Study on bioaerosol diffusion and deposition in a mobile BSL-4 laboratory based on air age analysis

Abstract

Biosafety issues have aroused global concern, especially after the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron strain of corona virus disease 2019 (COVID-19) caused incalculable human and property losses. Laboratory-acquired infections (LAIs) caused by improper operations or accidents are frequently reported. Research is urgently needed for a mobile biosafety level-4 (BSL-4) laboratory with a high risk for exposure. Deposition characteristics and the spatial distribution of bioaerosols under two typical cases were studied in this paper. Based on the age of air and simulation of airflow pattern, a detailed analysis of infection risk and the distribution of bioaerosols was conducted. The deposition characteristics of particles on different surfaces were analyzed based on particle tracking technology. The results showed that the removal rate of bioaerosols was lower in the space area of the laboratory from 1.6 m above the ground. The distribution of high-risk areas is affected by the coupling of equipment layout and pollution sources, mainly located downstream of the main airflow in the laboratory, and the particle concentration was eight times that of the low-risk areas. More than half of bioaerosol particles are deposited on laboratory equipment and walls. The number of particles deposited on the wall was the largest, accounting for 25.02% of the total. The unit area deposition ratio of the experimental table was the highest, which was 6.14 $\%/{\mathrm{m}}^2$. The main deposition area of each surface was determined, which could be of guiding significance to the determination of the key disinfection location of the mobile BSL-4 laboratory.

1. Introduction

The SARS-CoV-2 Omicron strain exhibits potent capabilities for immune evasion and viral entrance [1]. In the course of dealing with an epidemic outbreak and managing the emergent public health incidents, it would take a long time to construct a new fixed BSL-4 laboratory. A mobile laboratory with flexibility is urgently needed to prevent and control the epidemic in a timely and effective manner. Mobile biology laboratories have the characteristics of flexibility, modularization and automation. The mobile laboratory can be mainly transported by road, supplemented by railway, ship and air. In the case of asymptomatic transmission of novel coronavirus variants such as Omicron [2], the detection of pathogenic bioaerosols can be carried out by the mobile BSL-4 laboratory in epidemic areas, preventing the spread of viruses and ensuring the safe and closed discharge of highly infectious pollutants [3]. In order to adapt to different environments, the auxiliary equipment is modular, including generators, water treatment units, additional air filtration units, etc., which are designed into standardized modules for easy transportation. Specific modules were selected according to the actual situation and deployed to the site together with the mobile laboratory.

Mobile laboratories have been widely used for infectious diseases [4,5]. Raftery Philomena et al. completed the implementation of a rapid molecular diagnostic test run at a mobile laboratory in Liberia and analyzed case management and the subsequent impact on the Ebola virus disease (EVD) outbreak response [6]. A mobile laboratory provided a reliable solution for the rapid identification of the causative agent of non-malaria febrile illness (NMFI) in Africa, and it was proved that the results obtained in the mobile laboratory were reliable and consistent with those from a fixed laboratory [7]. In response to the COVID-19 pandemic, Xing Wanli et al. [8] used a stand-alone mobile laboratory that was integrated, automated and equipped with a 6-axis robot arm for specimen collection, inactivation, analysis, and reporting. The results showed the capability of a mobile laboratory to locally contain the spread of SARS-CoV-2. Aiming to optimize mobile laboratory testing, Ballard Susan A et al. [9] assessed the process of implementation, performance and impact of a mobile laboratory and investigated mobile laboratory testing optimization strategies.

As one of the main ways of novel coronavirus infection, bioaerosol transmission is of important research significance. Laboratory-acquired infections tend to occur in locations with a higher risk of potential infection, especially in BSL-4 laboratories. A number of studies have been carried out on bioaerosol particle diffusion in recent years. Yu et al. [10] began to analyze the diffusion and of pollutants in the air by computational fluid dynamics (CFD), and the CFD method was demonstrated to be able to predict the three-dimensional flow pattern of viral bioaerosols in indoor environments as they are diffused and suspended in the air. In a study of severe acute respiratory syndrome (SARS) virus transmission in wards, Qian et al. [11] used the infectious disease prediction equation coupled with the CFD method. The results matched the actual infection situation well. Zhang et al. [12] used the CFD program to predict the concentration distribution and dispersion of particles in ventilation rooms and test the effectiveness of different air supply systems in removing particles. It was found that particles were removed more effectively from the air distribution system under the floor than from the ceiling and side walls. Liu et al. [13] discussed the characteristics of diffusion and deposition of different kinds of bioaerosols, indicating that the deposition of particles on walls and equipment is improved by a reasonable distribution of experimental equipment. Feng et al. [14] indicated that a good indoor air flow organization strategy, reasonable and stable pressure gradient in the safety zone, and buffer room are important components of regulating air flow to prevent pollution in biological laboratories. To assess the potential for infection leakage, Wen et al. [15] tested and evaluated a laboratory high-efficiency particulate filter (HEPA) using bioaerosols. David L. Sewell [16] gave examples of possible laboratory infection risks for biosafety laboratory workers and indicated that some influenza incidents were caused by workers performing bioaerosol production procedures.

Most studies on mobile laboratories focused on their use in epidemic situations or shared early experiences with deployment and performance, such as on-site specimen collection and rapid identification, while few investigated the risk of exposure and infection when a leakage occurred in a mobile laboratory due to accidents or improper experimental operations. During the handling of highly infectious bioaerosols [17], ineffective discharge of pathogenic bioaerosols may result from improper operation, resulting in acquired infection in the laboratory [18]. Thus, it is critical to assess the potential infection risk of the mobile BSL-4 laboratory when a leakage occurs. In view of this knowledge gap, the purpose of this paper is to investigate the potential acquired infection risk under different typical leakage cases. The spatial distribution and deposition characteristics of bioaerosols due to accidents or improper experimental operations were explored based on air age and airflow pattern analysis. It also provides some scientific guidance to eliminate the bioaerosols from the local sources and grasp high-risk targeted disinfection locations from the perspective of protecting laboratory workers. In this study, our team cooperated with a company in Jiangsu province to build a mobile laboratory. Firstly, the air age transport equation was established to analyze the air age in the laboratory, and the areas with a low bioaerosol removal rate and a high infection risk were obtained. Secondly, considering the influence of pollution source location on infection risk, two typical cases of bioaerosol leakage were discussed and analyzed for spatial distribution and infection risk. Finally, for one of the conditions, the deposition characteristics of bioaerosols on different surfaces after a certain period of release were studied. The results could provide scientific reference for the formulation of safety protection measures and disinfection procedures, and reduce the possibility of laboratory-acquired infection.

2. Establishment of a mobile BSL-4 laboratory model

2.1. Physical model

Considering practicality and transportability, the rectangular shape is one of the typical shapes for mobile laboratories. For rapid processing of suspected case samples collected, Fall Amary et al. [19] deployed a mobile laboratory truck (MLT) in Senegal, which is a 6 m $\times$ 6 m truck equipped with an isolator for sample inactivation, a generator and batteries, and a molecular platform for pathogen detection.

In this study, in order to meet the requirements of traveling to remote areas and facilitating long-distance transportation, the laboratory form is designed as two square cabins connected by airtight soft connection channels. The carrier is a special dump truck. The laboratory can be operated directly on the truck or be deployed on solidified flat ground, which is convenient for personnel and logistics.

Fig. 1(a) shows a three-dimensional CFD simulation model of the mobile laboratory square cabin (the buffer is neglected). The internal size of the model is 6.2 m (X) by 2.2 m (Y) by 2.3 m (Z). The experimental model adopts the ventilation mode of upper supply and lower exhaust. Three evenly distributed air supply outlets and one exhaust outlet are set in the main laboratory. Air supply and exhaust outlets are 0.610 m by 0.305 m. According to negative pressure control and air volume calculation, airflow flows in from the air supply outlets at the top of the laboratory at a speed of 0.36 $m/s$. In order to meet the basic ventilation requirements, air flows out from the exhaust outlet 0.15 m away from the laboratory ground [20]. Keep the supply air volume less than the exhaust air volume in the laboratory and buffer zone to maintain negative pressure, and the ventilation number is set to 15 per hour according to design requirements.

Fig. 1.

Mobile BSL-4 laboratory simulation model.

Laboratory infrastructure conforms to the international standard “Laboratory Biosafety Manual” (third edition) in terms of its type, quantity, and location for BSL-4 laboratories [21]. “+" and "-" represent the upper and lower walls, left and right walls, and front and rear walls. The layout of the laboratory is completely designed according to the standards. Fig. 1(b) shows the top view of equipment layout. The flow rate and position of the air supply also meet specifications. Therefore, the results of this model can be used as a reference for other similar laboratories. In addition to a biosafety cabinet (BSC), the laboratory has common laboratory equipment, a cleaning table, and experimental tables. Boxes of rectangular shape are reasonable simplifications for these devices [22,23]. The device size parameters after simplification are listed in Table 1 . The equipment location is shown in Fig. 1(b).

Table 1.

Description of equipment sizes.

Description	X(m)	Y(m)	Z(m)
Biological safety cabinet	0.8	0.95	1.6
Laboratory cleaning region	0.5	1.0	1.2
Experimental table	0.5	3.2	1.1
Laboratory equipment (1)	0.8	1.25	2.0
Laboratory equipment (2)	0.8	1.3	1.0
Laboratory equipment (3)	0.8	1.5	1.0

2.2. Air volume control

The BSL-4 laboratory requires all fresh air exhausted. According to the standards, an adjacent room and the level 4 biosafety laboratory should have a difference in pressure of between −10 and −15 Pa. Because the pressure in the biological safety cabinet is required to be lower than the room pressure, excessive negative pressure in the laboratory requires greater air volume and allowance for leeway, which increases the energy consumption of the laboratory [24]. According to the air balance principle and air tightness test results of the level 4 biosafety laboratories, air supply and exhaust volumes were calculated for the room [25].

The exhaust air volume is mainly composed of the air volume permeated through the gap, the air volume introduced by opening and closing doors, and the air volume introduced by switching transfer window. The amount of exhaust air required to maintain negative pressure is defined by Eq. (1):

$$Q_o=3600E{F}_1\sqrt{\frac{2\Delta p}{\rho }}+tBE{F}_2\sqrt{\frac{2\Delta p}{\rho }}+0.5V_{1}B+q$$ (1)

where, $Q_o$ is the amount of exhaust air, $F_1$ is the slit area, $E$ is the flow coefficient, $\Delta p$ is the pressure gradient, $\rho$ is the air density, $F_2$ is the airtight door area, $t$ is the time to open the door, $V_1$ is the volume of the transfer window. $B$ is the frequency to open the door, 0.5 is the leakage rate, $q$ is the partial ventilation of biological safety cabinet.

The biosafety level 4 laboratory requires a full fresh air supply system, so the laboratory air supply is equal to the fresh air volume. Air supply depends on the air supply required by cleanliness and supplementary local exhaust. The amount of air supply is defined by Eq. (2) (3):

$$Q_i=\max \left(Q_r,q\right)$$ (2)

$$Q_r=nV$$ (3)

where, $n$ is the frequency of air change, $V$ is the room volume.

In order to maintain the negative pressure stably, the sum of supply air and leakage air volume should be equal to the exhaust air volume. The wind speed at each inlet is defined by Eq. (4):

$$v_i=\frac{Q_i}{3600\cdot N\cdot c_1\cdot d_1}$$ (4)

where, $N$ is the number of laboratory air supply outlets, $c_1$ is the length of the air supply outlet, $d_1$ is the width of the air supply outlet.

According to calculation results, the air velocity of the three air supply outlets in the biological laboratory should be more than 0.36 $m/s$.

2.3. Kinematic model of bioaerosol particles

In general, bioaerosol particles in the laboratory are subjected to multiple forces, including the additional mass force, Saffman force, drag force and the negative pressure gradient force. To make the numerical simulation results closer to the real results, other forces in the particle equilibrium equation play a crucial role. These forces can be partially ignored according to the size of particles and different indoor conditions.

The additional mass force is the additional force caused by the acceleration of the fluid around the particle. The additional mass force is defined by Eq. (5):

$$F_m=\frac{1}{2}\frac{\rho }{{\rho }_p}\frac{d}{dt}\left(u-u_p\right)$$ (5)

where, $u$ is the velocity of indoor air, $u_p$ is the velocity of bioaerosol, ${\rho }_p$ is the density of bioaerosol.

There is a transverse velocity gradient in indoor airflow, and the force exerted by bioaerosol particles under the influence of the velocity gradient is called the Saffman force. The Saffman force is defined by Eq. (6) [26]:

$$F_{\mathrm{S}}=1.62\mu d_p^2\frac{du/dy}{\sqrt{v|du/dy|}}\left(u-u_{px}\right)$$ (6)

where, $u_{px}$ is the velocity of particles in the X direction, $du/dy$ is the velocity gradient perpendicular to airflow to the laboratory wall.

Drag force is the resistance of air that particles overcome when they move relative to the surrounding air. Drag force is defined by Eq. (7) [27]:

$$F_D=\frac{18u_d}{{\rho }_p{d}_p^2}\cdot \frac{C_d\mathrm{R}\mathrm{e}}{24}\left(u_i-u_{pi}\right)$$ (7)

where, $u_d$ is the molecular dynamic viscosity of the air, ${\rho }_p$ is the density of the particle, $d_p$ is the diameter of the particle, $\mathrm{R}\mathrm{e}$ is the relative Reynolds number, $C_d$ is the fluid drag coefficient, $u_i$ is the velocity of indoor air, $u_{pi}$ is the velocity of bioaerosol.

Gravity deposition is another factor to consider because a bioaerosol particle's size varies. The gravity is defined by Eq. (8) [13]:

$$F_G=\frac{\pi }{6}{d}_{\mathrm{p}}^3{\rho }_{p}g$$ (8)

where, $d_p$ is the diameter of the particle, ${\rho }_p$ is the density of the particle.

Due to the high air tightness in the BSL-4 laboratory, the pressure change in the main laboratory area is very small, and the negative pressure gradient force is defined by Eq. (9) [28]:

$$F_P=\frac{\rho }{{\rho }_p}\cdot u_{pi}\cdot \frac{\partial u}{\partial x_i}$$ (9)

where, $\rho$ is the density of air. The negative pressure gradient force can be ignored because of the low density of air compared to bioaerosol particles.

A large negative pressure gradient in adjacent regions cannot be ignored when bioaerosol particles flowing between regions are considered as the research object, and the negative pressure gradient force is defined by Eq. (10):

$$F_P=\Delta P\cdot s$$ (10)

where, $\Delta P$ is the negative pressure gradient, $s$ is the projection area of spherical bioaerosol particles.

The density of bioaerosol particles in a laboratory is much higher than the density of its air. The order of magnitude of the additional mass force is small and can be ignored for the convenience of calculation.

In conclusion, based on Newton's Second Law:

$$m_p\frac{d\overrightarrow{u_p}}{dt}=\overrightarrow{F_S}+\overrightarrow{F_G}+\overrightarrow{F_D}+\overrightarrow{F_P}$$ (11)

where, $u_p$ is the velocity of particle motion.

2.4. The transport equation of air age

Local mean air age (MAA) is defined as the average time for air molecules to reach a certain position in the flow field after entering the room [29]. The age of air, which can measure the freshness of indoor air, is crucial for evaluating ventilation effects. The age of air is defined by Eq. (12) [30]:

$$\rho \frac{\partial \tau u_j}{\partial x_j}=\frac{\partial }{\partial x_j}\left[\left(2.88\rho \times {10}^{-5}+\frac{{\mu }_{\text{eff}}}{S{\mathrm{c}}_t}\right)\frac{\partial \tau }{\partial x_j}\right]+S_{\tau }$$ (12)

where, ${\mu }_{\text{eff}}$ is the effective turbulent viscosity of air defined by ${\mu }_{\text{eff}}=\mu +{\mu }_t$ in which $\mu$ and ${\mu }_t$ are the molecular and turbulent viscosities, respectively. $S{\mathrm{c}}_t$ is the turbulent Schmidt number for age of air and its typical values are in the range of 0.7–1.0. $S_{\tau }$ is the source term of $\tau$,which is usually set as a fixed value. In this study, $S_{\tau }$ = 1.

In air age models, tracer gas is widely used because it has been proven to be accurate in numerical simulations. Pollutants are removed more efficiently by the younger air age [31]. The air age transport equation can be derived from the calculation formula of air age and the corresponding transport equation of mass concentration [32]:

$$\frac{\partial \left(\mu {\tau }_P\right)}{\partial x}+\frac{\partial \left(\nu {\tau }_P\right)}{\partial y}+\frac{\partial \left(\omega {\tau }_P\right)}{\partial z}=\frac{\partial }{\partial x}\left(\alpha \frac{\partial {\tau }_P}{\partial x}\right)+\frac{\partial }{\partial y}\left(\alpha \frac{\partial {\tau }_P}{\partial y}\right)+\frac{\partial }{\partial z}\left(\alpha \frac{\partial {\tau }_P}{\partial z}\right)+1$$ (13)

where, ${\tau }_P$ is the age of air. $\mu$, $\nu$, $\omega$ are the velocities in the directions of the three axes, respectively. $\alpha$ is the diffusion coefficient. The form of local air recirculation zones caused by the turbulent diffusion of air passing through obstacles is reflected in the turbulent diffusion coefficient in the air transport equation.

3. Numerical simulation

3.1. Grid independence test

A great deal depends on the mesh density and structure in CFD numerical simulations [33,34]. Grids are used to discretize the whole biosafety laboratory. Grid independence verification is a prerequisite for numerical simulation [35], which is performed before adding pollution sources. The velocity changes of typical cross sections at position y = 1400 mm and z = 1400 mm in the main laboratory were compared respectively when the grid had 207587, 500563, and 1640296 cells. The results of grid independence test are shown in Fig. 2 .

Fig. 2.

Grid independence verification (t = 20s).

Fig. 2 shows that the velocity changes a lot with an increase in grids from 207587 to 500563, but the velocity field is almost unchanged with an increase in grids from 500563 to 1640296 in the typical section. Therefore, it can be believed that the model with the second type of grid is considered to be accurate enough. It is relatively reasonable to select the meshing method with a grid number of 500563 for subsequent simulations.

3.2. Boundary conditions

There is very little turbulence, low speed, and incompressible laboratory air selected for this study. The flow field is nearly not affected by particles since their volume fraction is small in the air [36]. According to existing literature, micron bioaerosols evaporate very quickly, so bioaerosols evaporating are not taken into account. Particle sizes measured in the experiment represent vaporized bioaerosol particles [37].

This study is mainly based on the boundary conditions:

• (a)The standard K-ε turbulence model was chosen since it has been demonstrated to be successful in modeling turbulence in buildings indoors [38].
• (b)Biosafety laboratories are strictly designed in accordance with laboratory design manuals. The standard wall function is used to simulate the near-wall region with low Reynolds number. The wall condition is defined as the non-slip wall boundary condition.
• (c)The main laboratory adopts a full fresh air system. The pressure is specified at the outlet and the velocity of air is specified at the inlet. The air supply outlet in the ventilation system is defined as the “velocity inlet”. And according to the calculation results, the air velocity of the three air supply outlets in the biological laboratory is set to 0.36 $m/s$.

4. Results and discussion

4.1. The age of air

Fig. 3 was the distribution of air particles of different ages of air (2 s–70 s). It could be found that the air age was related to the location of the air supply outlets and exhaust outlet. Fig. 3(a) and (b) showed that a younger age of air was found near the inlets and the area around the equipment near the exhaust outlet had an older air age, as shown in Fig. 3(f)–3(h). Secondly, there was relatively empty space and little air pressure around the air inlet, which allowed the air to diffuse around it, and there were areas with relatively old air age near the biological safety cabinet and the laboratory equipment (1) in Fig. 1. Bioaerosols can be removed more efficiently near the air supply outlets. Therefore, it can be inferred that the obstruction of equipment to air flow is one of the main factors for the decline of exhaust capacity. In addition, it can be observed that affected by airflow pattern, the majority of older air age areas were located above 1.6 m, which indicates poor bioaerosol removal efficiency.

Fig. 3.

The distribution of air particles of different ages of air (2s-70 s).

According to the results, the distribution of air ages was mostly determined by how the experimental equipment was arranged and how the air flowed in the BSL-4 laboratory. Although the removal efficiency of bioaerosols in areas above 1.6 m from the ground was low, airflow pattern, location of the pollution source and other factors affected the concentration as well, so the risk of infection associated with improper operations needs to be further investigated.

4.2. Simulation of airflow pattern

Airflow is one of the main factors in removing bioaerosols. An analysis of the airflow pattern in the key cross section (z = 1600 mm) was conducted to further study the distribution characteristics of bioaerosols after investigating the age of air. To visualize the airflow pattern, velocity distribution and streamline maps at different moments were employed to determine airflow characteristics. The airflow direction is represented by the streamline arrow.

Fig. 4 shows that the fresh air flows into the main laboratory vertically through three air supply outlets and spreads around. After 7 s, part of the air flow diffuses to the airtight door, hits the wall, and flows towards the exhaust outlet with the other part. At approximately t = 16 s, it can be found that the airflow velocity increases. Due to the constant collision between the airflow and the wall, the airflow velocity and direction change rapidly, and vortex areas of different scales occur around the corners of the laboratory. After 20 s, the main air flow gradually forms. Subsequently, a rewinding flow occurs in the downstream area due to the obstruction of the equipment after 28 s, especially near the surface of the laboratory equipment (1). Due to the influence of the main flow, the rewinding flow area gradually moves down and eventually forms a large vortex after 35 s. In general, the airflow direction is toward the exhaust outlet. The airflow near the air supply outlet with a high velocity (the maximum is 0.3 $m/s$) has a stronger capacity to carry bioaerosols. It is also consistent with the result of the air age analysis.

Fig. 4.

Velocity streamline maps at different times (z = 1600 mm).

The ability to remove pollutants is weaker in the reflux zone, which causes bioaerosols to fall out, thus increasing the risk of exposure. Considering the airflow characteristics of the selected section, bioaerosol accumulation and deposition tend to occur near the laboratory wall and in the downstream area.

4.3. Bioaerosol spatial distribution and infection risk assessment

The distribution and infection risk of bioaerosol particles released from different pollution sources were studied according to the results of air age analysis. Two typical cases of bioaerosol leakage are presented here. Laboratory equipment (3) and an experimental table are where the pollution originates. In both cases, the air supply volume was 770 $m^3/h$. The boundary conditions of the two cases are identical to the simulated settings except where the pollution source is located.

4.3.1. Bioaerosol spatial distribution of case 1

Fig. 5 shows the spatial distribution of bioaerosol particles after 1000 s of bioaerosol leakage in the case of the pollution source on the laboratory equipment (3). In this study, the bioaerosol number ratio is calculated based on the ratio of particles in different areas to the total number of bioaerosols in the respiratory area. The infection risk of the laboratory experimentalists in different areas was reflected in Case 1.

Fig. 5.

Bioaerosol spatial distribution in the respiratory area in Case 1.

The rate of pollutant production varies considerably under different conditions [39]. Case 1 and Case 2, respectively, describe two typical cases of bioaerosol leakage caused by accidents or improper experimental operations. In such cases, it is less similar to the sustained release of traditional human pollution sources. A large number of bioaerosol particles are released in a moment. After the leakage, the inlet air starts assimilating bioaerosols when it approaches the source. In Case 1, the distance between the source (1) and the air inlet was 1.9 m, and the time required for the air to approach the pollution source was about 6.1s. With the diffusion of bioaerosols in the laboratory for 1000s, bioaerosols as pollutants are distributed in most areas of the room, and the time before the inlet air approaches and assimilates the bioaerosols gradually becomes shorter. The distribution characteristics of bioaerosols after 1000 s of leakage were analyzed using the age of the air and simulation of airflow patterns.

Fig. 5 showed that bioaerosol concentration and infection risk were still high in some areas of the room, and they were distributed spatially in a regular manner after 1000 s. The high-risk areas are clearly located downstream of pollution sources. In an area that occupied 20% of the laboratory, bioaerosol particles accounted for more than 32% of the total, resulting in a higher risk of infection. The area with the younger air age was a low-risk area. The number of bioaerosol particles is 5% of the total in an area that occupied 20% of the laboratory. In addition, the risk of infection was lower near laboratory safety doors and cleaning tables. Due to its distance from pollution sources, the former occurred, so it was difficult for the air flow to carry bioaerosols to the area; the latter is because the air age was young and the bioaerosol removal rate was high. Therefore, the experimental personnel should be evacuated from the high-risk area first after bioaerosol leakage occurred in Case 1.

The results showed that various spatial regions showed varying levels of infection risk. Key factors include airflow patterns and equipment layout. Influenced by airflow patterns, bioaerosols tend to stay in the downstream of airflow, particularly around the exhaust outlet where the air age was old. At the same time, once the bioaerosol particles reach the downstream region, especially near the equipment, they are easily disturbed by the eddy current and remain there for a considerable amount of time, forming a high infection probability. Therefore, when bioaerosol leaks occur, it is crucial to optimize airflow patterns and equipment layout to reduce the risk of exposure.

4.3.2. Bioaerosol spatial distribution of case 2

Fig. 6 shows the spatial distribution of bioaerosol particles at 1000 s after bioaerosol leakage on an experimental table. Compared with Case 1, there were still bioaerosol accumulation areas in the laboratory near the exhaust outlet. In case 2, the distance between the source (2) and the air inlet was 1.2 m, and the time required for the air to reach the pollution source was about 3.8s.

Fig. 6.

Bioaerosol spatial distribution in the respiratory area in Case 2.

Compared with the results of Case 1, it can be found that the area with high infection risk in Case 2 occupied about 8% of the whole laboratory area, while the number of bioaerosol particles accounted for more than 18%. Bioaerosol particles accumulated further, and areas of high risk of infection migrated further downstream in the laboratory. Areas with a younger air age were low-risk areas, where bioaerosol particles accounted for 11% of the total number of particles in an area that occupied 28% of the total area of the laboratory.

There are two main reasons for this phenomenon: (1) Relative location of the source of pollution. In Case 1, the pollution source was far away from the air supply outlet, and the air age was older, which reduced the ability of airflow to remove bioaerosols. In Case 2, the pollution source was located near the air supply outlet, where the air age was relatively young, and the bioaerosol removal rate was high. (2) The effect of device layout. In case 1, there was a lot of equipment near the pollution source. Bioaerosol movement was affected by equipment arrangement and the pollution source at the same time. Due to the coupling effect of pollution source and equipment arrangement, the removal effect of air flow on bioaerosols is further weakened. Compared with Case 2, the high-risk areas of bioaerosol accumulation in Case 1 migrated to the middle of the laboratory. In Case 2, there was less equipment near the pollution source, so the airflow was not easily disturbed by the equipment, and it was easier to carry bioaerosols to the downstream area. At the same time, it tends to form eddy currents around the equipment due to the obstruction effect of equipment near the exhaust outlet, which leads to the aggregation of bioaerosol particles. The distribution of airflow is also consistent with this.

In conjunction with the air age analysis, it was found that older air age areas and high-risk areas, which were distributed downstream of the directional airflow near the experimental equipment, had similar distributions. In addition, affected by air flow patterns, the area with the younger air age was always low-risk. Therefore, it can be considered that the main factors affecting bioaerosol distribution and infection risk include airflow pattern, the location of air outlet and pollution source, and equipment layout.

4.3.3. Bioaerosol distribution and infection risk after increasing air supply

The actual demand for air supply varies according to the design requirements of different laboratories. At the same time, in order to analyze the relationship between air supply volume and bioaerosol spatial distribution, the spatial distribution of the above two cases under the condition of double air supply volume (1540 $m^3/h$) was analyzed.

By comparing Fig. 5(b) with Fig. 7 (a) of Case 1, it can be found that after the air supply volume was doubled, the high-risk distribution area of the laboratory in Case 1 decreased by about 30%, while the trend of high-risk distribution remained basically unchanged. Similarly, a similar conclusion can be drawn by comparing Fig. 6(b) with Fig. 7(b) of Case 2. There was still a great deal of danger in the middle and downstream areas in case 2. And the dangerous area was significantly reduced in size. The comparison of Case 1 and Case 2 showed that the risk of infection could be significantly reduced by increasing air supply when bioaerosol leaks occur in biosafety laboratories. On the one hand, increasing air supply can improve the bioaerosol removal ability of airflow in the laboratory and reducing the risk of infection. On the other hand, increasing air supply volume did not change the airflow pattern in the laboratory, so it basically did not affect the spatial distribution of bioaerosols.

Fig. 7.

Spatial distribution of bioaerosols in two cases with double air supply.

4.3.4. Surface deposition characteristics of bioaerosols

Another major cause of laboratory-acquired infections is bioaerosol deposition on experimental equipment. In addition, there are significant differences in deposition temporal and spatial characteristics between different surfaces in the laboratory, which requires further analysis.

Considering that there were more high-risk areas and their distribution was more complex in Case 1, the deposition characteristics of bioaerosols are further studied based on Case 1. Fig. 8 showed the deposition quantity and deposition ratio on different surfaces after the leakage on the laboratory equipment (3) occurred for 1000 s. A bioaerosol deposition ratio is calculated based on the percentage of particles deposited on each statistical surface to the number of pollutants originally released. Local deposition ratios are calculated by comparing the total number of particles deposited on different surfaces to the total amount deposited on each surface [13]. The bioaerosol deposition ratio and Local deposition ratios are defined by Eqs. (14), (15).

$$K=\frac{{\sum }_{i=1}^m{N}_i}{A}$$ (14)

$$\mathrm{k}=\frac{N_i}{{\sum }_{i=1}^m{N}_i}$$ (15)

where, $K$ is the deposition ratio, $m$ is surfaces where particles can be deposited and $N_i$ describes how many particles are deposited on the surface of unit $i$, $A$ is the total number of particles released and $k$ is the local deposition ratio.

Fig. 8.

Bioaerosol deposition on different surfaces.

Fig. 8 shows the degree of contamination on each surface of the laboratory. The laboratory walls were deposited with about 25% of the bioaerosol particles, accounting for 44.63% of the total deposited particles. Experimental equipment surface particles accounted for 13.74% of the total deposited particles, and the deposition ratio reached 7.7%. The number of particles on the laboratory ground accounted for 11.7% of all deposited particles, and the deposition ratio was 6.56%. There were the fewest bioaerosol particles deposited on the laboratory cleaning region, accounting for 3.42% of all depositied particles, and the particle deposition ratio was 1.92%. It is worth noting that the particle deposition ratio on the surface of the experimental table was 5.16%, but its unit area deposition ratio (6.14 $\%/m^2$) was the highest considering the difference in surface area. Therefore, even the small surface area of one device is a major surface for particle deposition. When sterilizing equipment, more attention should be paid to the main deposition surfaces.

The main sedimentary areas of several typical laboratory surfaces were studied further in order to study the specific deposition characteristics of bioaerosols on them.

Fig. 9, Fig. 10, Fig. 11 show the deposition distributions of bioaerosols on the wall, floor, and the main laboratory equipment in Case 1. Fig. 12, Fig. 13, Fig. 14 explore the formation mechanism of bioaerosol deposition in Case 1. Considering the existence of high-concentration particles in each distribution map, sedimentary distribution is difficult to highlight. In this way, the number of particles has been reduced proportionally. Only 1/10 of the actual bioaerosol particles are displayed.

Fig. 9.

Deposition characteristic of bioaerosols on the ground after 1000 s.

Fig. 10.

Deposition characteristic of bioaerosols on X- and X+ walls after 1000 s.

Fig. 11.

Deposition characteristic of bioaerosols on laboratory equipment after 1000 s.

Fig. 12.

Velocity streamline diagram at the cross-section (z = 100 mm) near the ground.

Fig. 13.

Velocity streamline diagram at the cross-section (x = 4500 mm) near the X+ wall.

Fig. 14.

Velocity streamline diagram at the cross-section (x = 500 mm) near the X-wall.

Deposition distribution on the ground is shown in Fig. 9. The main sedimentary areas were found around the laboratory cleaning region and the biological safety cabinet. The space near this area is low and narrow. As shown in Fig. 12, there is an obvious obstruction from the laboratory equipment (2) and laboratory cleaning region, resulting in rewinding airflow, which finally forms four main reflux zones. These can prevent bioaerosol particles from discharging with the airflow. As a result, a large amount of sediment is deposited. Combined with bioaerosol spatial distribution in the previous section, it can be found that the deposition area basically overlaps with the high-risk area.

Fig. 10 shows the deposition characteristic of bioaerosols on the walls. According to the deposition characteristics of the two walls, the X-wall deposited more bioaerosols than the X+ wall, which is affected by the airflow characteristics. The majority of the bioaerosol deposition was located at the edge of laboratory equipment (3) on the X+ wall. As shown in Fig. 13, many divergent airflow regions occur at the cross-section (x = 4500 mm), which is caused by the sudden change in airflow velocity and direction close to the X+ wall. Bioaerosols were easily adhered and deposited on the X+ wall under the influence of nearby eddy currents and divergent airflow. The deposition area on the X-wall surface was mainly located around the experimental table. It was close to the exhaust outlet, and the ability of air flow to remove bioaerosols was weak. As shown in Fig. 14, after 10 s, the main airflow hit the experimental table and then continued to move toward the exhaust outlet. After 31 s, the main vortex region occurred near the experimental table. Instead of being carried directly out of the room, the large amount of bioaerosols carried by the main flow of the laboratory stayed there for a long time, eventually depositing on the surface near the experimental table.

Fig. 11 shows the distribution of bioaerosol deposition on laboratory equipment after 1000 s. Due to the narrow space above the biological safety cabinet and the laboratory equipment (1) in Fig. 1, it was difficult for bioaerosols to reach the inside, so the distribution of bioaerosols was dense at the edges and sparse inside. The space of the experimental table was relatively low and there were many eddy current areas nearby, resulting in no effective discharge of bioaerosol particles with the airflow. This is also consistent with the characteristics of the main sedimentary areas on the X-wall. By analyzing the deposition of bioaerosol particles on each surface, it can be concluded that airflow characteristics play an important role in bioaerosol deposition.

5. Conclusions

In this study, the bioaerosol particle concentration and infection risk in main areas of BSL-4 laboratory under typical improper operation or accident were analyzed by means of air age analysis and particle tracking. The deposition characteristics of bioaerosol particles on laboratory walls and different equipment were further discussed. As a result of this study, the following conclusions were drawn:

• (1)The air age was mainly related to experimental equipment layout and the distribution of air supply and exhaust outlets in the BSL-4 laboratory. One of the primary reasons for the increase in air age and the decline in exhaust capacity was the blocking effect of equipment on air flow. In the area above 1.6 m, bioaerosol removal was low. Therefore, when carrying out laboratory disinfection, more attention should be paid to the disinfection of areas above 1.6 m.
• (2)The risk of infection is influenced by the coupling of the distribution of pollution sources and equipment layout. Compared with the experimental table, there was more equipment near the laboratory equipment (3) and there was a certain distance from the air supply outlet, so the air age was older. And when the leakage occurred at the laboratory equipment (3), the high-risk area occupied about 20% of the whole laboratory and migrated to the middle area. Therefore, the equipment layout of the laboratory should be orderly arranged to avoid accumulation, make full use of the bioaerosol removal ability of the main airflow in the laboratory, and reduce the risk of exposure.
• (3)The distribution of high-risk areas was basically the same as that of the older air age, mainly located downstream of the main air flow in the laboratory. And the concentration of bioaerosols was 8 times higher than that of a low-risk area. Therefore, when conducting laboratory experiments on pathogens such as Omicron, if bioaerosol leakage occurs, it is important for the experimenter to keep clear of the high-risk areas downstream of directional airflows at first. It is instructive to prevent acquired infection in the laboratory.
• (4)The high air supply volume can improve the bioaerosol removal ability of the airflow in the laboratory and effectively reduce the risk of infection. However, increasing the air supply volume does not change the airflow pattern in the laboratory, so the distribution characteristics of high-risk areas are not affected.
• (5)More than half of the bioaerosol particles were deposited on the walls and laboratory equipment. The experimental table exhibited the highest unit area deposition ratio (6.14 $\%/m^2$). The laboratory walls were deposited with about 25% of bioaerosol particles, which accounted for 44.63% of the total deposited particles. The experimental equipment was deposited with about 13.74% of bioaerosol particles, and the deposition ratio was 7.7%. Even the small surface area of one device is the main surface for particle deposition. Therefore, it is important to disinfect disinfection of main deposition surfaces such as the wall and experimental table, focusing on cleaning the main sedimentary area to reduce the possibility of LAIs.

In summary, the bioaerosol distribution and deposition characteristics in a mobile BSL-4 laboratory were investigated in detail using the computational fluid dynamics approach. The distribution of high-risk areas and main deposition areas under two typical cases were identified. The conclusions can provide a valuable scientific basis for laboratory disinfection. However, this study is based on a biosafety laboratory of a particular type. Therefore, optimizing and improving laboratory layout, including the location of equipment and airflow inlet or outlet, should be focused on in future studies.

Funding

This work was a special fund project of the Shanghai Municipal Civilian and Military Intergration Development, supported by the Shanghai Office of Civilian and Military Integration [2020-jmrh1-kj38].

Author contribution

Yan Wang: conceptualization, methodology, writing—review & editing, supervision; Chunyu Zhu: formal analysis, visualization, writing—original draft; Yi Gan: investigation, software; Jianbo Chen: investigation, software; Jiantao Miao: visualization, software, data curation.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The data that has been used is confidential.