Research on partially optimized air conditioning system in station space

. There are expectations recently for improving comfort in railway stations. The need to mitigate hot environments in stations in summer is especially strong. The purpose of this study was to investigate the achievement of thermal comfort in summer in railway stations while conserving energy, not by cooling down the whole area but by utilizing localized air conditioning systems. A field study was first conducted in an underground station concourse, measuring the thermal environment, surveying equipment that dissipates heat, and examining occupant density. Under existing conditions, the highest daily interior air temperature was almost the same as that outside the station, while the lowest temperature was higher than that outside; air velocity inside was low. Study results led to the conclusion that a localized air conditioning system utilizing air movement is effective in improving comfort, and to the development of a localized air conditioning system effective for special module concourses in railway stations. A planning method for selecting effective locations for the system is proposed. An evaluation of the system using actual devices was conducted in a mock-up spaces. The CFD simulation in an actual station concourse showed improved thermal comfort using the developed system.


Introduction
In recent years, an increasing number of station buildings have been equipped with air conditioning systems, coupled with commercial facilities on the premises. In addition, as air conditioning systems become common in railcars, expectations for improved station comfort are high, and there is a strong demand for relief from heat, especially during the summer months. On the other hand, in order not to increase the environmental burden, it is necessary to raise energysaving performance in station spaces with a view to decreasing CO2 emissions -simply installing air conditioning equipment is not enough to meet social demands for both comfortable environments and lower CO2 emissions. Nakano's coherent research on environmental adaptation and comfort evaluation in public spaces, including station buildings, has shown that the evaluation of a thermal environment is different from that of air conditioned spaces. (See reference 1.) On the other hand, few studies have been conducted on thermal environment control in station spaces using air conditioning equipment.
The purpose of this study was to develop a method to optimize air conditioning in certain portions of a station, rather than to air condition the entire station uniformly. First, the thermal environment in an actual station was investigated, and comfort levels were evaluated based on the data. Next, after clarifying the procedure for studying localized air conditioned areas, a prototype local air conditioning system was built, taking into consideration equipment installation in the station space. The realized thermal environment was evaluated through a full-scale experiment in simulated space. Finally, the influence on the thermal environment was evaluated by CFD analysis, assuming an actual station space. The following is a report on the results.

Overview of target space
The target space for this study was a station concourse used mainly by passengers. The case study used the underground concourse of a large terminal station in Tokyo, where approximately 560,000 passengers board and disembark trains each day. The target area was the south concourse (excluding stairways, stores, and office areas), which has an area of approximately 3,150m2 (Fig. 1). The space has a low ceiling height of 2.6m with pillars at almost constant intervals (5.6m to 6m). The concourse is adjacent to one private railway line and two subway lines with ticket gates and passageways on the north, west, and south sides. The passageways also serve as underground passageways to surrounding commercial facilities. Users of the space have for some time requested improvements to the thermal environment during the summer months.

Space utilization survey and environment measurements
Space utilization and the thermal environment were measured in summer 2017. In this report, the number of people in the space in the target area was surveyed to determine factors regarding the operation of the existing HVAC equipment and internal heat generation.

Ventilation and air conditioning systems and equipment
Ventilation and air conditioning in the concourse are independent from each other, each with two systems in operation. The ventilation system operates from 4:00 am to 1:00 am when trains are running, with a total outside air supply capacity of 67,500 m 3 /h. The air conditioning system consists of packaged air conditioners (PAC) and fan coil units with a total rated capacity (total heat) of 144 kW (Table 1).

Person count
The number of people in the space was surveyed twice on August 1 (a Tuesday) and 3 (a Thursday) in the morning, noon, and evening (12 times in total). The number of people in the space between the pillars ( Fig.1) was counted by converting the photographs into cubic panoramic images (using Theta Cube).
The count results shown in Fig. 2 indicate that the number of people present in the entire concourse at the maximum moment is approximately 800. The first count of August 3, when the largest number of people were counted, is shown in Fig. 3. Although the total number of passengers was higher in the morning and evening than in the daytime, there were more passengers going to the platform in the shortest distance from each ticket gate, and the number of passengers in the central part of the concourse was lower than that in the daytime in some columns.
Person distribution was expected to be 2.0 persons/m 2 , which is the limit density for walking 3 , although the density was actually about 0.5 persons/m 2 even in high-density areas.

Understanding the Actual Thermal Environment
To evaluate the thermal environment in the target space, the indoor temperature, humidity, and wind velocity were measured. However, since the space is inside a train station, the measuring devices were installed near the ceiling for the safety of passers-by. Measurements were taken from July 21 to August 6, 2017. The recording interval was 1 minute, and the measurement locations are shown in Fig. 4. August 6 (a Sunday) was selected as a representative day to analyze the thermal environment. The daily mean and maximum temperatures at the Tokyo District Meteorological Observatory on the day in question were 29.0°C and 33.5°C, respectively.

Air conditions affected by air conditioning system
Fig. 5 shows representative values of the air blown by the air-conditioner on the day in question. The south ticket gate air conditioner started operation a little after 3:00 a.m., and the temperature was measured at about 30°C with no significant change, although the absolute humidity decreased to 16 g/kg', indicating a change in conditions over time. On the other hand, the air conditioner at the north ticket gate remains in continuous operation for 24 hours/day, with an outlet temperature of about 18°C and an absolute humidity of about 15 g/kg'. The floor-mounted air conditioning unit has an outlet temperature of 14°C and an absolute humidity of about 12 g/kg'. Observatory. The maximum temperature in the measured space rose to about 33°C, almost the same as the outside air temperature, and the minimum temperature was more than 2°C higher than the outside air, with no noticeable time delay. The humidity change shows that from 16:00 to 17:00, the indoor humidity decreased as the outdoor humidity decreased, suggesting that the indoor humidity was considerably affected by the outdoor humidity. On the other hand, considering that the temperature at S-1 is lower than that at the other measurement points, and that the absolute humidity changes significantly over time, this is presumed to be affected by the air conditioner at the south ticket gate, and the same effect is also considered to be present at S-2. The temperature drop at the north gate is remarkable between midnight and 3:00 a.m., perhaps because the air conditioner remains in operation there, as well as because of the effect of the outside night air. The daily variation in the wall surface temperature was about 1°C, and the daily average wall surface temperature was about 31°C.

Analysis of thermal environment in the space
Since the number of measurement points and locations was limited when ascertaining actual conditions, the spatial distribution of temperature and wind velocity in the space was obtained by CFD, and SET* was calculated. Boundary conditions based on actual measurements are shown in Table 4, and CFD was used for steady-state calculations of incompressible fluids using the Boussinesq approximation. The number of computational meshes was about 18 million.

Case study of local air conditioning planning
Before installing local air conditioning in a station space, it is necessary to determine the method for planning the installation area that will be effective. According to the analysis using CFD and the actual conditions of the thermal environment identified in Chapter 2, local air conditioning was found to be effective at the installation site and its surrounding area. Here, the installation site and its surroundings are called "the area," and each part of the entire study space is divided into several areas. The station concourse in question is a space supported by pillars located approximately 6m x 6m apart, with basic components being the spaces between the pillars. The case study was conducted with the space between pillars being one unit of area.
(Step 1) The installation study was conducted based on the dissatisfied rate, i.e., the ratio of people who are thermally dissatisfied with the station space. The dissatisfied rate was obtained based on the literature 1 , which proposes a relationship between SET* and the rate of dissatisfied persons in station spaces.
The dissatisfied rate in the station concourse was calculated using the distribution of SET* shown in Fig.  8.
The number of dissatisfied persons in each area was calculated by multiplying the distribution of the number of persons present in the station concourse (obtained from the actual condition survey shown in Fig. 3 (c)) by the dissatisfied rate distribution. ( Step 2) The number of dissatisfied persons for the entire area is shown in Table 5. The dissatisfied rate for the entire station concourse was 18%, which is lower than the generally accepted value of 20%. On the other hand, looking at the distribution of the number of dissatisfied users, there are areas adjacent to each other where the number of dissatisfied users is high, and the introduction of local air conditioning for these areas will be considered. ( Step 3) Based on the count, the average number of people between pillars is about 8.5. If the allowable value of the dissatisfied rate is 20%, the number of dissatisfied persons between pillars exceeds the allowable value by more than 2 persons in the area. Areas with two or more dissatisfied persons are shown in Fig. 9 as candidate areas for local air conditioning installation.
(Step 4) After considering the conditions at the site and other factors, two areas where local air conditioning is to be installed were selected, as shown in Fig. 10. Fig. 11 shows a comparison between the dissatisfaction rate calculated by CFD simulation of the thermal environment in one case, where two local air conditioning systems were installed, and in the other E3S Web of Conferences 396, 01042 (2023) https://doi.org/10.1051/e3sconf/202339601042 IAQVEC2023 case, before installation. As a result of the installation of local air conditioning, the dissatisfied rate in the entire concourse was reduced from 18% to 15%, and the number of dissatisfied persons was reduced by 26 persons, indicating a significant improvement in comfort.
In addition, the reduction in energy consumption was 26% compared to an overall air conditioning system with SET*<28°C set for the entire station concourse, indicating that the energy-saving effect of local air conditioning was significant.

Local air conditioning design method based on the number of dissatisfied users
Based on the results of this case study, we propose a design method that ranks the areas to be locally air conditioned. Fig. 12 shows the flow of planning and design.
In Step 1, the number of dissatisfied persons is calculated by estimating the number of persons present in the divided areas using existing measurement results and flow line simulations. For each divided area, the dissatisfied rate is arrived at by calculating SET*, based on thermal environment evaluations using CFD simulation measurements, and multiplied by the number of persons in the area to calculate the number of dissatisfied persons in each area. Finally, the total number of dissatisfied persons is summed up to calculate the number of dissatisfied persons in the entire target space.
In Step 2, the allowable number of dissatisfied persons is calculated by multiplying the calculated number of persons present in the entire subject space by the allowable dissatisfied persons rate (= 20%). The total number of dissatisfied persons to be reduced is then calculated by comparing the number of dissatisfied persons in the subject space and the allowable number of dissatisfied persons.
In Step 3, the number of dissatisfied persons in each area and the drop in the number of dissatisfied persons when local air conditioning is applied are calculated, and then the area for local air conditioning is set. The drop in the number of dissatisfied persons when local air conditioning is applied is calculated by finding the amount of reduction in the dissatisfied rate based on SET*, from the specification of the local air conditioning to be applied and multiplying it by the number of persons present in the area. In Step 4, priority is given to the areas with the largest number of dissatisfied persons to be reduced when local air conditioning is applied. According to the priority order, the number of dissatisfied persons to be reduced in each area is summed up, and the priority order is determined according to the excess in the number of dissatisfied persons to be reduced in the space (calculated in Step 2).

Outline of the air conditioning system
Based on the method for appropriately planning and designing local air conditioning, as proposed in Chapter 3, a local air conditioning system was realized. Since the pillars in the target station are aligned at almost constant intervals (5.6m to 6m), the air conditioning system was designed with a 4m x 4m blowout area by combining 2m g g ( p ) x 2m modules, and the height of the system was kept within 0.45m because the ceiling is lowered by 0.7m (Fig. 13). In addition, a station is characterized by the fact that the users' positions are not fixed to specific locations between pillars. Therefore, we developed an air conditioning system that blows air at low velocity from a 4m x 4m ceiling air outlets to improve the airflow sensation in the entire space between pillars. Since it is necessary to reduce the air velocity plane distribution in the target space, the air outlet surface consists of a perforated plate (aperture ratio 54%) and a 0.5m square wind directional plate (0.5m square) which is capable of changing the blowing angle and blowing direction. The thermal environment target was set at SET* 29°C or lower, and the wind speed target was set at 0.2 to 0.5 m/s. 1 The evaluation height was set assuming the height of the head of a typical human body (1.5m above the floor). To remove heat generated by the equipment, a minimum amount of cooling was used (assuming a blowout temperature of 20°C, 1,600 m3/h x 2 units), and cool air from the air conditioner and circulating air from the target space (assuming an ambient temperature of 30°C, 3,600 m3/h x 4 units) were mixed in a chamber above the blowout surface to obtain a total air volume of 17,600 m 3 /h. The total air volume of 17,600 m3/h is discharged from the blow-off surface of the air blowing device (Fig. 14).

Device performance test
The air blowing apparatus performance test was conducted in a simulated space with a ceiling height of 2.6m between 6m x 6.25m spaced pillars, taking into consideration the target space. During the test, the air flow rate from the air conditioner and blower, as well as the blowing and suction temperatures, were measured continuously. Vertical temperature distributions (FL+2.5, 2.2, 1.7, 1.5, 1.1, 0.6, 0.3, 0.1 m) were measured at fixed points. A 6m x 6m flat surface was assumed as the target surface, and wind velocity distribution on the evaluation surface was measured at 144 points.
Testing was adjusted to the design air volume. The temperature difference between the average suction temperature (26.0 °C) and the air conditioner discharge temperature was 11.3 °C, with almost no vertical temperature difference during the test. Maximum wind speed at the evaluation surface was 1.63 m/s, the minimum was 0.1 m/s, and the average was 0.41 m/s. 51 % of the measurement points were within the target wind speed range (Fig. 15).

Evaluation of Thermal Environment
The developed air conditioning system is intended to provide a cool sensation to the human body through regulated air temperature and velocity. To evaluate the experienced thermal environment, a thermal mannequin with a body surface divided into 20 sections was used to measure equivalent temperatures at four points in the target spaces (C5, F3, F6, and H10, in Fig. 15) both when operation was stopped and during operation. Since heat generation inside the space was small at the time of the test and the ambient temperature could not be kept close to that of the actual summer survey, a correction of -14.4°C when there was no wind and -8°C when the air conditioning system was operating was added to 36.4°C on the right side of the control formula (Equation 1) for the average temperature 5 . Ts = 36.4-0.054Ql (1) where Ts is the average surface temperature, and Ql is the heat loss from the subject surface when clothed. was 11.2°C, and the airflow distribution at measurement locations can be judged to be close to that shown in Fig. 15. The temperature at H10 is lower in the vertical direction, while at F6 it is higher (Fig. 16). On the other hand, the wind velocity is higher at F6 and lower at H10 (Fig. 17). The equivalent temperature of the whole human body decreased by about 1.5°C compared to the no-wind condition (Fig. 18). In particular, equivalent temperature tends to decrease more in the upper body. The lowest equivalent temperature for the whole body was observed at F6 (29.4°C), which was 0.8°C lower than that at H10 (30.2°C). This is because of the wind velocity, which lowers the perceived temperature at F6. Compared to the no-wind condition (equivalent wholebody temperature 31.8°C), the temperatures at F6 and H10 were 2.4°C and 1.6°C lower, respectively. It can be judged that the cooling effect of the airflow, rather than the actual temperature, is apparent. Below are the results of CFD analysis of the thermal environment in summer when the local air conditioning system described in Chapter 3 was applied to the underground concourse of the train station.
Airflow characteristics of the local air conditioning system required for CFD analysis were measured in the same simulated space as in Chapter 4. A horizontal surface 50 mm directly below the outlet of a local air conditioning unit installed on the ceiling was used as the measurement target. Anemometers and ultrasonic anemometers were used to measure wind direction and velocity at 64 points.
Turbulence intensity I and turbulence scale L were calculated based on the instantaneous values of the wind speed data at each measurement point for the turbulent flow required as a boundary condition for CFD. Turbulence intensity I and turbulence scale L were calculated as follows = ( ) where ui is the instantaneous wind speed (m/s), is the mean wind speed (m/s), n is the averaged population (= 600), and L0 is the representative length (= 4 m).
The wind speed has a large distribution between 0.2 and 1.3 m/s. The wind direction tends to spread in all directions from the outlet. The values of turbulence intensity and turbulence scale were similar to those of a normal air conditioning outlet, although they were more distributed. Using these values as boundary conditions, a CFD analysis was performed for the underground station concourse.

CFD analysis
The number of meshes used in the calculations was approximately 25 million. Based on the design method proposed in Chapter 3, two cases were used: Case 1, in which the aforementioned local air conditioning system (4m x 4m) was installed at one location (one unit), and Case 2, in which two locations (one unit + two units) were installed (Fig. 19). Fig. 20 shows analysis results of air velocity and temperature at 1.5m above the floor. Case 1, for which the system was installed at only one location, showed an improvement in the thermal environment (increase in air velocity and decrease in temperature) only near the location where the local air conditioning system was installed, while Case 2, for which the system was installed at two locations, showed a wide range of improvement near the south ticket gate. The temperatures of SET* and PPD (percentage of persons dissatisfied) (Reference 1) converted from SET* are shown in Fig. 21. The thermal environment was greatly improved in Case 2 with two local air conditioning units installed compared to Case 1 with only one unit installed, with a maximum SET* of 5°C and an improved PPD rate of 15%, confirming the effectiveness of installing two units.

Conclusion
The purpose of this study was to effectively provide local air conditioning to achieve comfort while at the same time respecting energy-saving considerations during the summer season, rather than to uniformly air condition the entire station.
A survey of the concourse showed that the heat generated by lighting and other equipment was minor, while the station operating hours were long and the distribution of the number of people was about 0.5 persons/m 2 even in densely populated areas. Maximum temperature inside the space was almost the same as the outside air temperature, but the minimum temperature was more than 2°C higher than the outside air, and the time delay was not significant. Air velocity inside the space was low, suggesting that local air conditioning using air currents would be sufficient to improve the thermal environment.
As a method for appropriately planning and designing local air conditioning, we devised a method for ranking areas where local air conditioning is a candidate for installation and developed an air conditioning system that improves the thermal environment inside a station in summer mainly by using the sensation of airflow. Performance evaluation was conducted using a simulated space, and the effect of thermal environment improvement was evaluated using the equivalent temperature of a thermal mannequin. In addition, the effect of this system when applied to an actual underground station concourse was analyzed using CFD to confirm its effectiveness in improving the thermal environment.