Contents

Page

Abstract

1. Introduction

2. Methods

2.1 Selection of mobile air cleaning devices

2.1.1 Collection of information on available mobile air cleaning devices

2.1.2 Major specifications of mobile air cleaning devices

2.1.3 Final selection of mobile air cleaning devices

2.2 Assessment of mobile air cleaning devices

2.2.1 Particle decay test

2.2.2 Panel perception test

3. Results and Discussion

3.1 Results of each tested air cleaning device

3.2 Discussion

3.2.1 Particle removal rate and CADR

3.2.2 Sound pressure level and sound perception

3.2.3 Air velocity and air movement perception

4. Conclusions

References

Symbols and abbreviations

Appendix A Forms for perception tests

2. Methods

2.1 Selection of mobile air cleaning devices

2.1.1 Collection of information on available mobile air cleaning devices
To collect information on the availability of mobile air cleaning devices applicable to school classrooms,
existing products were searched within two ranges:

1) Professional organizations or associations, e.g., ISIAQ, ASHRAE, REHVA, etc.
2) Commercial dealers, e.g., Coolblue, MediaMarkt, BCC, Bol.com, Amazon.nl.

In total more than 300 products were found, and were further screened based on the following criteria,
which resulted in a preliminary list of 152 mobile air cleaning devices:

1) The manufacturer has its own official (professional) website.
2) The manufacturer’s main business is air handling products or it has more than one type of air

cleaning product.

3) The main air cleaning technology of the product is HEPA and/or electrostatic and/or plasma

(can also be combined with activated carbon and/or UV-C).

4) A list of technical specifications is provided for the product.
5) The product is available in or can be bought from the Netherlands.

2.1.2 Major specifications of mobile air cleaning devices
The 152 products were then categorized and compared using eight parameters, based on the
specifications provided by the manufacturer:
1) Air cleaning technology: The selected mobile air cleaning devices cover five main types of air

cleaning technology, namely: HEPA, EPA, electrostatic (ES), plasma (PL), activated carbon (AC),
and UV-C, which resulted in 15 types of combinations, as shown in Figure 1.

2) Airflow configuration: Airflow configuration of the mobile air cleaning device refers to the

configuration of how room air is sucked in and supplied by the device (i.e., the configuration of
airflow inlet and outlet). Figure 2 presents the 26 types of airflow configurations of the 152 products.
The most common configuration of airflow inlet and outlet is horizontally from the side, and
vertically from the top of the device, respectively.

3) Efficiency: The classification of the efficiency of air cleaning devices prescribed in the European

standard EN 1822-1:2019 [4] is presented in Table 1. According to the specifications provided by
the manufacturers, the particle removal efficiency of all the selected air cleaning devices is E12 or
higher, which is equivalent to a removal efficiency of 0.3 µm particles of ≥ 99.95%.

4) Fan capacity (or the CADR): The fan capacity means the maximum airflow rate the air cleaning

device can provide, usually in m3/h. Most of the selected air cleaning devices have multiple settings
(fan levels). For some devices, instead of fan capacity, CADR (clean air delivery rate) is specified,
which equals to the particle removal rate multiplied by the airflow rate, or the decrease of the
particle concentration multiplied by the room volume, and thus indicates both the efficiency and
fan capacity of the device. Since all the selected devices have an efficiency higher than 99.95%, the
CADR can thus be considered as approximately equal to the fan capacity.

5) Sound level: The sound (noise) level of the mobile air cleaning devices varies with the fan settings,

which can range from lower than 18 dB(A) to higher than 60 dB(A).

6) Dimensions: Normally the fan capacity of the mobile air cleaning device increases with the size.

However, the devices should also be able to fit in the school classrooms causing minimum hinder
to the teaching and learning activities.

7) Maintenance: The maintenance of the mobile air cleaning devices includes most importantly, the

filter life, and its cost. The additional AC filter may also add to the cost, however for many produces
the AC filter is combined with the main filter.

8) Price.

2.1.3 Final selection of mobile air cleaning devices
To select the proper mobile air cleaning devices for further testing, several criteria have been determined
considering both the technical requirements and feasibility of application for operating such devices in
school classrooms:
1) Considering the type of tests being performed on the devices in this project, as well as the

immatureness of real-life application, the mobile air cleaning devices that uses UV-C technology
were excluded.

2) To ensure better particle removal, H13 air filtering (and higher) is widely recommended. Hence,

the mobile air cleaning devices with an efficiency lower than H13 were excluded.

3) According to the Dutch Fresh Schools guideline [5], the noise level of installations in school

classrooms should be ≤ 35 dB(A). Therefore, the mobile air cleaning devices with a minimum noise
level higher than 35 dB(A) were excluded.

4) For ventilation in school classrooms, a ventilation rate of 8~10 L/s per person is recommended to

ensure occupants’ health and comfort [6]. Taking the student occupancy in a typical classroom as
30 persons, the total required ventilation rate is thus around 900~1000 m3/h. Therefore, when
operating the mobile air cleaning devices, the clean air delivery should be able to achieve such level.
With regards to the size and fan capacity of the products, it was determined that the maximum units
of devices that can be operated in one classroom is four, where each device should provide at least
250 m3/h of clean air. Hence, the mobile air cleaning devices with a maximum fan capacity lower
than 250 m3/h were excluded.

5) Considering the affordability of the schools, the total budget of mobile air cleaning devices in one

classroom is set to be 3000 euros. Thus, by multiplying the commercial unit price and number of
units needed in one classroom, those who reached a total cost higher than 3000 euros were excluded.

Besides the aforementioned criteria, further selection was made by reducing the number of products
from the same brands, as well as assessing the feasibility of the airflow configuration, which led to a
shortlist of 27 products. The whole process is illustrated in Figure 3.

In the end, eight types of mobile air cleaning devices (noted as AP) were selected, based on another
round  of  overall  comparison.  Each  of  these  eight  devices  presents  a  different  combination  of  air
purifying  technology,  airflow  configuration  and  fan  capacity. The  detailed  information  are  listed  in
Table 2. It is worth noting that considering the fan capacity of AP5, one unit is required; while for AP1,
AP4, AP6, and AP7, two units are required; and for AP2, AP3, and AP8, four units are required. The
manufacturers of the devices were reached out after the selection was completed, yet until the end of
the project the products of AP8 were not delivered, and thus had to be excluded from the study.

Table 2. Specifications of the selected mobile air cleaning devices.

Air cleaning device

Technology

Airflow configuration Fan capacity (m3/h) Efficiency Sound [dB(A)] Dimensions (cm)

Maintenance

Unit Price (€)

AP1

HEPA + AC

1000

H13

Max. 62

19.0 × 19.6 × 101.8

Filter life: 4380 h
Filter price (€): 75
AC life: combined
AC price (€): -

500

AP2

HEPA + ES + AC

610

H13

19-57

30.6 (Φ) × 70.5

Filter life: 24 months
Filter price (€): 100
AC life: combined
AC price (€): -

480

AP3

330

H13

19-53

27.0 × 27.0 × 50.0

Filter life: 10 years
Filter price (€): -

380

AP4

ES + AC

735

H13

27-55

34.0 × 34.0 × 85.5

Filter life: 12 months
Filter price (€): 160
AC life: combined
AC price (€): -

1100

AP5

ES + AC

1386

H13

33-49

38.0 × 38.0 × 76.0

Filter life: 12 months
Filter price (€): -
AC life: -
AC price (€): -

1900

AP6

HEPA

565

H13

18-51

33.2 x 33.6 x 60.6

Filter life: 12 months
Filter price (€): 90

500

AP7

HEPA

750

H13

26-65

68.8 (Φ) x 25.4

Filter life: 18-24
months
Filter price (€): 250

1500

AP8

HEPA + PL + AC

250

H13

27-54

24.1 (Φ) × 37.0

Filter life: 12 months
Filter price (€): 30

100

2.2 Assessment of mobile air cleaning devices
The assessment of the final selected air cleaning devices consisted of two parts, namely the particle
decay test and the panel perception test. The panel perception test also included physical measurements
of sound pressure level and air velocity. All the tests were conducted during May to July 2023, in the
Experience room of the SenseLab at the Delft University of Technology [7]. The Experience room has
a size of 6.1 (l) × 4.2 (b) × 2.7 (h) m3, with two windows and one door, and the interior is set as a
classroom, with tables, chairs, and a smartboard. The Experience room is equipped with both mixing
and displacement ventilation systems, with a maximum ventilation rate of 1500 m3/h, and openable
windows. The ventilation system is also equipped with a HEPA14 filter.

2.2.1 Particle decay test

Aerosol generator
An aerosol generator was adopted as the source of particles for the decay test, using an artificial saliva
liquid made of 98.5 wt% water + 1 wt% glycerine + 0.5 wt% NaCl (salt). The aerosol generator consists
of  an  HPLC  pump  (model:  SHIMADZU  LC-10AD),  a  PulmosprayTM  (Medspray),  and  an  air
compressor [8]. The PulmosprayTM contains a nozzle (spray chip), a liquid tube, and an air tube. When
the aerosol generator operates, the liquid is pumped from a bottle to the nozzle via the HPLC pump at
a flow rate of 0.8~0.9 ml/min. It becomes multiple parallel liquid jets passing the nozzle and then break
up into droplets, which are then mixed with the co-airflow to form a constant aerosol spray. The aerosol
generator was placed in the middle of the Experience room, with the spray facing the front of the room
(Figures 4 and 5).

Measurement instruments
The concentrations of particulate matter, PM2.5 and PM10, were measured by six SDS011 air quality
sensors, which were evenly placed on six tables in the Experience Room (Figure 5). The concentrations
of  VOC  (Volatile  organic  comppunds)  and  O3  (Ozone)  were  measured  by  a  Kanomax  Gasmaster
monitor  (model:  2750)  and  an Aeroqual  O3  monitor  (model:  Series  500),  respectively.  These  two
monitors were placed on one table, close to the center of the room (Figure 5).

Figure 4. Set-up of the aerosol generator.

Test conditions and procedure
To determine the settings for the formal tests, for each type of mobile air cleaning device, a pre-test was
performed to check the noise level of each setting. With the outcome two settings were selected for each
type of device, one being the highest possible setting with a noise level lower than 35 dB(A), and the
other being the maximum setting, noted as S1 and S2, respectively. Based on the number of units needed,
each type of device was also tested under different configurations (layouts in the room), as shown in
Figure 6. For the device with one unit, configuration 1 (C1) was with the device placed in the middle at
the back of the room, and configuration 2 (C2) in the front, slightly on the side to avoid blocking the
smartboard. For the device with two units, C1 was with one unit next to the smartboard in the front and
the other in the middle at the back, C2 was diagonally in two corners. For the device with four units,
there was only one configuration (C1), namely at four corners of the room. The total conditions tested
for each type of device are listed in Table 3.

Table 3. Tested conditions for the selected mobile air cleaning devices.

Mobile air cleaning device Tested settings*

Tested configurations

Number of conditions

AP1

S1 (L4), S2 (L10)

C1, C2

AP2

S1 (L1), S2 (L2)

AP3

S1 (L2), S2 (L4)

AP4

S1 (L1), S2 (L2)

C1, C2

AP5

S1 (L1), S2 (L5)

C1, C2

AP6

S1 (L4), S2 (L8)

C1, C2

AP7

S1 (L4), S2 (L8)

C1, C2

* S: setting. L: device setting level. C: configuration.

Each decay test consists of a build-up session and a decay session. For the build-up, the aerosol
generator was turned on for at least 1 h, to achieve a well-mixed condition of particles inside the room.
Two floor fans were used to help accelerating the process (Figure 5). When the particle concentrations
reached steady state, the build-up was completed and the aerosol generator and the fans were turned off,
while the mobile air cleaning device(s) was(were) turned on simultaneously, after which the decay
started. The decay session normally lasted for 1.5 h. During the test, everything was set to be remotely
controlled except for AP5. Before and after each test, the mixing ventilation was turned on at an airflow
rate of 1200 m3/h to help maintaining the particle concentrations in the room at a very low level (< 1
µg/m³).

Particle removal rate and CADR
The particle concentration decay can be described by equation (1)：

𝐶(𝑡) = 𝐶!𝑒"#$

(1)

where:

• 𝐶 is the particle concentration, µg/m³

• 𝐶! is the initial concentration of the decay process, µg/m³

• 𝑘 is the decay coefficient (here also the particle removal rate), h-1

• 𝑡 is the time after decay starts, h

Hence, the particle removal rate

𝑘 of the mobile air cleaning devices can be determined using non-linear

regression, here performed by IBM SPSS v.28. The CADR of the mobile air cleaning devices can be
calculated by equation (2):

𝐶𝐴𝐷𝑅 = 𝑘𝑉

(2)

where:

• 𝑘 is the particle removal rate, h-1

• 𝑉 is the volume of the room, m3

2.2.2 Panel perception test

Subjects and questionnaires
Eight PhD students (four males and four females, aged from 28 to 35 years) were recruited as subjects
for the perception tests. For each perception test, a panel of six subjects (three male and three female)
was formed to assess the sound and air movement created by the mobile air cleaning devices. For both
sound and air movement, the subjects were asked to vote for the perception (feel/not feel), as well as
intensity (quiet/loud, mild/strong) and assessment (pleasant/unpleasant) in the case they did sense any
sound  and/or  air  movement.  For  air  movement,  an  extra  question  was  added  to  ask  the  subjects  to
specify on which body part(s) they sensed the air movement (if there was any). The forms used are
presented in Appendix A.

Test conditions and procedure
Based on the results of the decay test, the configuration with a higher particle removal rate and CADR
was selected for each mobile air cleaning device to perform the perception test, combined with both
settings. For AP5, AP6, and AP7, the difference between the two configurations was not significant,
and thus both configurations were included for the perception test. When the subjects arrived, they were
asked to rest for 10 minutes while completing a general information form to report the clothing they
wore and their mood at the moment, after which they were seated at six tables in the Experience room
(same as the decay test) (Figure 7). Each type of device was tested during an independent session, which
started with an acclimatization period, and then the real test conditions began with the device(s) being
turned on. Each condition lasted for 6 minutes, during which the device(s) was(were) turned on for 5
minutes,  and  then  turned  off  during  the  last  minute.  The  subjects  were  asked  to  complete  the
questionnaire at the 4th minute. The conditions were randomly assigned for each type of device, and
after each session there was a break for switching the devices and preparing for the next session.

Figure 7. Set-up of the perception test.

Measurements of sound pressure level and air velocity
During the perception test, the sound pressure level was measured using a Norsonic sound analyzer
(model: Nor140), which was placed in the front of the room (Figure 7). Both sound pressure level and
air velocity were measured for the same conditions when the room was empty. For the sound pressure
level, the sound analyzer was placed at the same location as shown in Figure 7, and each measurement
lasted for 2 minutes. For the air velocity, a Dantec ComfortSense anemometer (model: 54T035) was
used, which was placed in front of each table where the subjects were seated, at a height of 1.1 m, and
each measurement lasted for 3 minutes.

3. Results and discussion

3.1 Results of each tested air cleaning device
Tables 4 to 10 present the detailed results of the tested mobile air cleaning devices (AP1-AP7).

Table 4. Results of AP1.

Condition

S1_C1

S2_C1

S1_C2

S2_C2

kPM2.5 (h-1) [mean (SD)]

2.216 (0.030)

5.061 (0.079)

2.862 (0.040)

5.234 (0.079)

CADRPM2.5 (m3/h) [mean (SD)]

153.3 (2.1)

350.1 (5.2)

198 (2.7)

362.1 (5.5)

kPM10 (h-1) [mean (SD)]

2.224 (0.055)

5.099 (0.095)

2.922 (0.092)

5.200 (0.172)

CADRPM10 (m3/h) [mean (SD)]

153.8 (3.8)

352.7 (6.6)

202.1 (6.3)

359.7 (11.9)

VOC (ppm) [mean (SD)]

5.1 (0.8)

5.5 (1.4)

4.7 (0.8)

4.0 (0.2)

O3 (ppm) [mean (SD)]

0.034 (0.003)

0.033 (0.004)

0.041 (0.002)

0.037 (0.003)

SPLempty [dB(A)] [mean (SD)]

32.4 (0.3)

47.9 (0.1)

SPLpanel [dB(A)] [mean (SD)]

34.2 (2.2)

47.9 (0.1)

Sound perception (%)

100

Sound intensity [mean (SD)]

2.0 (1.3)

4.3 (0.5)

Sound assessment [mean (SD)]

2.2 (1.0)

4.5 (0.8)

Sound unacceptability (%)a

Air velocity at 1.1 m (m/s) [mean (SD)]

0.006 (0.004)

0.006 (0.005)

Air movement perception (%)

Air movement body part(s)b

face(1); neck(1) face(1); neck(2)

Air movement intensity [mean (SD)]

1.0 (0)

2.0 (1.7)

Air movement assessment [mean (SD)]

1.5 (0.7)

2.3 (1.2)

Air movement unacceptability (%)c

a Percentage of sound assessment >3. b The numbers indicate number of subjects. c Percentage of air movement assessment >3.

Table 5. Results of AP2.

Condition

S1_C1

S2_C1

kPM2.5 (h-1) [mean (SD)]

8.684 (0.169)

13.562 (0.734)

CADRPM2.5 (m3/h) [mean (SD)]

600.7 (11.7)

938.1 (50.8)

kPM10 (h-1) [mean (SD)]

8.825 (0.763)

13.682 (1.117)

CADRPM10 (m3/h) [mean (SD)]

610.5 (52.8)

946.5 (77.2)

VOC (ppm) [mean (SD)]

1.4 (0.5)

3.6 (0.9)

O3 (ppm) [mean (SD)]

0.038 (0.005)

0.035 (0.001)

SPLempty [dB(A)] [mean (SD)]

33.1 (1.8)

43.3 (0)

SPLpanel [dB(A)] [mean (SD)]

34.3 (3.2)

45.7 (2.9)

Sound perception (%)

100

Sound intensity

2.2 (1.2)

3.7 (0.8)

Sound assessment

2.5 (1.0)

3.5 (1.0)

Sound unacceptability (%)a

Air velocity at 1.1 m (m/s) [mean (SD)]

0.025 (0.011)

0.077 (0.026)

Air movement perception (%)

Air movement body part(s)b

face (2); hand (1); arm (1)

face (3); neck (1); hand (2)

Air movement intensity

1.0 (0)

1.4 (0.5)

Air movement assessment

1.8 (0.5)

2.0 (0.7)

Air movement unacceptability (%)c

a Percentage of sound assessment >3. b The numbers indicate number of subjects. c Percentage of air movement assessment >3.

Table 6. Results of AP3.

Condition

S1_C1

S2_C1

kPM2.5 (h-1) [mean (SD)]

7.097 (0.707)

15.715 (0.369)

CADRPM2.5 (m3/h) [mean (SD)]

490.9 (48.9)

1087.1 (25.5)

kPM10 (h-1) [mean (SD)]

7.753 (0.936)

16.673 (0.950)

CADRPM10 (m3/h) [mean (SD)]

536.3 (64.8)

1153.3 (65.7)

VOC (ppm) [mean (SD)]

4.6 (0.5)

4.9 (0.4)

O3 (ppm) [mean (SD)]

0.040 (0.003)

0.044 (0.004)

SPLempty [dB(A)] [mean (SD)]

29.9 (0.5)

49.7 (0.1)

SPLpanel [dB(A)] [mean (SD)]

34.3 (0.2)

50.6 (1.5)

Sound perception (%)

100

Sound intensity

1.8 (0.8)

4.3 (0.8)

Sound assessment

2.2 (1.3)

4.0 (0.9)

Sound unacceptability (%)a

Air velocity at 1.1 m (m/s) [mean (SD)]

0.001 (0.001)

0.044 (0.029)

Air movement perception (%)

Air movement body part(s)b

face (2); chest (1)

Air movement intensity

1.5 (0.7)

Air movement assessment

2.0 (0)

Air movement unacceptability (%)c

a Percentage of sound assessment >3. b The numbers indicate number of subjects. c Percentage of air movement assessment >3.

Table 7. Results of AP4.

Condition

S1_C1

S2_C1

S1_C2

S2_C2

kPM2.5 (h-1) [mean (SD)]

5.935 (0.325)

14.252 (0.436)

6.119 (0.264)

14.359 (0.414)

CADRPM2.5 (m3/h) [mean (SD)]

410.5 (22.5)

985.9 (30.1)

423.3 (18.3)

993.3 (28.6)

kPM10 (h-1) [mean (SD)]

5.820 (0.530)

13.844 (1.181)

6.074 (0.503)

14.151 (0.962)

CADRPM10 (m3/h) [mean (SD)]

402.6 (36.7)

957.7 (81.7)

420.2 (34.8)

978.9 (66.5)

VOC (ppm) [mean (SD)]

4.5 (0.7)

4.7 (0.5)

5.0 (0)

4.9 (0.3)

O3 (ppm) [mean (SD)]

0.034 (0.002)

0.035 (0.002)

0.036 (0.002)

SPLempty [dB(A)] [mean (SD)]

28.3 (0.5)

45.0 (0)

SPLpanel [dB(A)] [mean (SD)]

30.5 (1.2)

45.4 (2.0)

Sound perception (%)

100

Sound intensity [mean (SD)]

1.3 (0.8)

3.7 (1.5)

Sound assessment [mean (SD)]

1.5 (0.8)

3.5 (1.0)

Sound unacceptability (%)a

Air velocity at 1.1 m (m/s) [mean (SD)]

0.004 (0.003)

0.043 (0.027)

Air movement perception (%)

Air movement body part(s)b

face (2); chest
(1);arm (1);
hand (2)

Air movement intensity [mean (SD)]

1.8 (0.5)

Air movement assessment [mean (SD)]

2.3 (1.0)

Air movement unacceptability (%)c

a Percentage of sound assessment >3. b The numbers indicate number of subjects. c Percentage of air movement assessment >3.

Table 8. Results of AP5.

Condition

S1_C1

S2_C1

S1_C2

S2_C2

kPM2.5 (h-1) [mean (SD)]

6.294 (0.277)

10.809 (0.255)

7.484 (0.913)

13.190 (1.189)

CADRPM2.5 (m3/h) [mean (SD)]

435.4 (19.1)

747.7 (17.6)

517.7 (63.1)

912.4 (82.3)

kPM10 (h-1) [mean (SD)]

6.136 (0.380)

10.774 (0.864)

7.867 (1.070)

14.446 (1.388)

CADRPM10 (m3/h) [mean (SD)]

424.4 (26.3)

745.3 (59.8)

544.2 (74.0)

999.3 (96.0)

VOC (ppm) [mean (SD)]

3.0 (0.2)

5.0 (0)

4.9 (0.3)

4.2 (0.5)

O3 (ppm) [mean (SD)]

0.040 (0.001)

0.037 (0.001)

0.046 (0.001)

0.046 (0.003)

SPLempty [dB(A)] [mean (SD)]

26.9 (0.4)

46.2 (0.1)

27.2 (0.3)

48.1 (0.4)

SPLpanel [dB(A)] [mean (SD)]

33.4 (5.7)

46.7 (1.5)

29.8 (2.6)

49.5 (3.3)

Sound perception (%)

100

Sound intensity [mean (SD)]

1.7 (1.2)

4.5 (0.5)

1.3 (0.8)

3.2 (1.2)

Sound assessment [mean (SD)]

2.0 (1.3)

4.3 (0.5)

2.0 (1.3)

3.3 (1.0)

Sound unacceptability (%)a

100

Air velocity at 1.1 m (m/s) [mean (SD)]

0.020 (0.009)

0.077 (0.032)

0.011 (0.011)

0.051 (0.034)

Air movement perception (%)

Air movement body part(s)b

face (1)

face(2); head(1)
arm(1) and (3);
thigh(1)

arm (1)

Air movement intensity [mean (SD)]

1.0 (-)

2.3 (1.5)

1.0 (-)

2.0 (-)

Air movement assessment [mean (SD)]

2.0 (-)

2.0 (1.0)

4.0 (-)

3.0 (-)

Air movement unacceptability (%)c

100

a Percentage of sound assessment >3. b The numbers indicate number of subjects. c Percentage of air movement assessment >3.

Table 9. Results of AP6.

Condition

S1_C1

S2_C1

S1_C2

S2_C2

kPM2.5 (h-1) [mean (SD)]

7.370 (0.121)

14.544 (0.590)

5.892 (0.168)

13.370 (0.207)

CADRPM2.5 (m3/h) [mean (SD)]

509.8 (8.4)

1006.1 (40.8)

407.5 (11.6)

924.8 (14.3)

kPM10 (h-1) [mean (SD)]

7.668 (0.226)

15.445 (0.888)

6.112 (0.207)

13.812 (0.509)

CADRPM10 (m3/h) [mean (SD)]

530.4 (15.6)

1068.4 (61.4)

422.8 (14.3)

955.4 (35.2)

VOC (ppm) [mean (SD)]

6.5 (0.5)

5.1 (0.8)

8.2 (0.4)

8.9 (0.3)

O3 (ppm) [mean (SD)]

0.044 (0.001)

0.046 (0.006)

0.037 (0.003)

0.037 (0.002)

SPLempty [dB(A)] [mean (SD)]

29.8 (0.2)

44.6 (0.3)

29.5 (0)

45.3 (0.1)

SPLpanel [dB(A)] [mean (SD)]

31.2 (1.6)

45.6 (1.8)

31.6 (2.0)

44.4 (1.2)

Sound perception (%)

100

Sound intensity [mean (SD)]

1.5 (0.8)

2.7 (0.8)

1.8 (0.8)

4.0 (0.6)

Sound assessment [mean (SD)]

2.0 (1.3)

2.8 (0.8)

2.0 (0.9)

3.8 (0.8)

Sound unacceptability (%)a

Air velocity at 1.1 m (m/s) [mean (SD)]

0.011 (0.006)

0.067 (0.024)

0.033 (0.022)

0.105 (0.037)

Air movement perception (%)

Air movement body part(s)b

face (1); neck
(1); shoulder (1)

face (2); neck
(1); shoulder
(1); hand (1)

face (3); head
(1); neck (1);
hand (2); thigh
(1); ankle (1)

Air movement intensity [mean (SD)]

1.3 (0.6)

1.5 (0.6)

1.6 (0.9)

Air movement assessment [mean (SD)]

2.0 (1.0)

1.8 (1.0)

1.6 (0.5)

Air movement unacceptability (%)c

a Percentage of sound assessment >3. b The numbers indicate number of subjects. c Percentage of air movement assessment >3.

Table 10. Results of AP7.

Condition

S1_C1

S2_C1

S1_C2

S2_C2

kPM2.5 (h-1) [mean (SD)]

9.527 (0.157)

19.701 (0.871)

11.248 (0.327)

19.284 (0.830)

CADRPM2.5 (m3/h) [mean (SD)]

659.0 (10.8)

1362.8 (60.3)

778.1 (22.6)

1334.0 (57.4)

kPM10 (h-1) [mean (SD)]

8.508 (0.172)

19.701 (0.871)

10.591 (0.343)

19.059 (0.506)

CADRPM10 (m3/h) [mean (SD)]

588.5 (11.9)

1362.8 (60.3)

732.6 (23.8)

1318.4 (35.0)

VOC (ppm) [mean (SD)]

0.6 (0.5)

1.0 (0.0)

1.8 (0.4)

2.0 (0.0)

O3 (ppm) [mean (SD)]

0.046 (0.006)

0.047 (0.006)

0.049 (0.005)

0.046 (0.007)

SPLempty [dB(A)] [mean (SD)]

33.4 (0.2)

49.2 (0.3)

32.1 (0.1)

50.5 (0.2)

SPLpanel [dB(A)] [mean (SD)]

35.6 (2.1)

53.1 (1.5)

35.2 (3.4)

52.3 (2.8)

Sound perception (%)

100

Sound intensity [mean (SD)]

2.7 (0.8)

4.7 (0.8)

1.7 (1.2)

4.2 (1.2)

Sound assessment [mean (SD)]

2.7 (0.8)

4.8 (0.4)

2.2 (1.2)

4.2 (1.0)

Sound unacceptability (%)a

100

Air velocity at 1.1 m (m/s) [mean (SD)]

0.065 (0.027)

0.197 (0.072)

0.094 (0.041)

0.268 (0.119)

Air movement perception (%)

Air movement body part(s)b

face (2); head
(1) ; neck (1);
chest (1); arm
(1); leg (1)

face (2); neck
(1); shoulder
(1); arm (1); leg
(2)

face (1); head
(1); arm (3);
hand (1); leg (1)

face (2); head
(1); neck (1);
arm (3); hand
(1); leg (1)

Air movement intensity [mean (SD)]

1.4 (0.5)

2.6 (0.9)

1.2 (0.4)

2.6 (0.5)

Air movement assessment [mean (SD)]

2.0 (0.7)

1.8 (0.8)

1.6 (0.9)

1.8 (0.4)

Air movement unacceptability (%)c

a Percentage of sound assessment >3. b The numbers indicate number of subjects. c Percentage of air movement assessment >3.

3.2 Discussion

3.2.1 Particle removal rate and CADR
The comparisons of particle removal rate and CADR of both PM2.5 and PM10 among the mobile air
cleaning devices are shown in Figures 8 to 11. As mentioned in section 2.1.3, for a typical classroom, a
clean air delivery rate of 900~1000 m3/h is desired. According to the results, a few of the selected mobile
air cleaning devices can reach such level (marked by the green line in Figures 10 and 11), but only under
the maximum setting, which inevitably cause a noise level way higher than 35 dB(A). However, since
real  classrooms  also  always  involve  vocal  activities  during  teaching  and  learning,  a  slightly  higher
background sound level can be considered acceptable if supported by the results of the perception test,
which  will  be  discussed  in  section  3.2.2.  It  is  also  worth  noting  that  for  AP7,  when  placed  as
configuration 1, the CADR of both PM2.5 and PM10 can almost reach 800 m3/h, which is approximately
7.5 l/s per person (assuming 30 students per classroom), and does fulfill the minimum requirement (6
l/s per person) prescribed by the Dutch Fresh Schools guideline [3].

Figure 11. CADR of PM10.

3.2.2 Sound pressure level and sound perception
The comparisons of sound perception, sound intensity, and sound assessment among the mobile air
cleaning devices are shown in Figures 12 to 14. Under most conditions the sound generated by the
mobile air cleaning devices are perceptible by all the subjects. For AP2, the condition that fulfills the
CADR requirement was S2_C1, of which the average noise level was 43.3 dB(A) (empty room), and
perception-wise the average sound intensity was 3.7 out of 5, with 50% of the subjects rated the sound
as acceptable (Table 5). For AP3, the condition that fulfilled the CADR requirement was S2_C1, of
which the average noise level was 49.7 dB(A) (empty room), and perception-wise the average sound
intensity was 4.3 out of 5, with 33% of the subjects rated the sound as acceptable (Table 6). For AP4,
the conditions that fulfilled the CADR requirement were S2 with both configurations, while S2_C2
showed  a  better  performance,  of  which  the  average  noise  level  was  45.0  dB(A)  (empty  room),  and
perception-wise the average sound intensity was 3.7 out of 5, with 50% of the subjects rated the sound
as acceptable (Table 7). For AP5, the condition that fulfilled the CADR requirement was S2_C2, of
which the average noise level was 48.1 dB(A) (empty room), and perception-wise the average sound
intensity was 3.2 out of 5, with 67% of the subjects rated the sound as acceptable (Table 8). For AP6,
the conditions that fulfilled the CADR requirement were S2 with both configurations, while S2_C1
showed  a  better  performance,  of  which  the  average  noise  level  was  44.6  dB(A)  (empty  room),  and
perception-wise the average sound intensity was 2.7 out of 5, with 83% of the subjects rated the sound
as  acceptable  (Table  9).  For AP7,  when  the  devices  were  operating  at  S1,  the  average  noise  level
remained lower than 35 dB(A) (empty room), and under condition S1_C2 the average sound intensity
was 1.7 out of 5, with 83% of the subjects rated the sound as acceptable. On the other hand, although
S2_C1 achieved the highest CADR, the average noise level was measured to be 49.2 dB(A) (empty
room), with an average sound intensity of 4.7 out of 5, and 100% of the subjects rated the sound as
unacceptable (Table 10).

4. Conclusions

In this project seven different types of mobile air cleaning devices were tested on their feasibility of
applying  them  into  school  classrooms  as  an  alternative  solution  when  ventilation  is  limited.  The
assessment consisted of a particle decay test and a panel perception test, which covered three key factors
of the utilization of mobile air cleaning devices, namely the particle removal ability (CADR), the noise
level, and the risk of draft discomfort.

As  the  results  showed  that  the  air  movement  generated  by  the  devices  has  in  fact  led  to  positive
perception,  the  only  concern  for  selecting  the  proper  device  ended  in  the  tradeoff  between  particle
removal efficiency and noise level. It has been indicated that to achieve the desired amount of clean air
delivery, none of the selected devices could maintain a noise level fulfilling the requirement of lower
than 35 dB(A) [4]. However, considering the results of the perception tests, several configurations can
be considered as acceptable for use in classrooms (see Table 11):

• AP2, with four units of devices operating at device setting L2;

• AP4, with two units of devices operating at device setting L2, configuration 2;

• AP6, with two units of devices operating at device setting L8, configuration 1;

• AP7, with two units of devices operating at device setting L4, configuration 2.

Table 11. Selected devices and configurations.

Air

cleaning

device

Technology

Airflow

pattern

Number of units

& configuration

Device

setting

CADR

[m3/h]

(PM2.5 and

PM10)

Sound

level

[dB(A)]

accepta

ble

AP2

HEPA + ES

+ AC

938
933

43.3

AP4

ES + AC

993
979

45.0

AP6

HEPA

1006
1068

44.6

AP7a

HEPA

778
733

32.1

a: Fulfils nearly the Dutch fresh school guidelines.

References

1. Ding E, Zhang D, & Bluyssen, PM. (2022). Ventilation regimes of school classrooms against

airborne transmission of infectious respiratory droplets: A review. Building and Environment 207:
108484. https://doi.org/10.1016/j.buildenv.2021.108484.

2. Ding E, Zhang D, Hamida A, García-Sánchez C, Jonker L, de Boer AR, Bruijning PCJL, Linde KJ,

Wouters IM, & Bluyssen PM. (2023). Ventilation and thermal conditions in secondary schools in
the Netherlands: Effects of COVID-19 pandemic control and prevention measures. Building and
Environment: 109922. https://doi.org/10.1016/j.buildenv.2022.109922.

3. Bluyssen PM, Ortiz M & Zhang D. (2021). The effect of a mobile HEPA filter system on ‘infectious’

aerosols, sound and air velocity in the SenseLab. Building and environment 188: 107475.
https://doi.org/10.1016/j.buildenv.2020.107475.

4. ECN (2019) High efficiency air filters (EPA, HEPA and ULPA) - Part 1: Classification, performance

testing, marking, EN 1822-1:2019, Brussels: European Committee for Standardization.

5. Netherlands Enterprise Agency (2021) Program of Requirements – Fresh Schools, May 2021.

Accessed:

Jul.

01,

2021.

(in

Dutch)

[Online].

Available:

https://www.rvo.nl/sites/default/files/2021/06/PvE-Frisse-Scholen-2021.pdf.

6. Qian H, Miao T, Liu L, Zheng X, Luo D, Li Y (2021) Indoor transmission of SARS- CoV-2, Indoor

Air 31(3):639-645, https://doi.org/10.1111/ina.12766.

7. Bluyssen PM, van Zeist F, Kurvers S, Tenpierik M, Pont S, Wolters B, van Hulst L, Meertins D.

(2018) The creation of Senselab: A laboratory for testing and experiencing single and combinations
of indoor environmental conditions, Intelligent Buildings International 10(1):5-18.
https://doi.org/10.1080/17508975.2017.1330187

8. Gaillard A, Lohse D, Bonn D, Yigit F (2023) Reconciling airborne disease transmission concerns

with energy saving requirements: the potential of UV-C pathogen deactivation and air distribution
optimization, Indoor Air 2023: 3927171, https://doi.org/10.1155/2023/3927172.

Questionnaire of Sound and Air Movement Perception

Part 1. Assessment of sound

1.1 Can you hear any sound at the location where you are sitting?

Yes

(If the answer is Yes, please continue with question 1.2 and 1.3; If the answer is No, you can
skip question 1.2 and 1.3)

1.2 How loud is the sound that you hear?

Quiet

Loud

1.3 What is your assessment of the sound that your hear?

Part 2. Assessment of air movement

2.1 Can you feel any air movement at the location where you are sitting?

Yes

(If the answer is Yes, please continue with question 2.2-2.4; If the answer is No, you can skip
question 2.2-2.4)

2.2 At which part(s) of your body do you feel the air movement? Please mark the body part(s)

with “x”.

2.3 How strong is the air movement that you feel?

Mild

Strong

2.4 What is your assessment of the air movement that you feel?