Dynamic Pore Network Modeling for Two-Phase Flow
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fluid flow modeling techniques for porous media are limited by low resolution and computational expense, leading to inaccurate representation of microscale porosities and fluid behavior, which hinders precise prediction of dynamic two-phase flow in water-wet porous media.
Innovation Solution
A method and apparatus using one or more CPUs to generate and process fracture pore network models, decompose them for parallel processing, and apply boundary conditions to predict dynamic two-phase fluid flow by determining volumetric flow rates and capillary pressures, identifying the highest displacement potential, and performing displacements based on these calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If low-resolution implementations are used to match currently available computational capabilities, then computational expense is reduced, but manufacturing precision (model accuracy) deteriorates
Solution Approach 1:
The pore network model is decomposed into multiple independent domains that can be processed in parallel. Each domain represents a subset of the full pore network, allowing computational tasks to be distributed across multiple CPU cores. This segmentation enables high-resolution modeling of microscale porosities while maintaining computational efficiency through parallel processing.
Solution Approach 2:
Pressure fields are pre-calculated for each possible movement of displacement fronts before the actual fluid flow simulation. This preliminary computation of pressure distributions allows the model to quickly determine displacement potentials without performing expensive real-time calculations during the dynamic simulation, thereby reducing overall computational expense while maintaining accuracy.
2Manufacturing precision
If high-resolution model input is used, then manufacturing precision (model accuracy) is improved, but use of energy (computational resources) increases
Solution Approach 1:
The high-resolution pore network model is divided into multiple domains that can be processed independently in parallel. This segmentation allows the computational resources to be efficiently utilized across multiple CPU cores, reducing the total energy consumption required to handle high-resolution input data while maintaining model accuracy.
Solution Approach 2:
The model calculates pressure fields only for specific possible movements of displacement fronts that are relevant to the current simulation state, rather than computing all possible pressure distributions. This partial action approach reduces computational resource usage while maintaining sufficient model accuracy for predicting dynamic two-phase fluid flow.
3Adaptability or versatility
If dynamic modeling of different fluid flow environments is improved, then adaptability is increased, but device complexity increases
Solution Approach 1:
The pore network model is designed to handle multiple fluid flow environments and displacement scenarios using a unified framework. The same model structure can simulate different two-phase flow conditions (e.g., water-wet and oil-wet conditions, different injection rates, various pressure gradients) without requiring separate specialized models, thereby increasing adaptability while controlling complexity.
Solution Approach 2:
Pressure fields for various possible movements are pre-calculated and stored, allowing the model to quickly adapt to different fluid flow environments during dynamic simulation. This preliminary preparation of pressure distributions enables the model to handle diverse flow conditions without increasing the complexity of the core simulation algorithm.
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 enables accurate, efficient, and repeatable prediction of fluid flow through porous media, reducing computational burdens and improving model accuracy for dynamic pore-scale modeling, particularly in water-wet fractured systems.
Implementation Method 1
generating a highest displacement potential for the set of possible movements using at least a set of capillary pressures
Implementation Method 2
predicting dynamic two-phase fluid flow in a water-wet porous medium
Data Source
AI summary
A method and system for predicting dynamic fluid flow in a water-wet porous medium by one or more central processing units (CPUs), comprising generating a set of possible movements of displacement fronts within a set of pore elements within a pore-network representation of a porous media or rough-walled fracture sample, based on the set of possible movements, generating pressure fields for each of the set of possible movements, based on the pressure fields, determining a highest displacement potential for the set of possible movements, and performing a displacement based on the highest displacement potential.