🔬 physics

A Streamline-Based Local Thermal Non-Equilibrium Framework for Heat Transport in Subsurface Porous Media

This study introduces a streamline-based local thermal non-equilibrium (LTNE) framework for subsurface heat transport that decouples fluid advection from multidimensional conduction and interphase exchange, demonstrating through benchmarks and heterogeneous simulations that significant local temperature differences between fluid and rock can persist even when macroscopic thermal responses appear near equilibrium.

Original authors: Mohammad Nezam Uddin, Yin Feng, Boyun Guo

Published 2026-07-07

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Mohammad Nezam Uddin, Yin Feng, Boyun Guo

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: The "Hot Rock, Cold Water" Problem

Imagine you are trying to cool down a giant, hot brick oven (the underground rock) by pumping cold water through it. This is how geothermal energy works: we pump cold water in, it picks up heat from the rock, and we pump hot water out to generate electricity.

For a long time, engineers have used a simple rule of thumb to predict how this works: The "Instant Hug" Theory.
They assumed that as soon as the cold water touches the hot rock, they instantly become the same temperature. It's like hugging someone; the theory says you instantly share the exact same body heat. This makes the math easy, but in reality, it takes time for heat to move from the rock into the water.

The New Discovery: The "Slow Handshake"

This paper introduces a new, more realistic way to look at the problem, called Local Thermal Non-Equilibrium (LTNE).

Instead of an "instant hug," the authors suggest a "slow handshake."

The Reality: When cold water rushes past hot rock, the water stays cold for a while, and the rock stays hot for a while. They don't instantly match temperatures.
The Problem with the Old Way: If you assume they match instantly, you might think the water coming out of the ground will cool down faster than it actually does. This could lead to bad guesses about how long a geothermal plant will last or how much energy it can produce.

How They Solved It: The "River Map" Method

The authors created a new computer model to track this "slow handshake." To do this, they used a clever trick called Streamline Simulation.

Imagine the underground rock is a giant, messy maze.

The Old Way (Grid): Imagine trying to track the water by looking at a giant chessboard. You check every single square to see where the water is. It's accurate but very slow and computationally heavy.
The New Way (Streamlines): Instead of a chessboard, imagine drawing invisible rivers (streamlines) that show exactly where the water is flowing. The computer then just follows these rivers like a train on a track.

The Magic Trick:
The authors split the problem into two parts that happen at the same time:

The River Ride: They calculate how the cold water moves along the "rivers" (streamlines). This is fast and easy.
The Handshake: They calculate how much heat the water steals from the rock as it passes by. This happens on the "chessboard" (the background grid) to account for the rock's heat storage.

By separating the "moving water" from the "staying rock," they can run the simulation much faster while still being very accurate.

What They Found (The Results)

1. The "Speed Test" (Verification)
They tested their new model against known math problems.

Result: Their new "River Map" method was incredibly accurate. It matched the perfect mathematical solutions almost perfectly (99.9% accuracy), even better than the old, slower methods.
The "Equilibrium" Check: They proved that if the heat exchange is super fast (like a very strong hug), their new model naturally turns into the old, simple model. This proves their new math is correct because it includes the old math as a special case.

2. The "Real World" Test (The Heterogeneous Reservoir)
They simulated a complex underground system with "fast lanes" (high permeability channels) and "slow lanes" (low permeability areas).

The Surprise: Even though the water coming out of the ground looked mostly the same whether they used the old "Instant Hug" model or the new "Slow Handshake" model, the inside of the system was very different.
The Temperature Gap: In the new model, the water and the rock stayed at very different temperatures for a long time.
- In the simple model, the difference was tiny (less than half a degree).
- In the new model, the difference was huge (up to 7.5 degrees).
The Output: Because the water stayed colder longer (it hadn't "hugged" the rock enough yet), the water coming out of the ground was 13 degrees hotter in the new model than the old model predicted after 400 days.

The Main Takeaway

The paper concludes that while the "Instant Hug" model (LTE) is okay for predicting the general temperature of water coming out of a well, it fails to see what is happening locally.

The Analogy:
Think of a crowded party.

The Old Model assumes everyone instantly agrees on the room temperature.
The New Model realizes that people in the corner (the rock) might still be sweating while the people rushing through the door (the water) are freezing. Even if the average temperature of the room looks normal, the people in the corner are experiencing a very different reality.

Why it matters:
If you are designing a geothermal plant, knowing that the rock is still much hotter than the water (even if the water coming out looks okay) tells you that there is still a lot of heat energy left in the ground that hasn't been used yet. The new model helps engineers see this hidden energy.

Summary of Claims

Method: They built a fast, accurate computer model that tracks water and rock temperatures separately.
Accuracy: It matches known math perfectly and is more precise than older grid-based methods.
Key Finding: In complex underground systems, the water and rock can have a large temperature difference (up to 7.5°C) even when the water coming out of the ground doesn't change much.
Limitation: The paper focuses on the physics of heat transport in porous media (like sand or rock) and does not discuss specific clinical uses or future applications beyond what is needed for geothermal and oil recovery modeling.

Technical Summary: A Streamline-Based Local Thermal Non-Equilibrium Framework for Heat Transport in Subsurface Porous Media

Problem Statement
Heat transport in subsurface porous media is critical for applications such as geothermal energy extraction, thermal enhanced oil recovery (EOR), and aquifer thermal energy storage. Conventional reservoir-scale simulations typically rely on the Local Thermal Equilibrium (LTE) assumption, which posits that fluid and rock phases within a representative elementary volume share a single temperature. While computationally efficient, LTE assumes rapid thermal equilibration between phases. This assumption may fail in scenarios where fluid residence times are short relative to interphase heat exchange, such as near injection wells, along preferential flow paths, in fractured formations, or under large temperature contrasts. In these conditions, Local Thermal Non-Equilibrium (LTNE) effects—where fluid and rock sustain distinct temperatures—can significantly alter thermal-front propagation and production temperature predictions. However, existing LTNE models are often limited to analytical solutions for idealized geometries or computationally intensive full-domain grid-based numerical models that do not explicitly exploit the flow-path structure of advection-dominated systems.

Methodology
The authors developed a streamline-based LTNE framework designed to capture separate fluid and rock temperatures while maintaining the computational efficiency of streamline simulation. The methodology integrates the following components:

Common Numerical Foundation: Both LTE and LTNE models utilize the same underlying pressure field, streamline geometry, mapping operators, and operator-splitting workflow. This ensures a direct, apples-to-apples comparison between the two formulations.
Operator Splitting: The transport process is decomposed into three distinct steps:
- Streamline Advection: The dominant flow-aligned transport is solved along regularized streamlines in a time-of-flight (TOF) coordinate. In the LTNE model, only the mobile fluid temperature ( $T_f$ ) undergoes advection, moving at the full interstitial velocity ( $\beta = 1$ ), whereas the rock temperature ( $T_r$ ) is stationary.
- Longitudinal Conduction: Heat conduction along the streamline direction is solved on the regularized TOF grid.
- Grid-Based Updates: Multidimensional conduction and the local fluid-rock heat exchange are handled on the background Eulerian grid.
Analytical Exchange Update: A key feature of the LTNE implementation is the use of an exact, conservative analytical solution to update the coupled fluid and rock temperatures at each grid cell during the exchange sub-step. This avoids numerical stiffness often associated with large exchange coefficients and ensures local conservation of energy.
Heat Transfer Coefficient ( $h_a$ ): The framework allows the volumetric fluid-rock heat-transfer coefficient to be specified as either a uniform parameter or a spatially variable field. For heterogeneous cases, $h_a$ is parameterized based on local permeability, linking high-permeability zones to smaller matrix blocks and stronger exchange.

Key Contributions

First Streamline LTNE Framework: To the authors' knowledge, this is the first report of a streamline-based LTNE framework for subsurface heat transport, extending the efficiency of streamline methods to two-temperature systems.
Decomposition of Physics: The model physically distinguishes between mobile-fluid advection (solved in 1D TOF space) and stationary-rock heat storage/exchange (solved on the background grid), preserving physical distinctions while retaining computational advantages.
Exact Exchange Solver: The implementation of an analytical, non-iterative exchange update ensures unconditional stability for the local exchange subproblem, even with high exchange coefficients.

Results
The framework was verified against analytical benchmarks and applied to a heterogeneous reservoir model (SPE10 Layer 36):

1D Verification:
- LTE: The streamline solution matched the Ogata–Banks analytical solution with an $R^2$ of 0.9928, outperforming a finite-difference method (FDM) on the same grid ( $R^2 = 0.9669$ ).
- LTNE: The fluid and rock temperature solutions matched the Schumann analytical solution with $R^2 = 0.9999$ , significantly outperforming the FDM ( $R^2 \approx 0.9975$ ).
- Convergence: As the volumetric exchange coefficient ( $h_a$ ) increased from 10 to 100 W m⁻³ °C⁻¹, the LTNE heat-capacity-weighted mean temperature converged to the LTE solution, with $R^2$ increasing from 0.9442 to 0.9996.
Heterogeneous Reservoir Case (500-day simulation):
- Thermal Breakthrough: Breakthrough occurred at approximately 200 days in a quarter-five-spot configuration.
- Production Temperature: The LTNE model predicted produced-fluid temperatures up to 13 °C higher than the LTE model near 400 days. This discrepancy arises because LTE forces instantaneous equilibration, whereas LTNE allows the rock to retain heat and slowly transfer it to the fluid.
- Spatial Variability: Introducing a spatially variable $h_a$ (based on permeability) had a negligible effect on the macroscopic production temperature curve compared to the uniform $h_a$ case. However, it significantly impacted local phase disequilibrium. The maximum fluid-rock temperature difference at 500 days increased from 0.43 °C (uniform case) to 7.57 °C (variable case), an eighteen-fold increase.

Significance and Claims
The paper claims that while macroscopic thermal responses (such as production temperature curves) may appear near equilibrium, substantial local thermal non-equilibrium can persist, particularly in heterogeneous systems with spatially variable exchange properties. The study demonstrates that a uniform exchange coefficient may accurately reproduce macroscopic production data while failing to represent the magnitude and spatial distribution of local temperature gaps. The proposed framework provides a computationally efficient tool to determine when LTE is adequate and when the retention of two-temperature effects is necessary for accurate subsurface heat-transport simulation.