3.2. Questionnaire Results
In general, occupants were slightly satisfied with the overall comfort in their offices (mean: 4.96). The most satisfying rated parameter was Light (mean: 5.57) followed by thermal (mean: 4.78), noise (mean: 4.54) and air quality (mean: 4.43) comfort. More specifically, occupants reported general moderate overall comfort (~44%) while only 15% were fully satisfied. Privacy (60%), decoration (42%) and glare (42%) were revealed to be the most dissatisfying parameters (values 1–3 from the 7-point scale). Furthermore, occupants reported complaints about very high temperatures (36%), dry and smelly air (36% and 31%) and noise inside the building (
Figure 3).
In order to identify the association between IEQ parameters and overall comfort, proportional odds ordinal logistic regression analysis was applied. Results showed that the greatest OR value (OR: 2.10,
p < 0.05) was found for overall temperature comfort. Similar studies also reveal that thermal environment that the most influential environmental aspect [
50,
51]. The second main parameter associated with overall comfort was overall noise (OR: 1.39,
p < 0.05). The importance of noise satisfaction was also raised in a study conducted in public and private buildings focusing on open-plan offices [
52]. Then follow the satisfaction with air quality (OR: 1.17,
p < 0.1) and light without statistical significance (OR: 1.03,
p > 0.1).
Figure 4 shows the prevalence of the SBS symptoms. In general, 63% reported at least one health symptom, while the most reported heath symptoms were headache (41%), dry eyes (30%) and sneezing, lethargy (16%, 15%). The least frequent were related to heart symptoms such as bradycardia (1%) and irregular heartbeats (2%); respiratory phlegm/mucus (2%); wheezing (2%); and rash or irritated skin (2%). Furthermore, if any participant reported at least one symptom that was classified in SBS group, the IAQ perception was examined for these two groups by the Mann–Whitney test. The IAQ perception of the SBS group (mean: 4.24) was significantly lower at
p-value level <0.1 than in the non-SBS group (mean: 4.76).
Occupants were asked to evaluate the satisfaction level of the perceived control of the IEQ parameters in their offices and then how many times per day they took an action to adjust these parameters. Perceived control of noise (mean = 3) exhibited the lowest score. On the other hand, the perceived control of light (mean = 5.3) was the highest, followed by the perceived control of shade (mean = 4.52), ventilation (mean = 4.51), and temperature (mean = 4.16). Regarding the exercise of control, there was an identifiable number of occupants who reported that they did not control the IEQ parameters. More specifically, during a working day, a percentage of up to 60% and 43% of occupants were not able to adjust the noise and shade, respectively, in their offices. In general, once per day they controlled the light levels (48%) and in the case of temperature and ventilation parameters, they preferred to adjust them several times during the day (
Figure 5).
In
Table 2, the correlations between the exercise control and occupants’ perception of control over the IEQ parameters and with their comfort satisfaction as a result of them are presented. The frequency with which employees used available controlling devices in their offices showed to have a moderate association with their control perception of ventilation and noise (Spearman cor. 0.424 and 0.430, respectively), while a weaker correlation appeared in shading control (Spearman cor. 0.248). Perceived overall comfort correlated significantly and positively with perceived control parameters and temperature control appeared to have the highest correlation (Spearman cor. 0.474). Controlling noise in offices remained the least affecting parameter both in noise comfort (Spearman cor. 0.252) as well as in overall comfort (Spearman cor. 0.242). These findings reinforced the idea that occupants prefer interaction with the building and to have some personal autonomy to reach their desired comfort level [
53]. Improved personal control of the IEQ by applying more available controls to the occupants could further satisfy their thermal comfort and IAQ needs, and also reduce the energy consumption by centralized HVAC systems, especially in open space offices [
54,
55]. These results are also enforce the findings in a recent study in which Hong et al. revealed the link between occupant and their behaviors of sharing energy and environment control systems and interactions with their colleagues [
56].
The questionnaire also gathered information about the occupants’ self-reported productivity ranging from −30% up to +30%.
Figure 6 shows that tself-reported productivity at workplace presents an increasing trend when the perception of comfort and control of the IEQ parameters is higher. More specifically, while occupants reported greater levels of overall comfort satisfaction, from 3.5 to 5, a remarkable increase in their productivity was noticed, i.e., from −30% to +10%. The same pattern is also revealed for the controlling parameters, especially that of temperature, while on the other hand noise control seems to be the least affecting parameter on productivity. The significance of perceived control in productivity was highlighted in another study conducted at 37 European offices, where a small increase in the low control area resulted in a clear increase in productivity [
34]. Also, a UK study discovered that the comfort, health, and productivity of workers were all related to their perceived personal thermal control [
57].
The participants, after completing the evaluation of the current IEQ perception in their offices, were asked to reevaluate their perception during the working hours before the lock-down period. The Wilcoxon statistical test was used to compare the pre- and post-lockdown period satisfaction scores.
Figure 7 presents the IEQ parameters that appeared to be statistically significant. The perception of overall comfort, temperature, light and noise was significantly higher during the time of the study. On the contrary, regarding privacy, the occupants’ perception was evaluated as significantly lower after returning to offices.
In the last part of the questionnaire, participants expressed their degree of acceptance of the lockdown aspects due to the pandemic, and the results are presented in
Table 3. In general, office occupants agreed to the applied polices and they seemed to accept the new conditions in their offices (mean score above the neutral score 3). On the other hand, at the time of the survey, they reported that they do not prefer to work from their homes anymore (score 2.54). The acceptance of new working conditions and the aspect of how the rest of the occupants in their office accept them, showed significant statistical correlation with their overall comfort (Spearman cor. 0.185,
p < 0.01 and 0.370
p < 0.01). The occupants’ acceptance of the COVID-19 countermeasures application, in order to ensure a healthy and safe environment at work, was significantly correlated to all overall IEQ parameters. Similar results were revealed as to whether participants believed that their colleagues considered the COVID-19 situation to be serious. These new findings were also reported in a study conducted in offices in Japan during the pandemic period, where it was highlighted that increasing satisfaction with COVID-19 countermeasures may increase occupants’ performance in the offices [
7]. Higher acceptance levels of the parameters, such as feeling comfortable in returning to workplace and the application of systems for social distancing and COVID-19 protection, resulted in fewer health complaints (Spearman cor. −0.246
p < 0.01 and −0.180
p < 0.01, respectively). As is depicted in
Figure 8, the social distancing protocol, use of masks and sanitizers (personal or stations) and the installation of barriers between the occupants were three of the most commonly applied protocols in the offices.
3.3. Indoor Air Pollutants Concentration in Offices
Descriptive data of indoor VOCs, aldehydes, O
3, NO
2 and SO
2 concentrations are presented in
Figure 9,
Figure 10 and
Figure 11. Among the measured VOCs, the highest median concentrations were found for benzene 11.0 μg m
−3 (max: 32.5 μg m
−3) and toluene 9.9 μg m
−3 (max: 32.9 μg m
−3), while formaldehyde and acetaldehyde showed the highest indoor concentrations of the investigated aldehydes, with median values 21.9 μg m
−3 (max: 34.8 μg m
−3) and 13.0 μg m
−3 (max: 40.3 μg m
−3) respectively. Median concentration for O
3 was 2.3 μg m
−3 (max: 7.6 μg m
−3), for NO
2 was 8.2 μg m
−3 (max: 25.1 μg m
−3) and for SO
2 was 0.1 μg m
−3 (max: 0.9 μg m
−3). Acrolein, crotonaldehyde, methacrolein and m-tolualdehyde were not detected in any building. Finally, in the majority of the buildings, octane, d -limonene and SO
2 concentrations were below the limit of detection values.
As a first evaluation, the computed median and mean values of pollutant concentrations of this study were compared with the relevant seasonal (summer) values reported in the OFFICAIR project [
17], a European project in which a wide scale of indoor air pollutants measurements in 37 office buildings were performed [
58]. Overall, among the VOCs, benzene, toluene, and d-limonene concentrations were higher in the investigated buildings compared to OFFICAIR corresponding values (1.4, 8.1 and 4.7 μg m
−3 respectively) while ethylbenzene (1.8 μg m
−3), a-pinene (4.2 μg m
−3) were of the same level. Aldehydes, formaldehyde, and acetaldehyde concentrations were higher, benzaldehyde was of the same order of magnitude while propionaldehyde, hexaldehyde and acrolein were present in lower levels. In addition, lower concentrations were observed for ozone and nitrogen dioxide. In another large-scale study in European countries, VOC measurements took place in 56 offices buildings [
59]. The reported concentrations of benzene (14.6 μg m
−3), toluene (35.1 μg m
−3), d-limonene (34.6 μg m
−3), Benzaldehyde (4.1 μg m
−3), o-xylene (10.2 μg m
−3) and p,m-xylene (22.2 μg m
−3) were much higher than the present results. Regarding the concentrations levels of ethylbenzene, o-xylene and p,m-xylene were of the same magnitude as a recent study conducted in office rooms in Poland; on the other hand, benzene concentrations were higher [
60].
Benzene, toluene and xylenes (BTEX) are volatile compounds related to outdoor sources, i.e., traffic, industrial emissions, combustions [
61,
62,
63,
64,
65]. Buildings B3, B4 and B6 exhibited elevated levels of these compounds, possibly because of their vicinity to high traffic roads, crossroads, industrial units or areas of tracks maneuvering. This assumption was verified by the lower or close to unity values of indoor to outdoor ratios (I/O), characteristically for benzene: 1.1 (B3), 0.5 (B4), 0.7 (B6). However, VOCs are associated with indoor sources, as well. Specifically, toluene has been characteristically associated with environmental tobacco smoke [
66]. Indeed, B1, B2, B3 present high levels of toluene likely due to smoking activity which was taking place either inside the office room (B1, I/O = 8) or in an adjacent outdoor smoking permitted area (B2, B3: I/0 = 2), where smoke penetrated indoors. Another indoor source of toluene and xylenes (usually combined with the presence of aldehydes) is emissions from building materials, paints, adhesives etc. [
18,
67,
68]. A typical case is the one of the recently renovated building B7 (toluene I/O = 2.1; m,p-xylene I/O = 22) as well as the offices of B1 (toluene I/O = 8) which is adjacent to a tiles and cement-based products unit. Last but not least, the measured concentrations inside a building are significantly influenced by the air renewal pattern of the building. Building 4, which is characterized by inadequate natural or mechanical ventilation conditions (B4 was rated as one of the lowest in perception for fresh air), presents comparatively higher concentrations of terpenes (limonene, a-pinene) and xylenes, probably due to pollutants accumulation inside the building. Measured d-limonene and a-pinene median concentrations were considerably below the short term critical exposure limit (CEL) of 45 mg m
−3 and 90 mg m
−3 respectively, developed within the EPHECT (Emissions, exposure patterns and health effects of consumer products in the EU), project based on sensory irritation as the critical effect [
69].
Carbonyls and aldehydes can originate from a wide variety of sources (building materials, equipment, adhesives, paints, smoking etc.,), the strength of which may significantly depend on the season i.e., differences in temperature, relative humidity and other environmental conditions [
18]. Aldehydes also originate from O
3-initiated reactions with many common and reactive VOCs that contain unsaturated carbon-carbon bonds, which react with O
3 much more rapidly than saturated organic compounds. In the present study, among the aldehydes, formaldehyde, acetaldehyde, and acetone showed a noticeable differentiation in the investigated buildings. For propionaldehyde, 2-butanone, benzaldehyde and hexaldehyde, values were almost similar among the office rooms. Remarkably, the indoor to outdoor ratio was significantly higher than unity in all buildings (I/O range for formaldehyde: 3.3–11.4, acetaldehyde: 2–20 and acetone: 2.5–11.3), indicating the prevalence of indoor sources inside the studied office buildings. Indeed, formaldehyde has been associated with emissions from electronic office equipment such as laser printers and photocopiers [
70]. Formaldehyde and acetaldehyde are associated with emissions from wooden-pressed products and wall paints [
71,
72,
73,
74,
75]. The median formaldehyde concentration (21.93 μgm
−3) was below the WHO Indoor Air Quality Guideline (IAQG) of 100 μgm
−3 derived in 2010 based on sensory irritation as the critical effect [
1] The campaign findings are in agreement with the literature, since similar materials and equipment are largely used in office rooms.
Nitrogen dioxide originates mainly from anthropogenic outdoor sources i.e., combustion of fossil fuels (coal, gas and oil). In the present study, indoor NO
2 is observed to be higher inside buildings attached to high traffic roads (Buildings 3,4 and 6). The same findings were also revealed in a recent study in schools and offices where indoor exposure to NO
2 from the infiltration of ambient air can be significant in urban areas, and in areas with high traffic volumes [
76]. However, in the cases of B2 and B4, the ΝO
2 I/O ratio was significantly higher than unity (3.8), implying the contribution of indoor sources and particularly combustion sources such as smoking, which is an activity that occasionally occurs inside the buildings. Regarding the O
3, I/O ratio, the values ranged from 0.01 to 0.13. Ozone is very reactive with high oxidizing potential and thus there are many possible scenarios for ozone decay indoors. Concerning O
3, there was an indication of strong sinks indoors as was also reported in office and school environments with similar values [
77,
78]. Building materials in offices and various types of design decorations can be considerable sinks for indoor ozone due to the irreversible heterogeneous reactions between ozone and material surfaces [
79,
80].
Concentrations of PM
1, PM
2.5 and PM
10 during the working hours are presented in
Table 4 (available for B3, B4, B5, B6). The highest mean values of fine and coarse fractions (PM
2.5, PM
10) were observed in B4 and B6 (though not important differences for PM
2.5) possibly because of their location (close to traffic roads and industrial units). The mean values of PM
2.5 and PM
10 for all buildings were below the EU annual limits (25 μg/m
3) and (40 μg/m
3) respectively [
81]. These results are similar to those of OFFICAIR project for Greece, where mean concentrations of PM
2.5 was 13.3 μg/m
3 and 14.3 μg/m
3 during summer and winter [
82] respectively, while being higher compared to mechanical ventilated offices [
83]. Maximum values of all fractions can be attributed to penetration events caused during building cleaning activities or instant high occupancy. The highest maximum values were noticed for B4 (the building with the poorest ventilation), indirectly implying pollutants accumulation.
3.4. Associations between Indoor Concentrations and SBS Symptoms/IEQ Comfort Perception
The variations in concentrations between the occupants belonging to the SBS group and the non-SBS group are shown in
Figure 12. By applying the Mann–Whitney U test between groups of reporting SBS versus number of occupants who reported SBS symptoms were exposed to higher values with significant (
p < 0.05) difference in NO
2, benzene, toluene, Propionaldehyde, 2-butanone, benzaldehyde, and hexaldehyde.
As a next step and for exploring indoor concentration variations in greater depth, tests were performed for each SBS symptom subgroup (
Table 5). For occupants who reported eye irritation, NO
2 (
p < 0.1), benzene and toluene, benzaldehyde and hexaldehyde (
p < 0.05), concentrations were higher in their offices than in those of occupants without symptoms. In offices where occupants with reported respiratory symptoms, NO
2 and toluene (
p < 0.05), 2-butanone (
p < 0.05) were also measured in higher levels. For the heart and skin symptoms groups, fewer compounds were revealed to be significantly higher (NO
2, Hexaldehyde (
p < 0.05) and toluene (
p < 0.1), respectively). For offices where occupants reported general symptoms, mean concentration values of NO
2, benzene and hexaldehyde (
p < 0.05) were found in higher concentrations.
Further to the SBS subgrouping, the analysis was expanded by separating occupants into two groups of satisfied and unsatisfied occupants, regarding the perception of IAQ.
Figure 13 presents the values for parameters related to IAQ, such as fresh air, odor, and overall perception. In general, the workplace of the IAQ-dissatisfied group appeared to be located in offices with higher pollutants concentrations. NO
2 appeared to be statistically significant (
p <0.05) for all IAQ parameters. In the case of odor or stuffy air, toluene and d-limonene were significantly higher.
3.5. Chronic Health Risk Assessment
Health risks due to inhalation exposures to benzene and formaldehyde for the staff in all investigated buildings were considered by applying both models, i.e., non-carcinogenic health risk and cancer risk assessment, using the methodologies described in
Section 2.5. The computed HQ for the non-carcinogenic health risks and LCR values are depicted in
Table 6 and
Table 7, respectively.
As shown in
Figure 9 and
Figure 10, in all monitored office buildings, the concentrations of formaldehyde were below the limit (CA
HCHO < 100 μg/m
3) set by WHO [
1]. On the other hand, concerning the benzene concentrations are concerned, only buildings 1 and 6 were within the acceptable range, building 5 was marginally above the permissible limits and all other buildings exceeded the recommended value of 5 μg/m
3 [
84] with a maximum value of 32.5 μg/m
3. In particular, WHO considers benzene as a pollutant without a safe level [
1].
The HQ value computed for the study of the non-carcinogenic health risk was higher than unity in all buildings for the pollutants under consideration, suggesting that workers in those particular office buildings might experience health problems. Such problems could include eyes and upper airways sensory irritation, along with lung effects involving asthma and allergies [
1].
The cancer risk assessment revealed that for both pollutants, LCR values were higher than 1.0 × 10
−6; however, values did not exceed the acceptable limits of cancer risk (10
−6–10
−4), for all investigated buildings. In addition, the highest LCR value for formaldehyde was observed in building 4, even though the highest formaldehyde concentration was measured in building 5. The same scenario appeared for the benzene study, where the maximum LCR value was computed for building 4 while building 2 benzene concentrations were the highest. This observation illustrates that pollutant concentrations should not be considered as the only factor affecting the quality of indoor environment. The risk assessment results in the case of formaldehyde showed that health problems could exist even though the pollutant concentration levels in the indoor environment were in low concentrations, as was also revealed in a recent study conducted in offices in China [
85].