Numerical Investigation on the Influence of Operation Mode of the Air-Conditioning and Oxygen Supply System on Energy Consumption of Plateau Train

Abstract

Problems of low temperature, low pressure, low oxygen, and high carbon dioxide (CO₂) concentration in the air-conditioning (AC) trains of the Qinghai-Tibet railway can affect the health and comfort of passengers and cause altitude sickness. When the Qinghai-Tibet railway train runs in high-altitude areas, it is necessary to supply oxygen and introduce fresh air to meet the limited oxygen (O₂) and CO₂ partial pressures (PPs) in the carriage. In this study, a numerical analysis of the correlation between the CO₂ PP and O₂ PP in AC trains and the air supply parameters (ASPs), fresh air volume (FAV), and oxygen supply volume (OSV) along the Qinghai-Tibet line in summer was conducted. The results show that the influence of the FAV on the energy consumption of the air-conditioning system (ACS) in different running areas is inconsistent, whereas the influence of the oxygen supply system (OSS) on energy consumption is significant. During the oxygen supply period, the FAV had the opposite effect on energy consumption of ACS and OSS. The energy consumption of the OSS was approximately five times that of the ACS. By studying the correlation between the internal environment and ASP of trains running in different regions, the operation modes of ACS and OSS can be reasonably set, which can effectively reduce energy consumption by 20–50%.

1. Introduction

The Qinghai-Tibet plateau has special environmental conditions, such as low air pressure, low temperature, low oxygen (O₂) content, and strong radiation [1]. Therefore, except for the air-conditioning system (ACS) required on trains on plains, trains on plateaus should be equipped with an oxygen supply system (OSS) to adapt to the environmental conditions along the Qinghai-Tibet line to maintain comfort and air quality [2,3]. Researchers have suggested that the mode of increasing O₂ through pressurisation in the aerospace field can also be used in the Qinghai-Tibet Railway train. However, the actual sealing conditions are not applicable for current trains [4]. At present, a diffusion O₂ supply combined with a distributed O₂ supply has become the best combination of O₂ supply modes. The lumped parameter method was used to simulate and calculate the O₂ content in a train along the Qinghai-Tibet Railway, and the curve of the minimum oxygen supply volume (OSV) along the line was obtained when the O₂ partial pressure (PP) was maintained at 13.7 kPa under the diffusion O₂ supply mode [5]. Chen [6] numerically analysed the effects of diffusion and distributed O₂ supplies under constant O₂ supply parameters. The results revealed that the combined O₂ supply of the two methods maximised the O₂ content in the passenger breathing zone. However, there was a restrictive relationship between the fresh air volume (FAV) and OSV of trains running in high-altitude areas. Wang et al. [7] calculated the FAV in different areas under the diffusion O₂ supply mode and provided a new idea to solve the restrictive relationship between fresh air and OSV, by controlling the concentration of pollutants and O₂ content in the carriage. When the Qinghai-Tibet train ran along the railway at different times, researchers calculated the theoretical relationship between the FAV and carbon dioxide (CO₂) concentration using the lumped parameter method and obtained the dynamic variation value of the minimum FAV while ensuring that the CO₂ concentration did not exceed the limit value [8]. Zhu et al. [9] studied the O₂ supply characteristics of a circular O₂ inlet for local O₂ diffusion in a high-altitude and low-pressure environment and found that the oxygen-enriched area formed by the diffusion of double O₂ inlets under the same O₂ flow was smaller than that of a single O₂ inlet. Jiang et al. [10] monitored the O₂ supply status of plateau trains and found that their O₂ supply equipment was inefficient. Therefore, the oxygen O₂ supply scheme for plateau trains should be optimised. Liu et al. [11] used questionnaire surveys and air quality detectors to analyse the effects of various air indices on human comfort in a train carriage on the Qinghai-Tibet railway line. The results showed that the temperature in the carriage did not meet the standard [12], and the CO₂ content was slightly higher at some stations. Therefore, the air circulation system of the train should be improved. Chen et al. [13] revised the evaluation index of passenger thermal comfort in plateau trains and inferred that atmospheric pressure and seat position were the main factors affecting the thermal comfort of passengers in carriages. Guo [14] studied the characteristics of the comprehensive comfort of the human body under normal and low air pressures and established a comprehensive comfort evaluation model under 80 kPa. The influencing factors and comfort zones of the evaluation model were obtained, and the expression of comprehensive human comfort under low pressure was explored. Wang et al. [15] numerically analysed the changes in the speed and temperature fields in a carriage along the Qinghai-Tibet railway in winter and obtained the distribution characteristics of various parameters in the carriage under different temperature conditions. Current research on the OSS and ACS of plateau trains has mainly focused on the O₂ supply method [9,16,17,18], environmental quality monitoring [10,11], human comfort in low-pressure environments [11,13,14,19], and thermal environment parameters in trains [15,20]. However, there have been no reports on the energy consumption of trains on the plateau. This study considered the CO₂ and O₂ PP as the control conditions of the internal environment of a train. Considering the attenuation and delay effects of a train enclosure structure on the processes of solar radiation and convection heat transfer, an unsteady heat transfer calculation method was used to obtain the inner wall temperatures of the train body and window at different times. Then, the entire line was divided into three sections based on the effect of the FAV on train energy consumption with changes in outdoor climate conditions, and typical stations were used to analyse the relationship between air supply parameters (ASPs) and energy consumption. The minimum energy consumption operation modes of the ACS and OSS that satisfied both comfort and hygiene conditions when the train passed through the entire line were obtained based on the above analysis results.

2. Physical and Mathematical Models

2.1. Physical Model

Currently, the train running on the Qinghai-Tibet railway line is a special 25 T plateau train produced by the Sifang and BSP companies [21]. The size of the train model was 25,500 × 3105 × 2500 mm. Ten rows of tables and chairs were placed in the model with a capacity of 98 people. The length and width of the windows were both 1050 mm. Luggage racks with a width of 500 mm were placed above the windows on both sides inside the train model. A roof orifice air inlet (50 × 50 mm) was used as the air supply. There were 64 rows in length and 10 air inlets per row. The front and rear doors of the carriage were used as return air outlets. The equipment and furnishings in the carriage were symmetrical around the middle section; therefore, half of the train was used as the calculation area. The physical model is illustrated in Figure 1.

2.2. Mathematical Model

The airflow organisation in the carriage was a mixed-convection heat transfer process that combines forced and natural convection, which is a turbulent flow state. In consideration of the high applicability of the RNG k−ε model under both high and low Reynolds numbers, we selected the RNG k−ε turbulence model to solve the flow and heat transfer processes in the carriage. The general governing equations describing flow and heat transfer can be written as follows [22]:

$$\frac{\partial (\rho u\varphi )}{\partial x}+\frac{\partial (\rho v\varphi )}{\partial y}+\frac{\partial (\rho w\varphi )}{\partial z}=\frac{\partial }{\partial x}(\mathsf{\Gamma }\frac{\partial \varphi }{\partial x})+\frac{\partial }{\partial y}(\mathsf{\Gamma }\frac{\partial \varphi }{\partial y})+\frac{\partial }{\partial z}(\mathsf{\Gamma }\frac{\partial \varphi }{\partial z})+S_{\varphi }$$

(1)

where $\mathsf{\Gamma }$ represents the diffusion coefficient; $\varphi$ is the general variable, which represent u, v, w, T, k, ε, respectively. Turbulent kinetic energy shear: $G_k={\eta }_t\left\{2[{(\frac{\partial u}{\partial x})}^2+{(\frac{\partial v}{\partial y})}^2+{(\frac{\partial w}{\partial z})}^2]+{(\frac{\partial u}{\partial y}+\frac{\partial v}{\partial x})}^2+{(\frac{\partial u}{\partial z}+\frac{\partial w}{\partial x})}^2+{(\frac{\partial v}{\partial z}+\frac{\partial w}{\partial y})}^2\right\}$. Turbulent kinetic energy buoyancy: $P_k=-\frac{{\eta }_{t}g\beta }{{\sigma }_t}\frac{\partial T}{\partial z}$. Turbulent viscosity: ${\eta }_t=c_{\mu }\rho k^2/\epsilon$. Thermal expansion coefficient: $\beta =1/(T_{ref}+273.15)$. Reference temperature is the set temperature in the train. The coefficients of the equations are listed in

Table 1. Coefficients in the control equation.

$\mathit{\varphi }$	$\mathsf{\Gamma }$	${\mathit{S}}_{\mathit{\varphi }}^{}$
u	${\eta }_{eff}=\eta +{\eta }_t$	$-\partial p/\partial x$
v	${\eta }_{eff}=\eta +{\eta }_t$	$-\partial p/\partial y$
w	${\eta }_{eff}=\eta +{\eta }_t$	$-\partial p/\partial z+\rho g\beta (T-T_c)$
T	$\eta /Pr+{\eta }_t/{\sigma }_t$
k	$\eta +{\eta }_t/{\sigma }_k$	$G_k+P_k-\rho \epsilon$
ε	$\eta +{\eta }_t/{\sigma }_{\epsilon }$	$c_1\rho S\epsilon -c_2\rho {\epsilon }^2/(k+\sqrt{v\epsilon })$

Values of various coefficients in Table 1: c_μ = 0.09, σ_k = 1.0, σ_ε = 1.2, c₂ = 1.9, c₁ = max[0.43, η/(5 + η)], η = Sk/ε, S = (2S_i,jS_i,j)^1/2, S_i,j = (∂u_i/∂x_j + ∂u_j/∂x_i)/2.

.

2.3. Mathematical Model Verification

The numerical results will be verified according to the experimental results in reference [23]. The classroom model is illustrated in Figure 2. Figure 3 shows a comparison between the experimental and simulated values. It can be observed that the experimental values are close to the simulation values in the same position (x = 9 m, y = 1.8 m, z = 0 m), and the maximum volume concentration differences of CO₂ are less than 0.005%. On the positions of teacher and student 2.1, several measuring points are distributed in the vertical direction. The maximum temperature differences between experimental values and simulated values are less than 0.5 °C. The reliability of the numerical method can be confirmed by the above results.

2.4. Assumptions and Boundary Conditions

For ease of calculation, the following assumptions were made:

(1)

Airflow in the carriage is a three-dimensional steady-state turbulent flow and the air is an incompressible gas. The change in air density adopts the Boussinesq assumption, and the air pressure in the carriage is approximately equal to the ambient pressure [5].

(2)

Air is a transparent radiation medium. The inner walls of the carriage, tables, chairs, and passenger surfaces are diffuse radiation gray bodies.

Boundary Conditions:

(1)

Inlet boundary: The air inlet is the speed inlet boundary; u = 0, v = 0, w = −1 m/s; the parameters of the adjustment working conditions are calculated based on the meteorological conditions along the railway line and different FAV. Turbulent kinetic energy and turbulent energy dissipation rate [24] were k = 0.004, ε = 0.0008.

(2)

Outlet boundary: The front and rear doors of the carriage were used as the outlet boundaries, P = P_out, $\partial k/\partial x=0$, and $\partial \epsilon /\partial x=0$.

(3)

Wall boundary: The middle section, luggage rack, and front and rear ends of the carriage were adiabatic boundaries, and the air–solid contact surface was a speed non-slip condition. In this study, the z264 train with the longest running period in summer was selected as the research object. The stations, altitudes, and arrival times (GMT + 8) are listed in

Table 2. Train passing stations and time (GMT + 8).

Number	1	2	3	4	5	6	7	8	9
Stations	Guangzhou	Chenzhou	Changsha	Wuchang	Zhengzhou	Xi’an	Lanzhou	Xining	Delingha
Altitude (m)	41.7	184.9	44.7	23.1	110.4	397.5	1517.3	2261.2	2881.5
Arrival time	11:52	15:34	19:08	22:34	3:33	9:36	16:26	19:11	23:33
Longitude (°)	113.26	113.04	113.02	114.32	113.67	108.97	103.86	101.82	97.39
Latitude (°)	23.15	25.81	28.20	30.53	34.75	34.29	36.04	36.63	37.32
Number	10	11	12	13	14	15	16	17	18
Stations	Golmud	Nachitai	Wudaoliang	Tuotuohe	Tanggula	Amdo	Nagqu	Damxung	Lhasa
Altitude (m)	2807.7	3575.2	4645.1	4533.1	5207.0	4801.4	4507.8	4200.7	3650.2
Arrival time	2:20	4:09	6:20	8:15	9:59	10:54	12:17	15:02	16:45
Longitude (°)	94.91	94.57	93.78	93.76	91.79	91.67	92.00	91.11	91.08
Latitude (°)	36.39	35.88	35.50	35.67	33.09	32.26	31.45	30.48	29.63

. Considering the attenuation and delay effects of the train enclosure structure on the process of solar radiation heat transfer, the unsteady heat transfer calculation method [25] was used to obtain the inner wall temperatures of the train body and window at different times as temperature boundary conditions for the numerical solution. The train windows were subject to solar radiation transmission effect based on the solar azimuth at different moments. The temperatures of the inner walls of the carriage when the train passed through different stations are shown in Figure 4.

(1)

Heat source boundary: Human body heat dissipation is an important part of the cooling load of a train and has a significant impact on the flow and temperature fields in the train. The heat dissipation was 116 W per person [24] and the heat was evenly distributed on the surface of the human body model.

(2)

Human breathing boundaries: Human respiration affects the concentrations of pollutants in carriages, thereby influencing the quality of fresh air. The amount of CO₂ produced by passengers was 18 L/(h·person) [26].

3. Numerical Methods

3.1. Meshing

In this study, an unstructured grid was used to divide the area under calculation. Considering the large gradient variation in various parameters in the breathing zone of passengers, air inlets, and air outlets, the grid adopted a method of dividing the carriage as a whole and local mesh encryption. After the trial calculations, when the number of grids exceeded 2,474,927, the change was less than 2% in the PP of CO₂ in the passenger breathing zone and 1% in the average temperature, as shown in Figure 5. Therefore, this set of grids could meet the requirements of grid independence.

3.2. Numerical Solution

The velocity-pressure coupling problem was tackled by applying the SIMPLE algorithm to solve the control equations [27]. The gradient term was discretised in a Green–Gaussian cell-based format. The momentum, energy, turbulent kinetic energy, turbulent energy dissipation rate, and component transport equations were discretised in the second-order upwind format. The convergence conditions for solving the governing equations were the same as those in [26].

4. Results and Discussions

Previous studies have investigated the influence of ASP on the temperature [13,15], velocity [15,22], and pollutant concentration fields [26] in Qinghai-Tibet railway trains when they passed through some stations. The results showed that adjusting the air supply temperature based on outdoor climatic conditions at different stations was beneficial for passenger comfort. For train z264, the interior temperature, CO₂, and O₂ contents in the train were numerically calculated when it crossed 18 stations in the plains and plateau. The control target was the requirements of temperature and oxygen PP in AC trains specified in the “TB/T 1932–2014” standard of the railway transportation industry in China [12]. The lumped-parameter method was used to establish an energy-balance model for the train.

$$\rho V\frac{dh}{d\tau }=Q_R+Q_{XF}+R+{\alpha }_n(T_i-T_n)A-Q_{PF}$$

(2)

where h is the air enthalpy, kJ/kg; Q_R is the passenger heat dissipation, W; Q_XF is the fresh air load, W; R is the solar radiation through train windows, W; α_n is the convective heat transfer coefficient of interior wall, W/(m²·K); T_i is the interior wall temperature, K; T_n is the interior air temperature, K; A is the interior wall area, m²; Q_PF is the heat removed by exhaust air, W.

The CO₂ concentration (<0.15%) specified in the standard [12] applies to plains; however, there is no standard for plateaus. The optimal value of CO₂ PP in the aerospace field is 500 Pa, which significantly surpasses the limit value in the standard [12]. In human respiratory physiology, pulmonary gas exchange is caused by the pressure difference between the alveolar gas and blood around the alveoli [28]. Therefore, referring to the upper limit of 0.15% CO₂ concentration in the plains [12], it was more reasonable to use a PP of 152 Pa (equivalent to 0.15% CO₂ concentration in the plains) as the CO₂ limit in the plateau. The equations for calculating the CO₂ and O₂ PP are as follows:

$$P_{C{O}_2}=P_A·y_{C{O}_2}$$

(3)

$$P_{O_2}=P_A·y_{O_2}$$

(4)

where y_CO2 and y_O2 are the CO₂ and O₂ volume concentrations, respectively, and P_A is the atmospheric pressure.

The above results indicate that the distribution of CO₂ in the carriage is nonuniform, and the average CO₂ concentration in the carriage does not represent the real level in the breathing zone of passengers.

A comparison between the average CO₂ PP in the carriage and the average CO₂ PP in the breathing zone under the same FAV (20 m³/h·person) is shown in Figure 6. It can be observed that the average CO₂ PP in the carriage is lower than that in the breathing zone when the train is running. Basically, if the upper limit of the average CO₂ PP in the carriage is used as the reference parameter for adjusting the FAV, the CO₂ released by passengers would exceed the allowable upper limit. Therefore, the average CO₂ concentration in the carriage does not represent the real level in the breathing zone of passengers, and the lumped parameter method is no longer applicable for the calculation of the CO₂ PP.

In summary, the ASP was adjusted based on the limits of the above standard [12], in which the CO₂ PP and O₂ PP calculations were based on the passenger breathing zone as the target area.

The energy consumption of the ACS and OSS under each condition was calculated to obtain the working conditions for minimum energy consumption that satisfied both comfort and hygiene conditions. Figure 7 shows a schematic of air circulation in the train. The energy consumption of the ACS is related to the altitude and ambient temperature outside the train. According to the heat exchange equation, the heat exchange rate is:

$$Q_{\mathrm{H}}=m_{\mathrm{H}}{c}_{\mathrm{p}}\Delta T={\rho }_{\mathrm{H}}{V}_{\mathrm{H}}{c}_{\mathrm{p}}\Delta T$$

(5)

where Q_H is the heat exchange rate of the heat exchanger at an altitude of H, W; ρ_H is the air density, kg/m³; V_H is the AC volume flow, m³/s; c_p is the specific heat at constant pressure, kJ/(kg·K).

The energy consumption of the oxygen supply equipment was chosen as 0.46 kW·h/m³ [18]. The supply air speed and temperature for the basic calculation conditions were taken from [22], and the FAV and supply oxygen concentrations were taken from the recommended values of reference [21] and are summarised in

Table 3. Basic calculation conditions.

Parameters	Velocity Speed	Temperature	Fresh Air	O2 Concentration (Nachitai–Lhasa)
value	1 m/s	18 °C	20 m3/(h·person)	25%

.

4.1. Analysis of Train Energy Consumption

The FAV mainly depends on the CO₂ content inside and outside the carriage and that produced by the passengers. Preliminary studies have indicated that the minimum FAV required by passengers increases with an increase in altitude, leading to dilution of the O₂ concentration, and increases the energy consumption of oxygen production [7]. Therefore, to design a reasonable installed capacity of the ACS and OSS, this study used the meteorological parameters of train z264 passing through each station as the calculation condition and calculated the lowest energy consumption of the ACS and OSS along the Qinghai-Tibet line.

The entire railway line can be divided into three sections based on the effect of FAV on train energy consumption with changes in outdoor climatic conditions: (1) non-oxygen supply areas where the enthalpy of ambient air is higher than the indoor enthalpy; (2) non-oxygen supply areas where the enthalpy of ambient air is lower than the indoor enthalpy; and (3) oxygen supply areas where the enthalpy of ambient air is lower than the indoor enthalpy. We analyse and discuss with respect to representative stations in the three sections.

The net heat input of each wall and heat dissipation of the heat source at the three representative stations are listed in

Table 4. The heat gains of the train.

Station	Enclosure Structure Heat Transfer Quantity/W						Solar Radiation/W		Passenger/W	Total Heat Gains/W
	Left Wall	Right Wall	Roof	Floor	Left Window	Right Window	Left Window	Right Window
Changsha	440.1	421.1	648.4	309.0	228.5	244.5	0	0	11368.0	13659.6
Xining	274.9	254.6	480.8	273.1	10.7	79.8	0	0	11368.0	12742.1
Wudaoliang	−293.1	−287.6	−673.8	−52.1	−190.5	−285.0	670	43	11368.0	10298.9

. A negative value indicates that heat was transferred from the interior to the exterior of the train.

4.1.1. No-Oxygen Supply Areas Where the Enthalpy of Ambient Air Is Higher than That of Indoor Enthalpy (Changsha Station)

Figure 8 shows the effect of FAV on the ACS energy consumption and CO₂ PP in the train under different supply air temperatures. The ACS energy consumption included two parts: cooling process after mixing fresh and returned air, and heating process before supplying air. The recommended value of 20 m³/(h·person) for FAV was chosen as the basic parameter for calculation [21]. The results indicate that the CO₂ average PP of the passenger breathing zone was only 136 Pa, which is lower than the upper limit of the CO₂ PP. The air quality in the carriage was good under basic conditions. A larger FAV corresponds to a larger fresh air load of the ACS, which leads to an increase in the ACS energy consumption. For FAVs of 0.4 m³/s, 0.3 m³/s, and 0.2 m³/s, we obtained 138, 143, and 154 Pa, respectively, as the average PP of CO₂ in the passenger breathing zone. As the FAV decreased to 0.2 m³/s, the average CO₂ pressure in the breathing zone exceeded the upper limit, and the air quality could not meet the hygiene requirement. The critical FAV that met the CO₂ PP requirement of the passenger breathing zone was 0.23 m³/s. Under this condition, fresh air load was the smallest and the ACS energy consumption was the lowest at 38.47 kW. When the supply air temperature was changed without changing the FAV, the ACS energy consumption exhibited a trend coincident with the supply air temperature. Under the premise of the standard [12], maximising the supply air temperature can achieve the lowest ACS energy consumption.

4.1.2. No-Oxygen Supply Areas Where the Enthalpy of Ambient Air Is Lower than That of Indoor Enthalpy (Xining Station)

Figure 9 shows the effect of different FAVs on the CO₂ PP and ACS energy consumption. Contrary to Changsha station, the enthalpy of the outdoor air at Xining station was lower than that of the indoor enthalpy; that is, the enthalpy of the mixed air containing fresh air and returned air was lower than that of the indoor enthalpy. As shown in Figure 10, the enthalpy difference in the air treatment process (C-L) decreased as the FAV increased. However, the residual heat and humidity in the train were both positive. Therefore, increasing the FAV helped reduce the cold load of the ACS and the CO₂ PP in the carriage simultaneously. The variation trend of ACS energy consumption with air supply temperature was opposite to that for the Changsha station. Increasing the supply air temperature increased the indoor temperature and simultaneously increased the enthalpy difference of C-L and L-S in the air treatment process. Therefore, using a lower supply air temperature and larger FAV can effectively reduce the energy consumption of the ACS. When a supply air temperature of 292 K and FAV of 0.544 m³/s were used, the temperature in the carriage was within the comfortable range of 24–28 °C, and it could be ensured that the CO₂ PP did not exceed the upper limit.

4.1.3. Oxygen Supply Areas Where the Enthalpy of Ambient Air Is Lower than That of Indoor Enthalpy (Wudaoliang Station)

Table 5. Calculation conditions of Wudaoliang railway station.

Calculation Conditions	Basic Condition	1	2	3	4	5
Temperature/K	291	291	291	292	293	294
Fresh air volume/m3·s−1	0.544	0.400	0.300	0.300	0.300	0.300
Oxygen supply volume/m3·s−1	0.176	0.154	0.138	0.138	0.124	0.102

lists the calculation results for Wudaoliang station, which has the lowest ambient temperature in the oxygen supply areas. The air supply conditions at Wudaoliang station are also listed in Table 5.

Figure 11 shows the effect of different FAV on the CO₂ PP and ACS energy consumption. The results indicate that when the FAV was 0.3 m³/s, the PP of CO₂ in the passenger breathing zone reached the maximum allowable value. At the same supply air temperature, energy consumption of the ACS reached a minimum value when the FAV was 0.4 m³/s. When the supply air temperature was 294 K, it could meet the minimum requirement of 24 °C in the carriage and maintain low ACS energy consumption.

Figure 12 shows the O₂ PP in the train and the total energy consumption of the ACS and OSS under different calculation conditions (Table 5). The effect of changing the FAV on the energy consumption of ACS and OSS revealed significant differences in the calculation results. Increasing the FAV simultaneously diluted the concentrations of CO₂ and O₂ in the train. This explains the restrictive relationship between O₂ PP, CO₂ PP, and FAV. Calculation conditions 1–3 in Table 5 indicate that reducing the FAV can simultaneously reduce the OSV while maintaining the indoor O₂ PP. Therefore, reducing the FAV can significantly reduce the energy consumption of the oxygen supply. Calculation conditions 3–5 further explain the importance of reducing the oxygen supply to save energy while maintaining FAV. When the OSVs were 0.138, 0.124, and 0.102 m³/s, the oxygen energy consumption was reduced by 21.59, 29.55, and 42.07%, respectively, compared to the basic condition. When the OSV was 0.102 m³/s, the O₂ PP in the carriage reached the lowest limit specified in the standard [12]. In Figure 12, it can be seen that the energy consumption of the ACS was the lowest under calculation condition 1. However, the ACS energy consumption accounted for a small proportion of the total energy consumption, indicating that the oxygen supply energy consumption had a greater impact on the total energy consumption. The importance of reducing OSS energy consumption was also revealed in reference [8], and measures were provided to reduce OSS energy consumption by reducing the FAV. Basically, the total energy consumption of the system achieved its minimum value under calculation condition 5, in which the oxygen energy consumption was the lowest. Therefore, when the train was running in oxygen supply areas, the adjustment of the FAV should be based on reducing the OSS energy consumption as measures to reduce the ACS energy consumption had little effect on the total energy consumption.

4.2. CO₂ Partial Pressure in the Running Train

Figure 13 shows the CO₂ PP variation curve under optimal ACS adjustment conditions. It is seen that when the train is running from Guangzhou to Lanzhou stations, the CO₂ PP in the breathing zone of the passengers is close to the upper limit value, and the ACS operates under minimum FAV. When the train is between Xining and Golmud stations, the FAV adopted the recommended value of the reference because increasing the FAV could reduce the ACS energy consumption [21]. Under these working conditions, the air quality in the train improved while reducing the energy consumption of the ACS. When the train ran in the O₂ supply areas (Nachitai-Lhasa) because the ACS energy consumption accounted for a small proportion of the total energy consumption, the aim of adjusting the FAV was to reduce the OSS energy consumption; therefore, the train was still running at the minimum FAV in these areas, and the CO₂ PP in the train increased again towards the upper limit.

4.3. O₂ Partial Pressure in the Running Train

Figure 14 shows the variation of the O₂ PP under optimal O₂ supply and no-oxygen supply conditions. It is observed that the variation trend of O₂ PP under no-oxygen supply conditions is opposite to the increasing altitude. Trains running in plains maintain a high O₂ PP level. When the train passed Xi’an station and continued to move west, with a rapid increase in altitude, the O₂ PP dropped sharply. The O₂ PP was the lowest (13.5 kPa) when the train arrived at Nachitai station. If O₂ supply measures had not been taken at this time, the air inside the carriage would have been in a hypoxic state for the rest of the journey. This result is consistent with the results in reference [5]; therefore, diffusion O₂ supply should be started after the Nachitai station, and the OSV should be adjusted under the lowest energy consumption condition to keep the O₂ PP above 13.5 kPa. When the train passed Tuotuohe station towards Tanggula, the altitude increased rapidly to 5072 m. Here, even if O₂ was supplied at the maximum concentration limit (25%), the PP of O₂ still could not reach 13.5 kPa. If 13.5 kPa was used as the O₂ PP adjustment target, the diffusion O₂ supply concentration would exceed 25%, which would not satisfy the requirements for a safe O₂ concentration range [21]. However, the safety problem was ignored in reference [5] to meet the O₂ PP, and the train running in this area should still use 25% oxygen concentration for oxygen supply. This suggests strengthening the use of distributed O₂ supply equipment and inspection of train stewards in this area [29].

4.4. Physical Field Distribution in the Train at Tanggula Station

Figure 15 shows the physical field distribution on the cross-section (y = 6 m) of the carriage when the train passed through Tanggula station under optimised air supply conditions. In this area, even if the maximum oxygen supply concentration was used, the average O₂ PP in the breathing zone of the passengers could not satisfy the minimum value of 13.5 kPa specified in the standard [12]. It is necessary to analyse each physical field in the train under optimal air supply conditions. The physical field in a train results from the combined effects of solar radiation, human body heat dissipation, AC, and passenger respiration. Furthermore, CO₂ and O₂ concentration fields in the train have a strong coupling effect with the temperature field [26].

The heat dissipation effect of the passengers led to a higher temperature above their heads. However, the temperature above the passenger heads next to the left window was higher. This was because the left wall faced the sun while traveling through Tanggula mountain station. Solar radiation through the train enclosure increases the temperature near the left wall through coupled heat transfer. Consequently, the buoyancy force in this zone increased, which caused the hot airflow to rise and formed a thermal stagnation zone under the luggage rack. However, the heat retention phenomenon in actual operation is also related to the migration and diffusion characteristics of pollutants, thus affecting ventilation efficiency [15]. The asymmetry of the cross-sectional flow field in the longitudinal direction of the carriage resulted in the accumulation of contaminants on the shady side, as shown in Figure 15a. In addition, the passageway area was cooler because of the lower air supply temperature from the middle of the roof.

The CO₂ PP gradient near left-seated passengers in the carriage was greater than that near the right-seated ones, whereas the CO₂ PP above the heads of right-seated passengers was significantly higher than that above the left-seated passengers. This was because the airflow in the carriage was caused by natural convection under a temperature difference and forced convection under mechanical ventilation. Owing to the effect of solar radiation, the higher-temperature airflow near the left side wall increased faster than that on the right side. Under the effect of the buoyancy force, the CO₂ exhaled by passengers migrated upward along the side wall with the transportation of the boundary layer. The CO₂ PP in the area above the heads of passengers was 139.32 Pa. The temperature near the right wall was relatively low; therefore, the rising speed of the airflow near the right wall was slower than that on the left wall. This would result in a slow dilution of the CO₂ concentration, causing the PP of CO₂ in this area to reach 141.49 Pa. The CO₂ PP in the passageways did not exceed the standard limit.

O₂ entered and circulated in the train through the air supply system. The movement characteristics of airflow in space determined the proportion of O₂ distributed to passengers at different positions. A supply airflow with high O₂ concentration was sent directly to the bottom of the train along the passageway and diffused at the left and right walls after encountering the bottom. Because of the blocking effect of the side walls and seats, the airflow formed a large-scale vortex under the passengers on both sides and moved up under the influence of the vortex. The updraft bypassing the seats accelerated and spread to the upper part of the train under the thermal buoyancy force formed by passenger heat dissipation. When the train passed the Tanggula mountain station, the left wall was sunny. The intense solar radiation through the window glass increased the air temperature near the inner surface of the window, resulting in a rise in the airflow speed under the effect of buoyancy, which was slightly greater than that on the shady side. A larger air diffusion zone with a high O₂ concentration was obtained on the sunny side (Figure 15d), which is consistent with the results shown in Figure 15b. Passengers consume a certain amount of O₂ and exhale CO₂ during respiration The gas generated by respiration moves with the airflow in the carriage to the roof such that the CO₂ PP near the top of the carriage is higher than that in the bottom, whereas the O₂ PP in the top is lower. Therefore, reasonable airflow organisation favours the removal of pollutants and rational distribution of O₂.

Overcoming the nonuniform phenomenon of temperature, air freshness, and oxygen concentration in trains is a technical improvement problem faced by air supply systems of plateau trains.

5. Optimal Control Scheme for Train Air-Conditioning and Oxygen Supply System

Figure 16 shows a comparison of the total energy consumption under basic and optimal conditions. The optimised ACS and OSS reduced energy consumption by varying degrees. In the no-oxygen supply areas, the working conditions were mainly adjusted by changing the FAV and supply air temperature. The aim of adjusting the FAV was to meet the hygiene conditions inside the carriage (CO₂ PP is the control objective) and adjust the supply air temperature required to meet the comfort range of 24–28 °C in the train. In the O₂ supply areas, the optimisation measures adjusted the FAV, air supply temperature, and OSV. The system energy consumption was affected by these three parameters. Figure 16 shows that the energy consumption of OSS has a significant impact on the total energy consumption. Even if the optimal ACS energy consumption of stations 11–18 was slightly higher than that of the basic working conditions, it was conducive to the reduction in the OSS energy consumption; thus, the total energy consumption was reduced. The adjustment parameters and energy-saving rates for each station are listed in

Table 6. Calculation conditions and energy consumption at each station.

Stations	Basic Parameters			Optimal Parameters			Basic Energy Consumption (kW)	Optimal Energy Consumption (kW)	Energy-Saving Rate (%)
	Tem (K)	Fresh Air (m3/s)	O2 (m3/s)	Tem (K)	Fresh Air (m3/s)	O2 (m3/s)
Guangzhou	291	0.544	-	292	0.200	-	55.61	39.32	29.29
Chenzhou				292	0.230		60.98	39.13	35.83
Changsha				293	0.230		58.68	38.47	34.44
Wuchang				293	0.230		49.33	35.48	28.08
Zhengzhou				293	0.230		41.72	32.59	21.88
Xi’an				292	0.250		44.25	33.74	23.75
Lanzhou				292	0.250		35.24	28.44	19.30
Xining				292	0.544		18.58	16.52	11.09
Delingha				293	0.544		11.19	7.26	35.12
Golmud				293	0.544		12.03	7.77	35.41
Nachitai			0.176	294	0.270	0.036	84.85	39.02	54.01
Wudaoliang				294	0.300	0.102	88.37	51.72	41.47
Tuotuohe				294	0.300	0.102	85.16	49.28	42.13
Tanggula				293	0.350	0.146	83.13	67.39	18.93
Amdo				293	0.300	0.124	86.68	62.28	28.15
Nagqu				293	0.300	0.102	88.14	54.30	38.39
Damxung				293	0.300	0.087	89.68	48.84	45.54
Lhasa				293	0.270	0.063	93.55	41.52	55.62

.

6. Conclusions

Under conditions that ensure the health and comfort of passengers during the operation of the plateau train, the energy consumption changes of the ACS and OSS along the Qinghai-Tibet railway line were studied by adjusting the ASP. The main conclusions are as follows.

(1)

The FAV in different areas increased with increasing altitude, and the O₂ demand of trains in the O₂ supply areas was positively correlated with altitude. The FAV and OSV reached maximum values of 0.350 m³/s and 0.146 m³/s, respectively, in the Tanggula mountainous area, which has the highest altitude.

(2)

When the train was running in areas where the ambient air enthalpy was higher than the indoor air enthalpy, reducing the FAV when the CO₂ PP was less than 152 Pa helped reduce the ACS energy consumption. When running in no-oxygen areas, where the enthalpy of the ambient air was lower than that of the indoor air, reducing the FAV led to an increase in the energy consumption of the ACS. The ACS in these areas should work under an FAV recommended by the specifications. When the train was running in the O₂ supply areas, the energy consumption of the ACS accounted for a small proportion of the total energy consumption, and the OSS energy consumption was significantly affected by the FAV. When the CO₂ PP was lower than the limit value, the minimum FAV should be used to minimise the energy consumption of the OSS. The maximum energy-saving rates of the system in the three areas were 37.16%, 35.41%, and 75.61%.

(3)

The air pressure in the Tanggula mountainous area was extremely low and the O₂ content was insufficient. The average O₂ PP of the passenger breathing zone could not satisfy the minimum standard requirement of 13.5 kPa under combined working conditions of a minimum FAV and maximum OSV. In this area, the use of distributed O₂ supply equipment and stewarded round-trip diagnosis should be strengthened.

(4)

As 25T trains are widely used in rail transit, their body structures cannot be easily changed. Optimising the air distribution is the only way to change the ASP. To improve the air composition inside a train, train body structure optimisation should be considered in the future.

In this study, the theoretical relationship among the FAV, OSV, and air parameters in a train running in different areas along the Qinghai-Tibet line was elucidated, and the dynamic operation modes of the ACS and OSS were proposed. Furthermore, the research results provide a healthy and comfortable interior environment for the plateau train and minimise its energy consumption.