A combined thermal-resistance-capacity and finite-element model for very fast and accurate short- and medium-term simulations of single U-tube borehole heat exchangers

Basic Information

• Paper ID: 2510.14421
• Title: A combined thermal-resistance-capacity and finite-element model for very fast and accurate short- and medium-term simulations of single U-tube borehole heat exchangers
• Authors: Enzo Zanchini, Francesco Zanchini, Claudia Naldi (University of Bologna)
• Classification: physics.app-ph
• Publication Year: 2024
• Paper Link: https://arxiv.org/abs/2510.14421

Abstract

Accurate design of ground source heat pump systems requires understanding the outlet fluid temperature of borehole heat exchangers (BHEs) over short and long-term periods. This paper focuses on short- and medium-term analysis. Within this timeframe, either 3D finite-element simulations or thermal-resistance-capacity models (TRCM) can be employed. The former offers high accuracy but is computationally time-consuming, while the latter is computationally fast but has limited accuracy due to simplifying assumptions. This paper proposes a novel approach combining the speed of TRCM with the accuracy of finite-element simulation for short- and medium-term simulations of single U-tube borehole heat exchangers. The method uses TRCM to estimate the thermal response of BHEs, then applies interpolation corrections based on a dataset generated from 54 finite-element simulations of different configurations. The model has been implemented as a C++ program and made available in the University of Bologna's open-source database, capable of providing time evolution of inlet/outlet and average fluid temperatures, average BHE surface temperature, 3D and effective borehole thermal resistance within two seconds.

Research Background and Motivation

Problem Description

Ground source heat pump systems, as efficient building heating and cooling technology, require accurate prediction of the thermal response characteristics of borehole heat exchangers. The key challenges are:

1. Contradiction between computational accuracy and efficiency: 3D finite-element simulation is accurate but time-consuming (hours); TRCM is fast but has limited accuracy
2. Short- and medium-term simulation requirements: Existing methods have accuracy or efficiency issues within the time range of 10⁻²·⁵ to 10³ hours
3. Practical engineering application needs: A simulation tool that is both fast and accurate is needed for system design and optimization

Limitations of Existing Methods

1. Analytical solution methods: Replace fluid with heating solids or surfaces, unable to account for heat exchange between upward and downward flows
2. 3D numerical simulation: High accuracy but single simulations require hours, unsuitable for rapid engineering design
3. Existing TRCM models: Adopt horizontally uniform surface temperature (HUST) assumption, leading to underestimation of medium-term thermal response

Research Motivation

Develop a hybrid simulation method combining the speed of TRCM with the accuracy of finite-element simulation to meet practical engineering requirements.

Core Contributions

1. Identified and quantified the impact of HUST assumption: Through comparison of finite-element simulations with and without highly conductive layers, demonstrated that HUST assumption underestimates T_fm, T_in, T_out, R_b3D, and R_beff
2. Developed an improved TRCM model: Employs a layered thermal network including fluid, pipe, grout, and formation nodes, improving simulation accuracy
3. Established a correction coefficient database: Constructed a time-dependent correction coefficient dataset through 54 finite-element simulations of different configurations
4. Implemented an efficient hybrid simulation method: Combining TRCM baseline calculations with interpolation corrections, achieving finite-element simulation accuracy within 2 seconds
5. Provided an open-source simulation tool: C++ program available for fast and accurate BHE short- and medium-term simulation

Methodology Details

Task Definition

Inputs: BHE geometric parameters (length L, radius r_b, pipe spacing s, etc.), thermophysical properties, flow rate, heat load Outputs: Time evolution of inlet/outlet temperatures (T_in, T_out), average fluid temperature (T_fm), BHE surface temperature (T_b), 3D thermal resistance (R_b3D), effective thermal resistance (R_beff) Time range: 10⁻²·⁵ to 10³ hours

Model Architecture

1. Finite-Element Reference Model

• Uses COMSOL Multiphysics heat transfer in solids interface to simulate solid domain heat conduction
• Employs pipe flow module to simulate fluid energy balance
• Mesh independence verification ensures accuracy
• Vertical coordinate scaling (factor 20) improves mesh quality

2. Improved TRCM Model

Layered structure: BHE and formation divided into m horizontal layers, each with height l = L/m

Thermal network: Each layer contains the following nodes

• Fluid nodes: f1 (downward), f2 (upward)
• Pipe nodes: p1, p2
• Grout nodes: gt1, gt2
• Borehole node: b
• Formation nodes: g1, g2, ..., gn

Thermal resistance calculations:

R₁^Δ = 2R_b                    (fluid to borehole wall)
R₁₂^Δ = 4R_b R_a/(4R_b - R_a)   (inter-fluid thermal resistance)
R_conv = 1/(2πr_pi h)           (convective thermal resistance)

Energy balance equations: Transient energy balance equations established for each node, forming a linear system for solution.

3. Correction Coefficient Method

Correction coefficient definition:

C_coeff = R_b3D,COMSOL / R_b3D,TRCM

Correction procedure:

1. TRCM calculates baseline thermal response
2. Obtain correction coefficients through parameter interpolation
3. Correct R_b3D and recalculate temperature field

Technical Innovations

1. Quantitative analysis of HUST impact: Quantifies the impact of HUST assumption on simulation accuracy by introducing highly conductive layers to enforce isothermal boundaries
2. Layered thermal network design: Compared to traditional TRCM, adds pipe and grout nodes for more accurate description of internal BHE heat transfer
3. A posteriori correction strategy: Leverages TRCM's accurate prediction of T_b characteristics, indirectly corrects other parameters through R_b3D correction
4. Multi-parameter interpolation technique: Performs parabolic interpolation based on four key parameters: borehole radius, pipe spacing, grout thermal conductivity, and pipe outer diameter

Experimental Setup

Simulation Configuration

Geometric parameter ranges:

• Borehole radius r_b: 64-76 mm
• Pipe spacing s: 27-47 mm
• Pipe outer diameter r_pe: 16, 20 mm
• Grout thermal conductivity k_gt: 1.0-2.2 W/(m·K)

Fixed parameters:

• BHE length: 100 m
• Burial depth: 1.8 m
• Formation thermal conductivity: 1.8 W/(m·K)
• Heat load: 50 W/m

Dataset Construction

• 54 different BHE parameter combinations
• COMSOL and TRCM comparative simulations for each combination
• Time points: 56 logarithmically equally-spaced points (10⁻²·⁵ to 10³ hours)
• Correction coefficient database generated for interpolation

Validation Cases

Three typical BHE configurations selected for validation:

• BHE 1: Standard configuration, used for correction coefficient calculation
• BHE 2: Configuration requiring parameter interpolation
• BHE 3: Small radius configuration requiring extrapolation

Experimental Results

Main Results

HUST Assumption Impact Analysis

Through comparison of simulations with and without highly conductive layers:

• T_fm, T_in, T_out are underestimated; T_b is almost unaffected
• R_b3D and R_beff are significantly underestimated
• Underestimation increases with time, particularly pronounced in medium-term (>4 hours)

Correction Effectiveness Verification

BHE 1 Results:

• T_fm root mean square error: 0.037°C
• T_in, T_out root mean square error: <0.038°C
• R_beff error at final time: 0.066%

BHE 2 Results (requiring interpolation):

• T_fm root mean square error: 0.051°C
• R_b3D error at final time: 0.13%
• R_beff error at final time: 0.23%

BHE 3 Results (requiring extrapolation):

• T_fm root mean square error: 0.052°C
• R_b3D error at final time: 0.20%
• R_beff error at final time: 0.40%

Computational Efficiency

• Corrected TRCM: Completed within 2 seconds
• COMSOL simulation: Hours
• Accuracy improvement: Achieves finite-element simulation accuracy

Comparative Analysis

Comparison with Pasquier-Marcotte artificial neural network:

• Both methods exhibit medium-term underestimation issues
• This paper's method effectively resolves this through correction coefficients

Related Work

Analytical Solution Methods

• Lamarche & Beauchamp: Short-term analytical solution, but cannot account for inter-fluid heat exchange
• Beier: Analytical model considering fluid energy balance, requires inverse Laplace transform

Numerical Simulation Methods

• Li & Zheng: 3D solid domain + 1D fluid energy balance
• Marcotte & Pasquier: Fully coupled 3D solid and fluid domain simulation

TRCM Methods

• Bauer et al.: Standard triangular thermal resistance network, employing HUST assumption
• Pasquier & Marcotte: Improved TRCM and development of quasi-3D model
• Ruiz-Calvo et al.: Borehole-formation model, avoiding isothermal assumption but with limited accuracy

Machine Learning Methods

• Pasquier et al.: Artificial neural network trained on 15,000 TRCM simulations

Conclusions and Discussion

Main Conclusions

1. HUST assumption is the fundamental cause of TRCM accuracy limitations, leading to systematic underestimation of medium-term thermal response
2. A posteriori correction strategy is effective: Leverages TRCM's advantage in accurately predicting T_b, achieving overall accuracy improvement through R_b3D correction
3. Hybrid method combines speed and accuracy: Achieves finite-element simulation accuracy within 2 seconds, meeting engineering application requirements
4. Interpolation and extrapolation perform well: Maintains high accuracy even for parameters outside the range

Limitations

1. Applicable only to single U-tube BHEs: Dual U-tube or coaxial types require reconstruction of correction database
2. Parameter range restrictions: Parameters outside interpolation range may affect accuracy
3. Physical interpretation of correction coefficients: Lacks deep analysis of relationship between correction coefficients and physical mechanisms
4. Long-term accuracy unverified: Long-term simulation accuracy beyond 1000 hours remains to be verified

Future Directions

1. Extension to other BHE types (dual U-tube, coaxial)
2. Consideration of complex heat transfer mechanisms such as groundwater seepage
3. Coupling with long-term g-function methods for full time-scale simulation
4. Development of adaptive correction strategies to reduce dependence on pre-constructed databases

In-Depth Evaluation

Strengths

1. Accurate problem identification: Thorough analysis of HUST assumption's physical mechanisms and impact degree
2. Ingenious method design: Fully leverages TRCM advantages, compensates for shortcomings through a posteriori correction
3. Comprehensive experimental design: 54 configurations cover main parameter space with sufficient verification
4. Significant engineering value: Open-source C++ program has practical application value
5. Rigorous accuracy verification: Multiple validation cases demonstrate method effectiveness

Weaknesses

1. Insufficient theoretical analysis: Lacks deep explanation of physical meaning of correction coefficients
2. Limited parameter sensitivity analysis: Insufficient analysis of each parameter's impact on correction effectiveness
3. Restricted applicability: Limited to single U-tube BHEs with limited generality
4. Missing long-term verification: Long-term simulation accuracy remains to be verified

Impact

1. Academic contribution: Provides new hybrid modeling approach for BHE simulation field
2. Engineering value: Significantly improves efficiency and accuracy of short- and medium-term BHE simulation
3. Open-source contribution: Provided C++ program promotes method dissemination and application
4. Methodological inspiration: A posteriori correction strategy can be generalized to other engineering simulation problems

Applicable Scenarios

1. Ground source heat pump system design: Rapid and accurate assessment of BHE short- and medium-term thermal performance
2. Thermal response test analysis: Assists in field test data interpretation and parameter identification
3. System optimization research: Provides efficient simulation tool for BHE parameter optimization
4. Teaching and research: Serves as standard tool for BHE heat transfer analysis

References

The paper cites 45 relevant references covering major methods and recent advances in BHE simulation, providing solid theoretical foundation for this research. Key references include:

• Claesson & Javed (2011): g-function analysis method
• Bauer et al. (2011): Standard TRCM method
• Pasquier & Marcotte (2014, 2020): Improved TRCM and machine learning methods
• Zanchini (2024): Comparative analysis of BHE simulation methods

Overall Assessment: This is a high-quality applied physics paper that addresses practical engineering requirements for ground source heat pump BHE simulation by proposing an innovative hybrid modeling method. The paper features in-depth theoretical analysis and comprehensive experimental verification, demonstrating significant academic value and engineering application prospects. The provision of open-source programs further enhances its impact and practical utility.