against flow regime transitions and pressure losses in the production tubing.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system achieves optimized gas lift parameters that enhance production efficiency and profitability by dynamically adjusting gas injection rates and choke sizes based on real-time data, resulting in increased oil production and reduced friction in the production tubing.
Implementation Method 1
Gas bubbles mix with the reservoir fluids, thus reducing the overall density of the mixture and improving lift
Implementation Method 2
decrease the density of the fluid, thereby decreasing the hydrostatic head to allow the reservoir pressure to act more favorably on the oil being lifted
Data Source
AI summary
A system and method for controlling a gas supply to provide gas lift for a production wellbore makes use of Bayesian optimization. A computing device controls a gas supply to inject gas into one or more wellbores. The computing device receives reservoir data associated with a subterranean reservoir to be penetrated by the wellbores and can simulate production using the reservoir data and using a physics-based or machine learning or hybrid physics-based machine learning model for the subterranean reservoir. The production simulation can provide production data. A Bayesian optimization of an objective function of the production data subject to any gas injection constraints can be performed to produce gas lift parameters. The gas lift parameters can be applied to the gas supply to control the injection of gas into the wellbore or wellbores.