Experimental and numerical study on thermal comfort and energy saving of stratified air distribution systems

Abstract

Stratified air distribution (STRAD) systems have better ventilation effectiveness and are more energy-saving, in comparison with the conventional mixing ventilation systems. By splitting the locations of return and exhaust grilles in a STRAD system, further energy saving potential is attainable. The primary aim of this study is to fill several key knowledge gaps for the design of STRAD systems, including the thermal comfort evaluation, the occupied zone cooling load calculation for the air flow rate determination, the cooling coil load calculation, and the locations of diffusers. The work that is presented here mainly consists of five parts: 1) comparison of two representative thermal comfort models, which are used to evaluate the thermal environments of STRAD systems, by coupling with computational fluid dynamic (CFD) simulations, 2) development of a cooling coil load calculation method and for the first time to conduct an experimental study in a full-scale chamber ventilated by a STRAD system with separate locations of return and exhaust grilles, 3) numerical investigation on the performances of STRAD systems with different heights of return grille location in a small office environment, 4) optimization of thermal comfort and maximization of energy saving for a lecture theatre with terraced floor conditioned by STRAD systems with different diffuser locations, 5) numerical assessments of alternative stratified air distribution designs for a terminal building with large floor area and high ceiling in terms of thermal comfort and energy efficiency . By coupling with CFD simulations, two representative thermal comfort models, the comprehensive UC Berkeley thermal comfort model and the simple ISO Standard 14505 method, are used to evaluate the thermal environments in a small office conditioned by a displacement ventilation system aided with personalized ventilation. The evaluation results indicate that in an overall thermally neutral environment, the ISO Standard 14505 is more sensitive to warming variation, while less sensitive to cooling variation, as compared with the UC Berkeley thermal comfort model. An innovative coupling procedure between the UC Berkeley thermal comfort model and CFD simulations is proposed to simplify operation and improve accuracy. The energy-saving principle for STRAD systems is clearly illustrated. The cooling coil load reduction is associated with but different from the space cooling load reduction. By splitting the locations of return and exhaust grilles, the energy saving potential of the cooling coil can be enhanced. A novel CFD-based cooling coil load calculation method is developed, which can be used to evaluate energy saving of the cooling coil in STRAD systems. Experimental study is conducted to demonstrate that this method is feasible to be adopted when designing a STRAD system. The experimental results also demonstrate that extra energy saving of the cooling coil load is attained by splitting the locations of return and exhaust grilles. The energy-saving capacity is directly proportional to the exhaust air temperature. When maintaining the exhaust grille at ceiling level and decreasing the return grille location from 2.4 m to 1.85 m, the exhaust air temperature is increased in two series of experiments. Correspondingly, energy saving of the cooling coil {439}Qcoil increases from 15.2% to 19.9% of the space cooling load with low supply air velocity, and from 11.5% to 19.5% with relatively high supply air velocity. However, the temperature difference between the head and ankle levels {439}thead-ankle also increases slightly as the return grille lowered from 2.4 m to 1.85 m. A CFD model validation is carried out by using both our own experimental results and the data from a published literature. The validation results reveal that the adopted CFD model is capable of assessing the ventilation performance of STRAD systems. A numerical study is carried out for an office setting, which is hypothetically located at the perimeter zone and ventilated by an under-floor air distribution system. The simulation result shows that decreasing the height of return grille will increase exhaust air temperature and significantly reduce the cooling coil load. As much as 17.0% of the space cooling load is reduced for the cooling coil load when lowering the return grille from ceiling level to 0.3 m above floor level. However, the risk of poor thermal comfort caused by too large temperature difference between the head and foot levels also increased, when the return grille is located close to the floor level. Locating the exhaust grilles at ceiling level and close to external wall and installing the return grilles at the upper boundary of the occupied zone are recommended in practice. It is demonstrated that the purpose to simultaneously achieve satisfied thermal comfort and prominent energy saving can be attained with a STRAD system via deliberate locations of supply outlets and return and exhaust grilles in relation to the heat source locations. The thermal environment in a large space lecture theatre with terraced floor and ventilated by STRAD systems is numerically investigated. The simulation results indicate that the air temperature around the upper human body at the back rows is somewhat higher than that at the front rows because of the upward thermal plumes flowing along the terraced seats. Thermal sensations for the lower body part of occupants are improved and more uniform thermal environment in the occupied zone is created by supplying air from desk-mounted grilles, as compared with supplying air at floor level only. Thermal environment is optimized when the air is supplied from most distributed locations. The highest energy efficiency is achieved when supplying air from floor-level and terraced grilles simultaneously. Furthermore, the influences of the return grilles height on the ventilation performance of STRAD systems are investigated. The simulation results indicate that decreasing the return grille locations worsens the thermal environment slightly. Satisfied thermal environment can also be realized in the large space when locating the return grilles at a low height and into the occupied zone. The results also demonstrate that more energy can be saved by splitting the locations of return and exhaust grilles. As the return grilles height decreased from 4.8 m to 2.3 m, the exhaust air temperature increased and the cooling coil reduction increased largely, from 14.8% to 21.4% of the space cooling load. However, decreasing the height of return grilles leads to a temperature rise of reversed air flowing from upper zone to the occupied zone, and thus an increase of the occupied zone cooling load Qoccupied. Two typical practical air distribution designs, the supplying of air horizontally at floor level or at mid-height (approximately 2 m above the floor level), are numerically investigated in terms of thermal comfort and energy saving in a terminal building with large floor area and high ceiling. The impact of diffuser locations on the ventilation performances is illustrated in detail. The simulation results show that a more uniform distribution of the supply air diffusers, i.e. installing a larger number of supply diffusers with a smaller area each around the occupied zone versus a smaller number with larger area each, can improve the thermal comfort of the occupants and increase the energy saving of the cooling coil, when the air is supplied at floor level and returned at mid-height. Meanwhile, special attentions should be given to the airstream collision regions to avoid un-comfortable local drafts when additional supply diffusers are installed. The locations of return grilles have no obvious effects on the thermal environments of the tested conditions. However, locating the return grilles at the exterior walls blocks free upward convection of warmed air along with the exterior wall, and much more convection heat releases to the occupied zone and contributes to the cooling coil load. When the air is supplied at mid-height and returned at floor level, the thermal environment is more uniform in the occupied zone compared to that when the air is supplied at floor level and returned at mid-height. But a high local draft risk existed for the occupants exposed directly to the cold supply air jets. The enhanced solar radiation intensity strengthens the upward thermal buoyancy flowing along the exterior wall, which is beneficial to the reduction of cooling coil load. The intensified thermal buoyancy flow in the occupied zone uncurled the supplying air streamlines when the air is supplied at mid-height, which is beneficial to reducing the entrainment of warm air from upper zone to the occupied zone. Thus, more energy saving is achieved for the cooling coil. However, the undesired temperature gradient in the region exposed directly to the solar radiation also increased when the air is supplied at floor level. A database of effective cooling load factors for different heat sources in three common types of space is built based on a total of 56 additional simulation cases. The database can be adopted by engineers to calculate the occupied zone cooling load and then determine the required supply airflow rate properly. The developed cooling coil load calculation method is concluded and special design considerations for STRAD systems are also presented.