This study aims to design and implement a computational fluid dynamics (CFD)-based simulation experiment on residential indoor gas leakage dispersion, exploring the effects of different factors on the dispersion process. As urbanization accelerates and natural gas usage increases, indoor gas leakage accidents pose severe threats to residents' safety. Although on-site experiments are limited by safety concerns and costs, CFD simulations offer a powerful alternative. This experiment is designed to deepen students' understanding of gas leakage dispersion processes, help them master CFD techniques for engineering applications, and enhance their safety awareness and risk assessment skills.

A two-bedroom apartment layout was modeled using ICEM software to simulate realistic residential indoor gas leakage scenarios. ANSYS Fluent was used for numerical simulations, and four scenarios were designed to investigate the effects of ventilation, leakage rate, and gas type on dispersion patterns. The realizable k-ε turbulence model was used, with second-order upwind schemes for spatial discretization and the SIMPLE algorithm for pressure-velocity coupling. The scenarios included the following: (1) unventilated methane leakage at 0.2 g/s, (2) natural ventilation at an air change rate (ACH) of 2 h⁻¹, (3) increased leakage rate of 0.4 g/s, and (4) liquid petroleum gas (LPG) leakage to compare dispersion patterns between different gas types.

The study revealed significant insights into gas dispersion patterns under various conditions. In the unventilated methane leakage scenario (i.e., 0.2 g/s), kitchen concentrations reached the alarm threshold (i.e., 0.7% mass fraction) above 1.7 m height after 50 minutes. By the 5-hour mark, the kitchen ceiling concentration peaked at 2.67%, approaching the lower explosive limit. The introduction of ventilation (i.e., 2 h⁻¹ACH) significantly altered the dispersion dynamics, reducing the kitchen's peak concentration by 62% compared to unventilated conditions after 2 hours of leakage. After 5 hours of ventilated leakage, only the upper part of the kitchen exceeded the alarm threshold. Increasing the leakage rate to 0.4 g/s resulted in a more rapid concentration buildup, with the alarm threshold in the kitchen reached after just 2 minutes. At the 1-hour mark, maximum concentrations were 57% higher than in the lower leakage rate scenario. LPG leakage (i.e., 0.2 g/s) exhibited distinctly different behavior due to its higher density. A significant vertical concentration gradient was observed in the kitchen, with higher concentration at the floor level. The lateral spread of LPG was also slower compared to that of methane. These findings have important implications for gas detector placement. For methane, detectors should be positioned just below the ceiling for fast response. Conversely, for LPG, floor-level detectors will be more efficient.

This CFD-based simulation experiment has provided valuable quantitative insights into indoor gas leakage dispersion under various conditions. The findings underscore the critical role of ventilation, leakage rate, and gas properties in determining dispersion patterns and concentration buildup. This study demonstrates the effectiveness of CFD simulations in analyzing complex indoor environments and has significant implications for residential gas safety management and accident prevention. By integrating such simulation experiments into engineering education, students can enhance their practical skills and prepare for real-world challenges in their future careers. The flexible experiment design allows for the exploration of multiple influencing factors, making it suitable for inquiry-based learning and fostering students’ innovative thinking and research skills.

The development of passive buildings is essential for promoting the green and low-carbon transformation of China's construction industry and achieving the “carbon peaking and carbon neutrality” goals. This transformation urgently requires interdisciplinary talents in relevant professions. However, traditional university teaching methods for passive buildings struggle to meet the demands of cultivating such talents. To address this issue, this study proposed a novel teaching method for passive buildings by integrating virtual simulation and field investigation. This approach aims to comprehensively enhance the quality of talent cultivation in passive building design and operation management, providing new insights for passive building talent education.

The "virtual-reality integration" teaching model was adopted, combining EnergyPlus building energy simulation with on-site investigations and energy monitoring of actual passive buildings. Taking a passive house in Qingdao, China, as a case study, an EnergyPlus simulation model was established. Through field investigations, energy monitoring, mutual verification with simulation results, as well as energy-saving potential analysis and sensitivity factor analysis based on simulations, a hierarchical and integrated innovative, comprehensive experiment was designed.

The new method demonstrated strong operability, with high conformity between simulation results and measured data. The energy-saving effect of the passive building was effectively validated. The mean bias error and root mean square error of the simulated annual energy consumption were 2.64% and 21.04%, respectively. The simulated indoor average temperatures in winter and summer were close to the measured values, meeting accuracy requirements. During the cooling season, air conditioning and fresh air handling equipment dominated energy consumption. Effective passive building design significantly reduced this portion of energy consumption. The annual heating and cooling energy consumption indices of the case passive building were 14.08 kWh/m²·a and 8.37 kWh/m²·a, respectively, achieving 55.26% annual energy savings compared to traditional buildings. The air-conditioning energy-saving rate was 56.95%. Key passive strategies contributing to energy savings included exterior wall insulation, window insulation, and shading systems, accounting for 26.54%, 23.58%, and 23.97% of the total savings, respectively. Additionally, passive buildings ensured a more stable indoor temperature compared to traditional buildings, with an annual indoor average temperature of 20.5±3.0 ℃ and higher thermal comfort levels. Grey relational analysis showed that the average correlation degree ranking of various passive measures was exterior wall (0.853) > window (0.714) > airtightness (0.699) > heat recovery ventilation (0.655) > shading (0.507), with the thermal transmittance of exterior walls having the highest sensitivity to annual energy consumption.

Compared to traditional buildings, passive buildings achieve significant energy savings and provide a more stable indoor environment with higher comfort levels, making them an important direction for future building development. Through this designed experiment, the energy-saving effects of different passive design schemes can be effectively evaluated, providing a basis for further design optimizations. This "virtual-reality integration" experimental teaching model improves students' modeling and simulation analysis abilities, hands-on practical skills, innovative thinking, and scientific literacy. This approach enables the organic integration of theory and practice, providing a new path for cultivating versatile passive building talents.