logoShaping Tomorrow's Built Environment Today

STBE Featured Paper, May 2019

Using Two Modeling Approaches to Analyze Heat Transfer Processes

From eSociety, May 2019

An article appearing in the May/June edition of Science and Technology for the Built Environment analyzes the heat transfer processes of a ground heat exchanger (GHE) system when adopting two computational fluid dynamics (CFD) modeling approaches.

The article, “Numerical Modeling of Thermal Response of a Ground Heat Exchanger with Single U-shaped Tube,” shows the significance of adopting proper thermal boundary conditions in numerical simulations. This can contribute toward the GHE system performance assessment and optimization.

One author, Wenxin Li with the National Centre for International Research of Low-Carbon and Green Buildings at Chongqing University, explains the article’s significance and challenges.

1. What is the significance of this research?

As an efficient and environmentally friendly approach, the ground source heat pump (GSHP) system adopts shallow geothermal energy to heat or cool the buildings, and its efficiency was largely determined by the thermal performance of ground heat exchangers (GHEs).

The accuracy prediction of GHEs’ performance is important for the system design, evaluation and prediction; however, it is impossible to guarantee that simulations have the exact same calculation conditions and results as the experimental one.

Through a successful indoor experiment, this study found that such simulation uncertainty could be different when adopting two commonly used boundary conditions: one gave accurate results; the other over-predicted by up to 2°C (3.6°F).

Heat transfer process and the uncertainties were analyzed in detail, and this study highlighted the significance of adopting proper thermal boundary conditions in modeling, especially in numerical simulations.

2. Why is it important to explore this topic now?

We normally adopt the commonly used boundary conditions in modeling without thinking about their uncertainties. However, the discrepancy between experimental and modeling results are unavoidable, and even the insignificant one at the very beginning (or every time-step) can be accumulated and enlarged to be unacceptable if the boundary condition of prescribed heating load was used.

It doesn’t mean we cannot use that boundary condition for simulation. It only reminds us to treat the uncertainties carefully when using this approach. Moreover, since these two approaches cannot guarantee the same prediction accuracy, the boundary conditions used for long-term predictions should be consistent with the one used for validation.

3. What lessons, facts and/or guidance can an engineer working in the field take away from this research?

This work focused more on the modeling methods, and it recommends paying attention to the selections of proper boundary conditions in both numerical and analytical simulations. The approach of inputting recorded inlet temperature gave accurate results, while the one approach with prescribed heating load gave unacceptable results. However, the latter is more frequently used because it can be obtained in advance.

Thus, maybe the latter method can be corrected with a correlation coefficient calculated by comparing these approaches, which may largely improve the accuracy of experiments with unavoidable uncertainties for industrial purposes.

4. How can this research further the industry’s knowledge on this topic?

Experimental data from laboratory and field tests are important for system validation and design.

For a short-term prediction of a well-controlled laboratory device, a large discrepancy was found in predictions with different boundary conditions, not to mention the experiments affected by other unavoidable factors. Therefore, the boundary conditions of the modeling methods should be properly addressed.

On the other hand, since the experimental uncertainties are complex, the modified boundary condition considering the uncertainties may provide an alternative approach to get accurate predictions in modeling.

5. Were there any surprises or unforeseen challenges for you when preparing this research?

This work aims to validate the experimental data at first; however, when we used two kinds of the most commonly used boundary conditions for modeling, we got different results but didn’t know the reasons (e.g., were they experimental or calculation uncertainties?). It’s so weird, but interesting, so we did further analysis of the heat flux and thermal distributions to find the reasons.