Thermal Environment, Ventilation Effectiveness, and Infection Risk of Restaurant with Impinging Jet Ventilation

. The basic strategy of the impinging jet ventilation system (IJV) is to keep high air quality within the occupied zone. In IJV, the air supplied from ducts impinges on the floor and spread through the lower level of the room. When the air reaches the heat sources, it rises upward and is exhausted around the ceiling. Most of the previous works about IJV focus on simple room set-ups, whereas the studies focus on realistic set-ups are limited. To investigate the applicability of IJV to a more practical situation, CFD analysis was conducted in a room that simulated a restaurant. In this numerical study, thirty-two seated occupants were distributed in the restaurant hall as the heat and contaminant sources (contaminant was released from their mouth), while a large amount of heat was generated by the equipment in the kitchen. The indoor environment of IJV was compared to that of the mixing ventilation system (MV) in terms of thermal environment, ventilation effectiveness, and infection risk. The results showed that IJV were superior to MV in terms of air conditioning efficiency, pollutant removal efficiency, and infection risk reduction. In conclusion, the results suggest that IJV can improve the indoor environment.


Fig. 1 Plan view (IJV, MV)
because they exhaust air without diffusing pollutants throughout the room. In addition, IJV can be used for both heating and cooling, and have the advantage under cooling mode that it prevents the foot area supercooling in the case of cooling compared with displacement ventilation. Currently IJV research is mostly fundamental, and there are few studies on its application to specific spaces. Therefore, this study examines the feasibility of introducing IJV in a specific space, assuming a restaurant, using CFD analysis. The widely used MV will also be simulated for comparison and evaluation in terms of thermal environment, ventilation efficiency, and infection risk.

Analysis overview
In this study, analysis was conducted on the restaurantlike shown in Fig1. Table 1 shows the analysis method,  and Tables 2-a and 2-b show the boundary conditions. The supply air temperature was set at 21.45 °C for both the MV and IJV. For exhaust air, the total flow rate was set at 3,840 m 3 /h from two exhaust hoods in the kitchen and four ceiling exhaust vents on the seating side for both IJV and MV. The exhaust flow rate of the hoods was set at 3,200 m 3 /h to achieve a surface air velocity of 0.3 m/s or higher at the hood openings, and was exhausted through 0.2 × 0.2 m exhaust vents for each hood. On the seating side, the room air ethausted through 0.3 × 0.3 m exhaust ports at total rate of 640 m 3 /h, equivalent to 20 m 3 /h/person. For the IJV, twenty ducts with an air supply area of 0.3 × 0.3 m were installed with the supply end 0.6 m above the floor, and the air velocity was set to 1.185 m/s. For the MV, eight cuboids of 0.2 × 0.2 × 0.05 m were installed on the ceiling to simulate an anemostat-type vent, referring to the previous study 2) . The air velocity was 1.33 m/s in the vertical downward direction and 2.00 m/s in the horizontal direction.
In the room, thirty-two seated human simulators generated heat at 92 W each, assuming a seated eating aposition 3) . The human simulators were 0.45 × 0.35 × 1.0 m cuboids that located 0.4 m above the floors 4) . The lower surface of the cuboid was insulated, and the heat was generated from all the other surfaces. The human simulator had a heat-generating area of 1.76 m 2 , which was set with reference to previous studies 5) . As shown in Fig 2, 0.05 × 0.05 m mouth opening placed at a height of 1.1 m above the floor surface, and contaminated gas at 32 °C was blown out at 0.01 m/s. The physical properties of the contaminated gas was the same as air, and the ventilation efficiency and infection risk were evaluated by varying the location of the pollutant source. The ventilation efficiency is investigated by releasing the contaminated air from all the human mouth, whereas the infection risk was investigated by setting contaminant source as Human-4 at the counter, Human-17 at the tables, and Human-32 at the back of the room.
The kitchen equipment is shown in Table 2-b. The electric griddle top surface was given a temperature of 250 °C and a convective heat transfer coefficient of 180 W/(m 2 ・K), and the electric fryer top surface was given a heat generation of 740 W, referring to the study by Kondo et al. 6) . The pots (3 pans) on the IH generated heat on the sides at 200 W/piece, and the top surface produced steam at 100 °C at 0.056 m/s. Heat generation from lighting was uniformly applied to the ceiling surface at the location shown in Fig 1 at    and 0.9 for the walls (including the heating elements), all other surfaces, which were assumed to be adiabatic boundaries. In this study, the sensible heat component was given as a boundary condition.

Evaluation Index
To evaluate the difference in contaminant removal performance by ventilation method, the contaminant removal efficiency 7) is calculated for the entire room and within the occupied are. In addition, the infection probability 8) of a person in the room is defined from the perspective of infection control. The contaminant removal effectiveness [-] is an index that expresses the efficiency with which contaminants in the air are removed from a room, and is defined by the following equation.
[ppm] is the exhaust contaminant concentration in the steady state and [ppm] is the room average concentration. The efficiency of contaminant removal in the fully mixed state is = 1 and when > 1, the efficiency is higher than mixed state. The contaminant removal effectiveness in the occupied zone [-] is defined as follows, using the height up to 1.3 m as the occupied zone when seated position, as in (Eq. 1).
[ p p m ] i s t h e a v e r a g e o f c o n t a m i n a n t concentration within the occupied zone. In addition to the above, the probability of infection is calculated based on the Wells-Riley model 8) using the following equation.

Where
[ppm] is the time-averaged viral concentration, [m 3 /h] is the respiratory volume, and [h] is the time that people spent in the room. Here, = 0.6 and = 2 were used, and the local quanta concentration, , was obtained assuming the contaminant gas emission rate from the infected person to be the quanta emission rate of 42 [quanta/h] 8) . Fig 3 and 4 show the temperature distribution in each of the cross sections shown in Fig 1. As for the plume from the kitchen appliances, both MV and IJV were directly caught by the hood and exhausted. In the IJV, temperature stratification was formed in the entire room, and the temperature difference between the head and ankle was about 2 ℃ 9) , thus, the discomfort caused by this temperature difference is not a concern under the present conditions. On the other hand, in the MV, the room air is mixed therefore the temperature distribution is generally uniform. The temperature in the occupied zone when seated was about 24 ℃ in the IJV and about 26 ℃ in the MV, indicating that the IJV had higher cooling efficiency in the occupied zone. Fig 5 and 6 show the air velocity distribution at foot and neck height, respectively. These are the heights at which draft risk is considered to be high. The Building Sanitation Law 10) requires that indoor air velocity to be less than 0.5 m/s. MV meets this requirement for the whole area. On the other hand, IJV satisfies this criterion sufficiently near the seating area, but the air velocity near the ducts is found to be higher. This shows that duct layout planning is important, since air velocities at the seating area may vary depending on the location of the ducts and the blowing velocity.  Table 3 is 1.340 for the IJV and 0.826 for the MV, indicating that the IJV has better contaminant removal efficiency. Moreover the pollutant removal effectiveness within the occupied zone, which is 3.472 for the IJV and 0.963 for the MV, making the trend more pronounced. probability P=1%, 5%, and 10% are also added with white thin dashed, thick dashed, and solid lines respectively. In the case that the occupant at counter seat Human-4 releases contaminant shown in Fig 9, in IJV the contaminants generated from the mouth rose in the plume and went to the hood in the kitchen, where they were directly captured by hood and exhausted, so the contaminant generally did not reach the mouths of other occupants.On the other hand, in the MV, the contaminants spread before being exhausted and reached other occupants. Fig 10 shows the case that contaminant is released from Human-17 at table, where in both IJV and MV, contaminants was spread more when compared to the case that the source was released from the occupant at counter, however the spread of contaminants is more suppressed in IJV. Fig 11 shows the case that the contaminant is released from Human-32, whose seat is at the far end of the room. The normalized concentrations show that in both IJV and MV, the concentrations around the source  is high. This may be due to the fact that although the air inlet for both IJV and MV is located close to Human-32 but the distance to the exhaust port is relatively far, thus the contaminant was assumed to be stagnated at the back of the room. In addition, the spread of contaminants was more suppressed in the IJV compared to the MV. Comparing the above three conditions, the case with contaminant release from Human-32, which is far from the exhaust port had higher contaminant concentrations at the mouth of other occupants, therefore, it is shown to be important to carefully plan the placement of exhaust port.

Risk of Infection
In the MV, the area where the probability of infection exceeds 1% covered about a half of the room regardless of the location of the infected person, whereas in the IJV, this area is suppressed particularly in the cases with contaminant released from Human-4 and Human-32. It is suggested that it is possible to decrease the regions that the infection probability exceeds 1% by installing IJV if compared to MV.

Conclusions
In this study, CFD analysis was performed to investigate the introduction of the IJV in restaurants with kitchens by comparing IJV and MV. IJV showed higher performance than MV in terms of air conditioning efficiency and pollutant removal efficiency in the occupied zone and infection risk. In addition, the importance of planning the placement of air supply and exhaust systems was demonstrated. Future work will include a study on draft risk and the spread of kitchen contaminants.