How do you choose the best multiphase flow models for your reservoir and fluid?

1 Reservoir and fluid properties

The first step is to collect and analyze the relevant reservoir and fluid properties, such as porosity, permeability, pressure, temperature, saturation, viscosity, density, and composition. These properties determine the flow regimes, pressure drops, phase behavior, and transport phenomena of the multiphase flow. You need to use reliable laboratory measurements, field data, and correlations to estimate these properties and their variations over time and space.

2 Flow models classification

The second step is to understand the different types of multiphase flow models and how they are classified. Generally, there are three main categories: empirical, mechanistic, and numerical models. Empirical models are based on experimental data and correlations, and they are simple and fast to use, but they may not be applicable to different conditions or systems. Mechanistic models are based on physical principles and equations, and they can capture more details and phenomena, but they may require more input data and computational resources. Numerical models are based on discretizing and solving the governing equations of the multiphase flow, and they can handle complex geometries and interactions, but they may be very time-consuming and sensitive to the grid size and boundary conditions.

3 Model selection criteria

The third step is to define the criteria for selecting the best multiphase flow models for your reservoir and fluid. Accuracy, robustness, simplicity, efficiency, and flexibility are all important factors to consider. Accuracy measures how well the model matches observed or expected results, whereas robustness refers to its stability and reliability under different scenarios and uncertainties. Simplicity is important for how easy and intuitive the model is to use and understand, whereas efficiency relates to how fast and economical it is to run and maintain. Flexibility looks at how adaptable and customizable the model is to different situations and objectives. Weighing the pros and cons of each model according to these criteria and your specific needs and preferences is essential.

4 Model validation and verification

The fourth step is to validate and verify the chosen multiphase flow models for your reservoir and fluid. Validation is the process of comparing the model predictions with the actual or experimental data to assess the accuracy and applicability of the model. Verification is the process of checking the model implementation and assumptions to ensure the correctness and consistency of the model. You need to use appropriate methods and metrics to perform validation and verification, such as graphical plots, statistical tests, sensitivity analysis, and error estimation.

5 Model sensitivity and uncertainty

The fifth step is to evaluate the sensitivity and uncertainty of the multiphase flow models for your reservoir and fluid. Sensitivity is the measure of how much the model output changes with respect to a change in the model input or parameter. Uncertainty is the measure of how much the model output deviates from the true or expected value due to the lack of knowledge or information. You need to use suitable techniques and tools to conduct sensitivity and uncertainty analysis, such as Monte Carlo simulation, Latin hypercube sampling, response surface method, and probabilistic risk assessment.

6 Model optimization and improvement

The sixth step is to optimize and improve the multiphase flow models for your reservoir and fluid. Optimization is the process of finding the optimal values of the model input or parameter that maximize or minimize a certain objective function, such as production rate, recovery factor, or net present value. Improvement is the process of modifying or updating the model to enhance its performance, accuracy, or applicability. You need to use effective methods and algorithms to carry out optimization and improvement, such as gradient-based, evolutionary, or hybrid methods, and incorporate feedback and learning mechanisms to refine the model over time.

7 Here’s what else to consider

This is a space to share examples, stories, or insights that don’t fit into any of the previous sections. What else would you like to add?