The Effective Use of the Piston Effect,Natural Cold Sources,and Energy Saving in Beijing Subways

Abstract

With the development of the subway, the energy consumption of subway ventilation and air-conditioning systems is increasing rapidly. To solve this problem, this paper describes the environment of a subway station in Beijing subway line no. 2 from July to September, 2011, and uses numerical methods to quantitatively analyze and evaluate the contributions of the subway piston effect and natural outdoor cold sources to energy saving in the air-conditioning system and improvement of the subway environment. The results show that the piston effect can be used for ventilation in each area of the subway station. During summer, part of the subway air-conditioning season, a natural cold source can introduce a cold capacity of 5,100 GJ to the subway station.

1. Introduction

At the end of 2012, the Beijing subway had 16 lines, a total mileage of 440 km, and daily passenger traffic of 8 million [1]. As a consequence, the required energy consumption of subway air-conditioning and ventilation systems to maintain the environmental quality of subway stations with high passenger flow, high personnel density, and closed space is quite large. According to data from 2003 [2], the ticket revenue of Guangzhou subway line nos. 1 and 2 offset operation consumption, to which the energy consumption of the air-conditioning and ventilation systems made a critical contribution, and of which the fans in the system consume half [3]. Recently, according to statistics from the Beijing Urban Construction Institute [4], the total energy consumption of the air-conditioning and ventilation systems in the Beijing subway in 2011 reached 300,300,000 kWh. Therefore, energy saving and green operation of air-conditioning and ventilation systems in subways are important problems.

Beijing, as a cold region in China, in the summer, its day-night temperature difference is large; both the outdoor temperature and enthalpy in the morning/evening peak hours are obviously lower than the standard values of the subway environment which is described in the Code for Design of Metro (GB50157-2003) [5]: “In summer, the dry-bulb temperature and the relative humidity of the platform are 29°C and 65%, respectively (equivalent to an enthalpy of 70.8 kJ/kg), and that of the concourse are 30°C and 65%, respectively (equivalent to an enthalpy of 74.3 kJ/kg)”; that is, there is an advantage of taking use of the outdoor natural cold source in Beijing subways. Figure 1 shows the weather conditions of Beijing in the summer of 2011; it can be found that although the period from mid-June to mid-August has a weather condition of a high temperature, there are few days of which the air temperature is above 32°C, and a few days of which the air enthalpy is above 78 kJ/kg. The day-night temperature difference is up to 10°C. The hours that the enthalpy of the outdoor air is higher than the design value only accounts for 19.2% of the total hours of the subway air-conditioning season in summer, and these are mainly from late July to mid-August.

Figure 1:

Outdoor air temperature and enthalpy in Beijing (Junuary. 15th–September 15th, 2011).

Figure 2 is an air flow schematic that shows the impact of the metro piston effect on the air flow of a subway station and its ground entrances, analyzed based on aerodynamics. Due to the piston effect, when a train enters a station without SDSs, positive pressure in front of the train pushes some air in the tunnel flow into the platform of the station, which is then exhausted to the outdoors along the following route: platform, concourse, corridor, ground entrances (black arrows in Figure 2). Similarly, when the train leaves the station, the negative pressure behind the train sucks outdoor air into the tunnel in the opposite direction: ground entrances, corridor, concourse, platform (green arrows in Figure 2). The piston effect can introduce fresh air into the station and exhaust air from the tunnel to outside, resulting in a “cross ventilation” effect. Thus, an outdoor natural cold source could be introduced into the subway station, which is of great significance to energy saving and green operation of the subway air-conditioning and ventilation systems.

Figure 2:

Air flow schematic of the piston effect in an island-type station.

Many scholars have studied the influence on the thermal environment and air quality in the subway station caused by the piston effect. Kim and Kim [6] evaluated ventilation performance and found that the optimum location of the vent shaft with respect to maximizing ventilation performance is near the station. Juraeva et al. [7] used ANSYS CFX software to simulate the unsteady flow field in a subway tunnel and found that the air curtain would be better arranged between the natural and mechanical shafts to optimize the ventilation system. Wang et al. [8] summarized the results of temperature measurements for the underground space in the Beijing subway and found that the temperature in the ticket hall is influenced by not only atmospheric temperature but also ventilation induced by piston action due to train movement. With field tests and CFD (Computational Fluid Dynamics), Li [9] analyzed the influence of piston wind on the environment of each area in the subway station. Cao et al. [10] proposed an SDS with adjustable vents, a system that can use the piston process to introduce outside air, thereby reducing ventilation energy consumption by 50% and reducing total annual energy consumption by 30%. These studies mainly focus on the influence of the piston effect on the environment in the subway stations, but there is little or no mention of the effective use of the piston effect.

This paper studies the Beijing subway. Combining field measurements with numerical simulation, this study quantitatively analyzes and evaluates the subway piston effect and the use of natural outside cold sources, which contribute to reduce energy consumption by air-conditioning and ventilation systems and improve air quality in the Beijing subway during the summer. This study also seeks to provide a reference for safe and green methods of Beijing subway system operation.

2. Object of Study

2.1. Overview of the Measured Station

This study describes one subway station in Beijing (subway line no. 2) as the key object of study (Figure 3). This subway station (hereinafter referred to as the measured station) has two ground entrances. To comprehensively determine the influence of the subway piston effect on ventilation in a subway station, this study conducted field measurements during both off-peak (14:00∼16:00) and peak hours (17:00∼19:00) of subway operation to obtain the air velocity and temperature in the two ground entrances A/B.

Figure 3:

Actual civil engineering schematic diagram of the subway station structure.

2.2. Instrumentation and Measured Points

Hot-ball anemometers were used to test the air velocity and temperature (testo-435, China Agent: Testo Ltd., speed range: 0∼+20 m/s, precision: ±0.03 m/s, resolution ratio: 0.01 m/s). Considering the field conditions and flow vortices near the corners, the black points in Figure 4 show the two data collection points, which were both placed in the centers of the corridor areas. The air velocity at each measured point represents the average air velocity of the entire corresponding cross section.

Figure 4:

Arrangement of measured points (data collection points).

3. Methodology

3.1. Estimation of Air Volume through Ground Entrances

To quantitatively analyze the ventilation resulting from the piston effect, the air volume can be calculated as follows:

$G=F{\int }_{t_{\mathrm{1}}}^{t_{\mathrm{2}}}v\hspace{0.17em}\text{d}\tau ,$

(1)

where G is the air volume through the ground entrances, m³; F is the section area of the ground entrances, m²; v is the measured air velocity, m/s; τ is the time; t₁ is the start time of the air flow, s; and t₂ is the end time of the air flow, s.

3.2. Numerical Analysis

Due to the piston effect, outdoor fresh air can be introduced to each area of the station, flowing along the following path: ground entrance, corridor, concourse, platform. According to subway operation conditions (train speed and departure interval during peak, off-peak, and low-ebb hours), this section combines SES and FLUENT to determine the contribution of suction to the air quality in each station area.

The tunnel is analyzed using the SES using the following calculation principle: the length of the tunnel connected to the measured station is much longer than its hydraulic diameter. The air flow model using five conjoined structures, three stations, and two tunnels from SES is shown in Figure 5. The velocities of the four tunnel portals from SES are written as a UDF file to be assigned to FLUENT as boundary conditions. During station operation, crowds of people occupy relatively little space in the platform, concourses, and entrance passageways, and only slightly affect the resistance of air flow; thus, crowds of people are not added to the model in FLUENT.

Figure 5:

Node diagram of the SES numerical simulation physical model.

3.2.1. Basic Governing Equations

The continuity equation, momentum equation, and turbulence model of the standard k-∊ equation (2) are used as the basic governing equations [11] to describe the air flow state in the station:

${\eta }_t=c_{\mu }^{\prime }\rho k^{\mathrm{1}/\mathrm{2}}l=\left(c_{\mu }\rho k^{\mathrm{2}}\right)/∊,$

(2)

where η _t , c_μ′, ρ, k, l, c_μ, and ∊ are the turbulence viscosity coefficient, empirical coefficient, fluid density, turbulent kinetic energy, turbulent length scale, c_μ = c_μ′ · c _D (c _D is the empirical constant), and pulsation kinetic energy dissipation rate, respectively.

3.2.2. Mesh Generation and Boundary Conditions

(1) Mesh Generation. Considering the geometrically complex structure of the simulated station, an unstructured grid is adopted in model computation; in addition, because the complex structure of the stairs connecting the platform and concourses would result in complex nearby air flow, the grid of the stairs is locally refined. Figure 6 shows the local mesh generation of the simulated station. The total grid number is 203,383.

Figure 6:

Mesh generation of the simulated station.

(2) Boundary Conditions. The following are boundary conditions for FLUENT. All model walls are set as nonsliding surface walls. The model has a total of two ground entrances and four tunnels; the two ground entrances are designed with pressure-outlet boundary conditions equal to the atmospheric pressure. The four tunnel portals are set up as velocity-inlet; the specific velocities are obtained by SES simulation. From the SES modeling approach, the node diagram of a model with five conjoined structures, including the simulated station, adjoining stations, and metro tunnels, is shown in Figure 5. Nodes 13 and 95 in Figure 5 are pressure boundaries; the values of both are equal to the atmospheric pressure.

The SES boundary conditions are as follows: (1) the train operation mode is set as explicit and the speed variation of the train running from station I to the measured station and from the measured station to station II are set as measured results; (2) the departure interval is set at 300 s, and the parking time and time difference between up and down trains arriving at the simulated station are set at 30 and 98 s, respectively, both of which are obtained by field measurement in off-peak hours; (3) the calculation time is set at 1500 s.

Figure 7 shows the air velocities of the four tunnel portals in one cycle (300 s) using SES. The labels 53, 56, 153, and 156 correspond with the node numbers of the four tunnel portals in Figure 5. The calculation results obtained by SES simulation are written to the UDF file as the boundary conditions in the FLUENT simulation process.

Figure 7:

Air velocities of the simulated station's four tunnel portals from SES simulation.

3.2.3. Model Solution

The k-∊ two-equation model was used in this study to solve the three-dimensional transient incompressible air flow problem in the simulated station. The UDF method was used to process the air velocities of the simulated station's four tunnel portals in Figure 7 as the boundary conditions for FLUENT. To distinguish between air exhausted from tunnels and fresh air sucked in from outside, the two air types are separately defined as piston-wind and freshair in the simulation process; the physical parameters are set equal. In addition, convergence criteria of the continuity and momentum equations are set at the software default value of 1 × 10⁻³.

3.2.4. Model Validation

Figure 8 compares the numerical and measured air velocities in entrances A and B, in which a positive value indicates that the air is sucked into the station, while negative means that the air is pushed outside. The accuracy of the measured results is certainly affected by air flow instability and the restriction of site conditions during the field measurement process. Some appropriate simplifications and assumptions during the simulation process would cause certain error, yet the numerical results agree generally well with the measured results. In one cycle (300 s), the measured average air flow velocities on the cross section of entrances A and B are 1.13 and 1.14 m/s, respectively; the numerical values obtained by FLUENT simulation are 1.33 and 1.18 m/s, respectively.

Figure 8:

Numerical and measured air velocities on the entrance cross sections.

3.3. Energy Efficiency Evaluation

The air velocity field can be calculated using the numerical method described in Section 3.2. According to diagram in Figure 2, we can assume that, when fresh air is sucked into the subway station, the cooling or heating capacity associated with fresh air can enter the station; similarly, the heating capacity associated with exhausted wind from the tunnel can enter the station. To analyze the contribution of fresh air and exhausted wind from the tunnel to the environment of the station in the summer, this study introduces the following indices.

(1) Heating Capacity from Fresh Air. According to subway operation and the numerical method described in Section 3.2, the air velocity field and fresh air volume in each area of the station can be obtained, and the contribution of fresh air through the ground entrances to the environment of the station can be quantitatively evaluated has seen below:

$Q_f^i=G_f^i\left(h_f-h_n^i\right),$

(3)

where i denotes the areas of the station, including the platform, concourse, and corridor; Q _f ⁱ is the heating capacity from fresh air (Q _f ⁱ > 0 indicates a heating capacity; Q _f ⁱ < 0 indicates a cooling capacity), kJ; G _f ⁱ is the volume of fresh air flow into each station area, m³/s; h _f is the enthalpy of outdoor fresh air, kJ/kg; h _n ⁱ is the design enthalpy of each station area, kJ/kg.

(2) Heating Capacity from Exhausted Wind from the Tunnel. The contribution of the exhausted wind from the tunnel to the environment of the station can be quantitatively evaluated based on the below:

$Q_t^i=G_t^i\left(h_t-h_n^i\right),$

(4)

where Q _t ⁱ is the heating capacity from exhausted wind from the tunnel, kJ; G _t ⁱ is the volume of exhausted wind flow into each station area, m³/s; h _t is the enthalpy of the exhausted wind, kJ/kg.

(3) Total Heating Capacity of the Piston Wind into the Station. Due to the piston effect, the heating capacity into each station area is the sum of (3) and (4), as shown in (5). The total heating capacity into the station is the sum of those in each station area (6). Consider

$Q\left(i\right)=\sum _{\mathrm{0}}^{\tau }\left(Q_f\left(i\right){+Q}_t\left(i\right)\right)$

(5)

$Q=\sum _{\mathrm{0}}^{\tau }{Q}_f\left(i\right)+\sum _{\mathrm{0}}^{\tau }{Q}_t\left(i\right).$

(6)

4. Results and Discussion

4.1. Air Volume in the Ground Entrances

The effects of the piston wind on the air flow characteristics of both station entrances are similar [12]; this section focuses on results measured at entrance A.

Figure 9(a) shows a periodic change in the air flow direction and air temperature on the cross section of entrance A in the off-peak hours. These values are affected by the running direction and time of the train and the positions of other trains in the tunnels. Figure 9(b) reflects a complete periodic change of the air flow characteristics, which lasts approximately 300 s, equivalent to the subway train departure interval.

Figure 9:

Air flow characteristics in off-peak hours at the measured station.

Figure 10(a) shows a periodic change in the air flow direction and air temperature on the cross section of entrance A in peak hours. The departure interval gets shorter, which contributes to shorten the distance between two trains in the tunnel; as a consequence, the comprehensive effect of positive pressure at the front and negative pressure at the back of the trains becomes more pronounced. However, the regular changes of the piston effect can be determined. Figure 10(b) reflects a complete periodic change of the air flow characteristics of entrance A in peak hours, which lasts approximately 120 s, equivalent to the subway train departure interval.

Figure 10:

Air flow characteristics in peak hours at the measured station.

Combining the measured results for off-peak hours and (1), fresh air was introduced into the station through the two ground entrances at a rate of 71,400 m³/h, and air was pushed out from the station through the two ground entrances at a rate of 54,400 m³/h. In peak hours, fresh air was introduced into the station through the two ground entrances at 60,520 m³/h, and air was pushed out from the station through the two ground entrances at 60,520 m³/h. According to the standard of the demand for fresh air, “for a closed system, the fresh air volume supplied for each passenger in every hour should not be less than 12.6 m³ and the total fresh air volume supplied should not be less than 10% of the total air volume supplied,” as described in the Code [5]. In off-peak hours (peak hours), the piston wind caused by the subway train entering and leaving the station for each trip is equivalent to the fresh air supply for 5,660 (4,800) people/h in the subway station.

4.2. Air Volume in Each Area

Based on the result measured in Section 4.1 and the numerical calculation method in Section 3.2, during off-peak hours, the volume variations with time of fresh and exhaust air during one cycle (300 s) in each station area can be determined.

Figure 11 shows that the first half (0∼150 s) of the cycle is the period in which fresh air is introduced, and the latter half (150∼300 s) of the cycle is the period in which inside air is exhausted. As the fresh and exhaust air continue to flow, the mixing of air in all areas of the station is increasingly intense; fresh air constantly mixes with air from the tunnels, and the fresh air volume is constantly consumed but replaced by fresh air in the next cycle.

Figure 11:

Variation of volume of fresh air with time in each station area.

Figure 12 shows that the amount of fresh air through the ground entrances in 300 s is 7,820 m³, which is affected by piston wind; 33% (2,550 m³) contributes to the corridor area, 15% (1,217 m³) is occupied by the concourse, and approximately 52% (4,053 m³) flows into the platform. Figure 12(b) shows that the amount of exhausting air from the tunnel is approximately 4,566 m3, which is affected by piston wind. The air mixes with other nearby air during the process of passing the station; the mixing proportion is 68.4% platform (2,624 m3), 13.3% concourse (509 m3), and 18.4% corridor (704 m3).

Figure 12:

Variation of volume of fresh and piston air with time.

4.3. Energy-Saving Effect Estimation of the Piston Effect

4.3.1. Daily Energy-Saving Effect

Figure 13 shows the estimated heat in each station area caused by the piston effect during the subway operation period, obtained through the method described in Section 3 under the weather conditions of July 27th shown in Figure 2. The temperature of the exhausted air from the tunnel is 35°C, and its enthalpy is 86.2 kJ/kg. As the corridor is an unconditioned zone, the parameters of the air in the corridor are the measured results. According to Figure 13, during morning peak hours (7:00∼9:00), as the enthalpy of the fresh air is relatively lower and the departure interval is shorter, the total cold energy in each area from fresh air is greater than the total heat energy in each area of exhausted air from the tunnel. During off-peak hours (10:00∼17:00), although the enthalpy of the fresh air increases, it is lower than the design enthalpy value in each area, and the cold energy from fresh air is slightly greater than the heat energy of exhausted air from the tunnel. During evening peak hours, the enthalpy of fresh air is close to the design enthalpy value in each area, and the cold energy from fresh air cannot match the heat energy of exhausted air from the tunnel. Using (5), we can calculate that the daily energy-saving effect result from the piston effect is a cold energy of 691 MJ.

Figure 13:

Cold energy into the station from piston wind on July 27th.

Similarly, on July 27th, 2011, the enthalpy of the outdoors is lower than the design enthalpy of the platform during operation hours and is lower than the design enthalpy of the concourse before 19:00 (Figure 14). As a consequence, the cold energy of fresh air can counteract the heat energy of exhausted air from the tunnel, contributing a natural cold energy of 691 MJ to the station.

Figure 14:

Heat capacity into the station over time (July 27th, 2011). Brown indicates the heat from the tunnel, while blue indicates the heat from fresh air.

4.3.2. Energy-Saving Estimation in Summer

From the analysis above, the daily variation of heat energy into each station area caused by the piston effect can be calculated during the subway station's air-conditioning season (Jun. 15th–Sep. 16th), as shown in Figure 15. According to (6), the cold energy into the station is approximately 510 GJ per summer. The results show that the utilization of the piston effect can provide a natural cold source for the Beijing subway station in summer.

Figure 15:

Cooling capacity into the station from piston-wind and outdoor air enthalpy (Jun. 15th∼Sep. 15th).

5. Conclusion

This study focuses on the effective use of the piston effect and natural cold sources in subway stations. A Beijing subway station is the key object of study. By combining field measurements with numerical simulation, it is found that the comprehensive effect of the subway piston effect and natural cold source is reduction of the energy consumption of the air-conditioning and ventilation systems and improvement of the air quality in the Beijing subway in summer. The results are summarized below.

1. From the results of field measurements, due to the piston effect, the air flow characteristics of the two entrances are similar. During off-peak hours, fresh air was introduced into the station through the two ground entrances at a rate of 71,400 m³/h, and air was pushed out from the station through the two ground entrances at 54,400 m³/h. During peak hours, fresh air was introduced into the station through the two ground entrances at 60,520 m³/h, while air was pushed out from the station through the two ground entrances at 60520 m³/h.

2. During a calculation period (300 s, equivalent to a departure interval) in off-peak hours, the total amount of suction fresh air is 7820 m³, of which 2550 m³ (33%) flows into the corridor, 1217 m³ (12%) flows into the concourse, and 4053 m³ (52%) flows into the platform, and the total amount of air flowing from the tunnels is 4556 m³.

3. The results show that, even during one hot summer air-conditioning season in Beijing, a cooling capacity of 5.1 GJ can be introduced into the station.