Proxy Models for Wellbore Material Property Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In wellbore environments, traditional methods for determining material properties using ultrasonic pulse echo face challenges due to difficulties in mapping ultrasonic wave responses to specific properties, especially in conditions like extreme temperatures and pressures, and are hindered by resource-intensive physics-based numerical simulations that are costly and time-consuming.
Innovation Solution
The implementation of proxy models, specifically artificial intelligence and machine learning models such as Fourier Neural Operators, U-Net networks, and Physics-Informed Neural Networks, to efficiently map material properties to ultrasonic wave responses, reducing computational burden and enabling real-time predictions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If physics-based numerical simulations are used to determine material properties, then measurement precision is improved, but productivity deteriorates due to resource-intensive computations causing expensive delays
Solution Approach 1:
The patent creates a proxy model that copies the input-output behavior of complex physics-based numerical simulations. Instead of running resource-intensive simulations, the system trains a machine learning model on simulation data to replicate their results. This allows rapid prediction of material properties without repeatedly executing full physics simulations, thus maintaining measurement precision while dramatically improving productivity.
Solution Approach 2:
The system performs preliminary action by pre-computing physics-based simulations to train the proxy model offline. During actual wellbore operations, the already-trained model provides instant predictions without requiring real-time computational resources. This shifts the computational burden to a preliminary training phase, enabling fast deployment in time-critical drilling operations.
2Device complexity
If traditional mapping methods are used to relate ultrasonic wave responses to material properties, then device complexity is reduced, but measurement precision deteriorates in extreme downhole conditions
Solution Approach 1:
The patent introduces an intermediary layer - the proxy model - that mediates between ultrasonic wave response measurements and material property determinations. This machine learning model learns complex, non-linear relationships from training data and acts as a sophisticated translator between the two domains. The intermediary handles the complexity of extreme condition physics internally, allowing the overall system to remain relatively simple while achieving high measurement precision.
Solution Approach 2:
The system adapts to extreme downhole conditions by changing the parameters and relationships within the proxy model based on training data from various conditions. The model learns how material properties and ultrasonic responses change under different temperatures, pressures, and formation characteristics. This allows the system to maintain measurement precision across varying downhole conditions without increasing physical device complexity.
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
This approach allows for accurate, efficient, and real-time estimation of material properties in wellbore environments, replacing traditional numerical simulations with faster and more cost-effective AI/ML models that provide broader case coverage and immediate on-site predictions.
Implementation Method 1
an ultrasonic transceiver can transmit an ultrasonic wave through a wellbore system
Implementation Method 2
receive an ultrasonic wave response from the transmitted ultrasonic wave
Data Source
AI summary
Described herein are systems and techniques for predicting sample characteristics in a wellbore. An example method can include determining a set of values of estimated characteristics of a sample in a wellbore; determining, via a proxy model, a predicted ultrasonic wave response corresponding to the set of values of the estimated characteristics of the sample; based on a comparison of the predicted ultrasonic wave response with a measured ultrasonic wave response associated with the sample, determining an error associated with the predicted ultrasonic wave response; determining whether the error associated with the predicted ultrasonic wave response is below a threshold; and determining whether to update the set of values of the estimated characteristics of the sample based on determining whether the error is below the threshold.