Contaminant removal effectiveness evaluation of ventilation, UR-UVGI, and air cleaner using CFD in negative pressure isolation ward

. A negative pressure isolation ward (NPIW) is designed to create negative pressure to prevent the spread of airborne viruses. Therefore, it is necessary to increase the contaminant removal effect in the NPIW. To reduce indoor air contaminants, upper room-ultraviolet germicidal irradiation (UR-UVGI) and air cleaner are being used. Unlike the UR-UVGI, the air cleaner generates airflow. Therefore, the contaminant removal effect may differ according to the ventilation methods. This study is to compare ventilation, UR-UVGI, and air cleaner contaminant removal effects using CFD and Contaminant Removal Effectiveness (CRE) in the NPIW. As a result, the CRE of the ceiling and wall exhaust methods were different, and the CRE of the UR-UVGI and air cleaner cases were also different according to the two ventilation methods. Therefore, when applying UR-UVGI and air cleaner, it is necessary to consider the ventilation method.


Introduction
Casualties occurred due to severe acute respiratory syndrome (SARS) in 2003, middle east respiratory syndrome (MERS) in 2015, and coronavirus disease (COVID-19) in 2019. COVID-19 caused about 610 million people (about 6 million deaths) casualties worldwide as of December 8, 2022 [1]. Infected people are quarantined in healthcare facilities to prevent the spread of infectious diseases. When isolating infected people, the healthcare facility used is a negative pressure isolation ward (NPIW). The NPIW is designed to form negative pressure step by step from the corridor to the toilet to prevent the spread of airborne viruses (Fig. 1.). However, the access of medical staff for medical treatment causes door openings, which can break the pressure difference and spread airborne viruses to adjacent rooms. Therefore, it is necessary to reduce the concentration of airborne viruses in the NPIW to prevent the spread of infection. The NPIW is designed with high air change per hour to reduce the airborne virus concentration. Also, exhaust diffusers are placed near the patient to exhaust airborne viruses as quickly as possible. The exhaust diffusers are  [2]. Qian et al. analyzed the contaminant removal effect of three ventilation methods in a multibed airborne infection isolation room [3]. However, previous studies did not consider additional devices installed in hospital ward to reduce the concentration of airborne viruses. Additional devices installed in the NPIW include the upper room-ultraviolet germicidal irradiation (UR-UVGI) system and air cleaner. The UR-UVGI system sanitizes the upper part of the room. However, due to the location of exhaust diffusers, contaminants might not spread to the upper part of the room, which can lower the contaminant removal efficiency. Unlike the UR-UVGI system, the air cleaner generates airflow, so it may affect the room airflow which was planned to reduce contaminants effectively. Bang et al. analyzed the contaminant reduction effect according to the location of the UR-UVGI system in NPIW [4]. Qian et al. confirmed the particle matter removal effect according to the airflow rate of the air cleaner and the operation of the heating, ventilation, and air conditioning system [5]. However, the location of the exhaust diffusers was not considered when analyzing the contaminant reduction effect of the additional device. This study aims to compare the contaminant removal efficiency according to the exhaust methods (ceiling and wall exhaust). In addition, the contaminant removal efficiency when installing the UR-UVGI system and air cleaner for each exhaust method is compared.

Method
This study was conducted at the NPIW experimental chamber in Sejong University, Seoul, Korea. The contaminant removal efficiency analysis intends to use contaminant removal effectiveness (CRE), an index for evaluating indoor ventilation efficiency. In addition, computational fluid dynamics (CFD) was used to overcome experimental limitations and observe the distribution of contaminants throughout the NPIW. Fig. 2. is a domain of the NPIW experimental chamber for CFD simulation. One supply and one exhaust diffuser are located on the ceiling in the anteroom. The makeup supply diffuser in the anteroom is installed to simulate the air flowing into the NPIW from the adjacent room. One supply diffuser is located on the ceiling in the ward. In the ceiling exhaust method (CEM), an exhaust diffuser is located on the ceiling close to the patient. In the wall exhaust method (WEM), two exhaust diffusers are located on the wall near the patient. The CFD analysis conditions are shown in Table 1. The supply and exhaust airflow rate were measured when the air handling unit (AHU) in the experiment chamber was operating. Air change per hour in the ward meets 6 ACH or more based on supply airflow rate. The pressure difference between the anteroom and the ward is more than -2.5 Pa. This CFD analysis was set as a passive scalar, assuming that airborne viruses are particles with a diameter of 5 μm or less capable of airborne propagation. Therefore, this study assumes that airborne viruses will follow the airflow pattern. The contaminant was generated at a concentration of 1 for relative comparison because the concentration of the airborne viruses could not be measured. The CFD analysis proceeded with steady-state analysis.  The CFD analysis are six cases as shown in Table 2. The analysis was conducted by dividing CEM and WEM. In addition, in the case where no additional device was installed, only the UR-UVGI system and the air cleaner were installed for each CEM and WEM, which were then analyzed.

UR-UVGI system
The UR-UVGI system utilized the device that was applied to Bang et al., and the method of calculating the UV intensity distribution and the CFD analysis method of the UR-UVGI are the same as those proposed by Bang et al. [4]. The UV sterilization coefficient is 0.3512 m²/J of coronavirus proposed in Walker and Ko [6]. The UV intensity at 1 m from the UR-UVGI system is 0.99 W/m². The installation location was installed on the upper part of the patient, as shown in Fig. 3, and its height was 2.3 m.

Air cleaner
The air cleaner is Blueair's Classic 290i, and the specification is shown in Table 3. In the CFD analysis, the low airflow rate (128 m 3 /h) was applied. The air inlet is at the front of the air cleaner, and the air outlet is at the top of the air cleaner. Air is supplied at 26.2 degrees, not perpendicular to the air outlet surface. The filter of the air cleaner removes 99.97% of 0.1 μm particles. Therefore, it was set to be supplied at 0.03% of the contaminant concentration inhaled in CFD. The location of the air cleaner was installed near the door, as shown in Fig. 3., to reduce the contaminant concentration near the door.

CRE
The CRE is a useful the ability of contaminants removal when the location of contaminant source is known. Therefore, when the location of the contaminant source (= patient) is known, such as in NPIW, the ability of contaminant removal can be confirmed through the CRE. The CRE is calculated as: where Ce is the contaminant concentration in the exhaust air, Ci is the mean contaminant concentration in the room, and Cs is the contaminant concentration in the supply air. If Cs is zero, the CRE is the ratio of Ce to Ci. show the airflow and temperature distribution in the ward. A cross section of the airflow distribution is located at the patient's mouth, and a cross section of the temperature distribution is located at the center of the patient's body. The airflow in CEM is an updraft. The air generated from the patient's mouth moves to the upper part of the room and then to both sides of the wall. The airflow in WEM is also updraft like that in CEM. However, the air generated from the patient's mouth does not rise vertically but instead curves to the patient's left side and moves clockwise. In the temperature distribution of the CEM, the heat generated by the patient rises vertically, like airflow distribution. Therefore, the temperature of the left and right sides of the patient is relatively lower than that of the upper part of the ward. The temperature distribution of the WEM, like the airflow distribution, moves the heat generated from the patient to the left and rises clockwise.  Comparing the concentration distributions of Case C1 and W1, in which no additional devices are installed, it can be confirmed that Case C1 has a lower concentration than Case W1. In the WEM, the contaminants generated by the patient are not exhausted through the exhaust diffusers next to the patient but moved to the upper part of the ward by an updraft. Therefore, the contaminant concentration in the entire ward of the WEM is higher than that of the CEM. The contaminant concentration in the case where the UR-UVGI system and air cleaner are installed is lower than in Case C1 and W1. In Case C2 and W2 (installed UR-UVGI system), it can be confirmed that the contaminant concentration is significantly lower than in Case C1 and W1. In Case C3 and W3 (installed air cleaner), the contaminant concentration was higher than in Case C2 and W2, but lower than in Case C1 and W1. The mean contaminant concentration in the ward was calculated to quantitatively confirm the contaminant concentration. Fig. 7. is a graph showing the value multiplied by 10 4 of the mean contaminant concentration in ward by cases. Case W1 has 1.99 times higher mean contaminant concentration in the ward than Case C1. The contaminant reduction rate of Case C2 and C3 compared to Case C1 are 81.1 % and 34.5 %, respectively. The contaminant reduction rate of Case W2 and W3 compared to Case W1 are 75.5 % and 15.9 %, respectively. The contaminant reduction rate of the CEM is 2.35 times higher UR-UVGI system than air cleaner. The contaminant reduction rate of the WEM is 4.75 times higher than that of the UR-UVGI system compared to the air cleaner. When the UR-UVGI system was installed on the CEM and WEM, the contaminant reduction rate was the highest.

CRE by cases
The CRE calculated by cases is shown in Table 4. Ce of Case C1 and W1 are similar, but Ci is lower in Case C1. Therefore, the CRE of Case C1 is twice as high as that of Case W1. In Case C2 and W2, Ci was significantly lowered by the UR-UVGI system. However, Ce was also lowered, so the CRE was calculated lower than Case C1 and Case W1. In Case C3 and W3, where the air cleaner was installed, Ce and Ci decreased compared to Case C1 and W1. However, the reduction rate of Ci was greater than that of Ce, so the CRE was calculated higher than in Case C1 and W1.

Distribution of airflow and temperature
In this analysis domain, the ceiling exhaust is installed near the patient. In the patient's mouth, 37°C air is supplied at 0.39 m/s (0.5 m 3 /h), and the patient is set to have a thermal condition specification. Therefore, in the CEM, the air flow rises like a 'T' shape due to exhaust airflow and convection caused by the temperature and moves to both sides of the wall. The exhaust diffusers of the WEM are installed near the patient but are located below the patient. Therefore, the contaminant generated from the patient's mouth is not directly exhausted through the exhaust diffusers but instead rises. However, the air flow does not rise vertically because the exhaust diffusers are located below the patient. Under the influence of the air supply diffuser on the door side and the inflowing air, the updraft moves clockwise. A temperature distribution appears like the airflow distribution formed in the ward.

Distribution of contaminant concentration
The CEM and WEM create updrafts. Contaminants from the patient's mouth are directed to the upper part of the room along the formed updraft. The CEM can directly exhaust contaminants that have migrated to the upper part of the ward. However, the WEM has exhaust diffusers located at the bottom of the wall, so contaminants that have moved to the upper part of the room cannot be immediately exhausted. Therefore, the WEM spreads contaminants throughout the ward more than the CEM. However, in this study, contaminants were assumed to be gaseous matter and analyzed. It is expected that different results will be obtained if the contaminants are assumed to be particulate matter.
The contaminant reduction effect of the UR-UVGI system is better for the CEM than the WEM. In the CEM, contaminants generated by the patient move directly to the area irradiated with UV, but in the WEM, contaminants move along the airflow to the left of the patient and move clockwise. In addition, the WEM has a faster airflow speed at the upper part of the ward than the CEM. Therefore, the residence time of contaminants in the UV irradiation area of the WEM is shorter than that of the CEM. For this reason, it is considered that the contaminant reduction effect of the UR-UVGI system in the CEM is better than that of the WEM. The UR-UVGI system showed a better contaminant reduction effect than the air cleaner. In this study, the air cleaner was installed near the door to reduce contaminants near the door. Therefore, since the air cleaner is far from the patient, it is considered that the contaminant reduction effect of the air cleaner is not high. The contaminant reduction effect according to the location of the air cleaner will be confirmed in a future study.

CRE by cases
According to the results of this study, the CRE in cases where the UR-UVGI system was installed was lower than that in cases where only ventilation was operated.
On the other hand, the CRE in cases where the air cleaner was installed was higher than that in cases where only ventilation was operated. The CRE is used as an indicator of ventilation efficiency, but when devices such as the UR-UVGI system and air cleaner are installed, the CRE can vary even in the same space and ventilation conditions. After the UR-UVGI system was installed, the contaminant concentration in the ward was greatly reduced, but the CRE was lower than before the installation. Contaminants generated by the patient are moved to the upper part of the ward and directly reduced by the UR-UVGI system located above the patient. As the contaminants near the patient are reduced, the contaminant concentration in the air exhausted through the exhaust diffusers installed near the patient is reduced more than the contaminant concentration in the ward. Therefore, when the UR-UVGI system is installed, it is considered that the CRE is lower than when only ventilation is operated. The contaminant concentration in the ward decreased less when the air cleaner was installed than with the UR-UVGI system. However, the CRE after the air cleaner installation was higher than before the installation and when the UR-UVGI was installed. Since the air cleaner is installed near the door, it is difficult to reduce contaminants near the patient, and contaminant concentration in the ward are reduced through recirculation. Therefore, when the air cleaner is installed, it is judged that the CRE is higher than when only ventilation is operated.
As a result of this study, the UR-UVGI system and air cleaner were fixed at one point. However, if the location of the UR-UVGI system and air cleaner is different, it is expected that the results of this study will also be different. Therefore, the CRE according to the location of the UR-UVGI system and air cleaner will be confirmed in the future study.
Using the CREratio, it is possible to determine which of the contaminant concentration reduction rate in the exhaust air and the mean contaminant concentration reduction rate in the ward is higher, according to the additional device. The CREratio is the ratio of the CRE for a ward without additional devices to the CRE for a ward with additional devices. The CRE ratio is calculated as:  (3) and (4). When Re is low and Ri is high, the CREratio is high. Therefore, if the CREratio is greater than 1, it means that Ri is greater than Re, and if it is less than 1, Re is greater than Ri.

Conclusion
In this study, the ventilation method in the NPIW was divided into the CEM and WEM, and the contaminant removal effect was compared. In addition, contaminant removal effects were compared when the UR-UVGI system and air cleaner were installed for each exhaust method. The comparison results are as follows: 1) The UR-UVGI system and air cleaner are effective in removing contaminants.
2) The UR-UVGI system is more effective in reducing contaminants than air cleaner. 3) The CEM showed 5.6 % and 18.6 % higher contaminant reduction rate than the WEM when UR-UVGI system and air cleaner were installed, respectively. 4) When the contaminant is assumed to be gaseous matter, the CRE of the CEM is twice as high as that of the WEM. 5) When additional devices are installed, the CRE is different even in the same space and ventilation conditions. 6) By checking whether CREratio is greater than or less than 1, it is possible to determine which of Re (the contaminant concentration reduction rate in the exhaust air) and Ri (the mean contaminant concentration reduction in the ward) is higher.
When applying the UR-UVGI system and air cleaner, it is necessary to consider the ventilation method. In addition, if additional devices are installed in the ward, a low CRE does not mean low ventilation efficiency. In future studies, the contaminant reduction effect and CRE will be confirmed by considering the location of the UR-UVGI system and air cleaner, the air change per hour, and the pressure difference in the ward.