The Impact of Indoor Living Wall System on Air Quality: A Comparative Monitoring Test in Building Corridors

: Living wall systems have been widely recognized as one of the promising approaches for building applications due to their aesthetic value and ecological beneﬁts. Compared with outdoor living wall systems, indoor living wall systems (ILWS) play a more vital role in indoor air quality. The aim of this study is to investigate the effects of ILWS on indoor air quality. In an ofﬁce building, two parallel corridors were selected as comparative groups. A 10.6 m 2 ILWS was installed on the sidewall of the west corridor while the east corridor was empty. Some important parameters, including indoor air temperature, relative humidity, concentrations of carbon dioxide (CO 2 ), and particulate matter (PM) were obtained based on the actual environment monitoring. According to the statistical analysis of the data, there were signiﬁcant differences in the concentrations of CO 2 and PMs in the corridors with and without ILWS, which indicated that CO 2 and PM 2.5 removal rate ranged from 12% to 17% and 8% to 14%, respectively. The temperature difference is quite small (0.13 ◦ C on average), while relative humidity slightly increased by 3.1–6.4% with the presence of the ILWS.


Introduction
Currently, city development trends in the vertical direction owing to the rapid increase in population, especially in developing countries [1,2]. The growth of density and reduction of green spaces has caused a series of urban environmental health problems including the heat island effect [3] and air particle pollution [4]. In recent years, the purification and thermal effects of green plants on the urban environment have received increasing attention in the academic community [5][6][7]. Their ecological effects have also been verified, such as pollution reduction [8,9], noise reduction [10][11][12], increase in biodiversity [13][14][15], and energy consumption reduction [16][17][18][19][20][21][22]. Relevant research studies are mostly focused on outdoor greening, such as three-dimensional greening in urban centers [23], greening of roofs [24], and building skins [25]. Hence, the vertical greenery system (VGS) becomes an ideal typology of greening in a crowded city, since it effectively uses the vertical space to accommodate the green plants. VGS can be categorized into green façade and the living wall system (LWS) based on structure configurations. A green façade refers to the climbing plants rooting at the ground level. The vegetation can either climb on the wall directly or indirectly with trellises. By contrast, an LWS refers to vegetation grown in felt or modular systems attached to the walls. The felt or modular system holds the substrate media which can be watered mechanically [26]. With a number of planting units, LWS can be flexibly

Location of the Study
The site is located in Nanjing city (118°38′24″ E, 32°04′48″ N), which is situated in the hot summer and cold winter climatic zone. Based on the China Meteorological Data Network [43], as shown in Figure 1, the maximum ambient air temperature is about 37.2 °C in July, while the minimum is −5.6 °C in January. In terms of relative humidity, despite the large fluctuations from 15% to 100%, most of the values are concentrated in the 60-90% range, with an average of over 70%.

Monitoring Conditions
The monitoring test was carried out on the first floor of a six-story office building (built in 2010) at the university park as shown in Figure 2. To be more specific, the ILWS is arranged on the wall of the stairwell at the western end of the corridor, which has an area of approximately 10.6 m 2 as shown in Figure 3. The ILWS has been constructed and run for more than 2 years. The ILWS consists of a modular planting pot array, a drip irrigation network, and a light-emitting diode (LED) lighting system. Specifically, the water at the bottom of the tank is pumped to each level of the planting pot array, which is then distributed by the drip irrigation to each pot and finally circulated back to the water tank by gravity. The vegetation of ILWS consists of three shade-requiring landscaping plants (Table 1). By contrast, the wall without ILWS is at the eastern end of the corridor. The corridors are enclosed by glazing that separates the indoor corridor space from the courtyard. The building information can be found in Table 2.

Monitored Parameters and Sensors
A series of parameters, including CO2 and PMs concentration, temperature, and relative humidity, of the two corridors, were compared using recorded monitoring data. In doing this, several devices were adopted, including three air quality monitors, a photosynthetically active radiation (PAR) meter, two anemometers, and two pairs of passenger flow counters, as presented in Figure 4. Table 3 summarized the measurement methods of the concentration of CO2, PMs (average PMs concentration was recorded for three mutually exclusive PMs fractions: PM0.3-1, PM1-2.5, PM2.5-10), temperature, and relative humidity. The three air quality monitors were located in two corridors and one outside the building. Monitors in each corridor were positioned at a 2 m height level where the air movement was relatively stable with minimum human interference. The third monitor was set outside the exterior window of the first floor (ground floor height: 4 m), at 2 m level to the floor, i.e., 6 m above the ground. The PAR meter measures the photosynthetic energy from natural and artificial light. It was placed at six measuring points on the surface of the ILWS

Monitoring Conditions
The monitoring test was carried out on the first floor of a six-story office building (built in 2010) at the university park as shown in Figure 2. To be more specific, the ILWS is arranged on the wall of the stairwell at the western end of the corridor, which has an area of approximately 10.6 m 2 as shown in Figure 3. The ILWS has been constructed and run for more than 2 years. The ILWS consists of a modular planting pot array, a drip irrigation network, and a light-emitting diode (LED) lighting system. Specifically, the water at the bottom of the tank is pumped to each level of the planting pot array, which is then distributed by the drip irrigation to each pot and finally circulated back to the water tank by gravity. The vegetation of ILWS consists of three shade-requiring landscaping plants (Table 1). By contrast, the wall without ILWS is at the eastern end of the corridor. The corridors are enclosed by glazing that separates the indoor corridor space from the courtyard. The building information can be found in Table 2. average PPFD on each point. The four anemometers, setting at the middle of each corridor, record the indoor wind speed. The number of people passing by the corridors was recorded by the passenger flow counters that were positioned at the midline of each corridor. All devices were checked and calibrated before use. The monitoring period was recorded from November 2018 to August 2019. Three independent measurements were conducted in both heating and cooling seasons, in which each measurement contains 15 days of records.

Monitored Parameters and Sensors
A series of parameters, including CO 2 and PMs concentration, temperature, and relative humidity, of the two corridors, were compared using recorded monitoring data. In doing this, several devices were adopted, including three air quality monitors, a photosynthetically active radiation (PAR) meter, two anemometers, and two pairs of passenger flow counters, as presented in Figure 4. Table 3 summarized the measurement methods of the concentration of CO 2 , PMs (average PMs concentration was recorded for three mutually exclusive PMs fractions: PM 0.3-1 , PM 1-2.5 , PM 2.5-10 ), temperature, and relative humidity. The three air quality monitors were located in two corridors and one outside the building. Monitors in each corridor were positioned at a 2 m height level where the air movement was relatively stable with minimum human interference. The third monitor was set outside the exterior window of the first floor (ground floor height: 4 m), at 2 m level to the floor, i.e., 6 m above the ground. The PAR meter measures the photosynthetic energy from natural and artificial light. It was placed at six measuring points on the surface of the ILWS ( Figure 3C) in order to obtain the photosynthetic photon flux density (PPFD) values at each point on different sunny and cloudy days. These data were then used to calculate the average PPFD on each point. The four anemometers, setting at the middle of each corridor, Sustainability 2021, 13, 7884 5 of 21 record the indoor wind speed. The number of people passing by the corridors was recorded by the passenger flow counters that were positioned at the midline of each corridor. All devices were checked and calibrated before use. The monitoring period was recorded from November 2018 to August 2019. Three independent measurements were conducted in both heating and cooling seasons, in which each measurement contains 15 days of records.

Data Analysis
As the forms of the two corridors are the same, the air change rate of the corridors depends only on the air movement speed, considering no other mechanical ventilating device is used. The anemometers were set at the middle of each corridor at the height of 1.5 m. The statistical data of 15 days' records (15 min interval), with and without ILWS, were compared in heating and cooling seasons. Through independent sample t-test analysis (Table 4), there was no statistically significant difference in air movement speed between the compared corridors. Table 4. Comparisons on indoor air movement speed between the corridors.

Season
Average Air Movement Speed (m/s)

95% Confidence Interval of the Difference
With ILWS The passenger flow counters were applied to check the equality of the intensity of use between the corridors. The counters recorded the daily and total numbers during June, July, and August in the cooling season and November, December, and January in the heating season. Then the average daily number was calculated and compared. Again, the independent sample t-test (Table 5) shows that there was no statistically significant difference in daily users between the compared corridors. Table 5. Comparisons on the number of people passing through the corridors.

Season
Average

95% Confidence Interval of the Difference
With ILWS (C W ) The light intensity of natural sunlight varied from time to time, but the light intensity of artificial LED light was stable. In order to figure out which one dominated energy source in plants' photosynthesis, and quantitatively evaluate daily PAR from that source, PAR meters were used. They were positioned at six measuring points on the surface of ILWS to record the photosynthetic photon flux density (PPFD) value at daytime (sunlight only) and night (LED light only). Daily PAR can be calculated by multiply PPFD and 10 hours' photoperiod (8:00 AM to 18:00 PM).
The daily photosynthetically active radiation (PAR) values at each measurement point ( Figure 3C) on the surface of the ILWS were measured, shown in Figure 5. Since the PAR received from the natural light was relatively small, whether on sunny or cloudy days, the ILWS mainly relied on the LED artificial light sources to provide energy for photosynthesis. The average photosynthetic photon flux density (PPFD) was 35.3 µmol/s and the daily PAR was calculated to be 1.3 mol/day at 10 hours' photoperiod from 8:00 AM to 18:00 PM. In this study, parameters of the indoor air quality on both sides of the corridors with and without ILWS were measured, including CO2 concentration, PM concentration, temperature, and relative humidity. The obtained data of each parameter were then analyzed, using Mean, Mean Relative Error, and Independent Sample t-test, to examine whether there was a statistically significant difference. It can be inferred that the differences in parameters measured in the two corridors were caused by the ILWS, for the following reasons: (1) The form of two corridors are completely the same and they are symmetrical in the floor plan; (2) There was no statistically significant difference in the air change rate; (3) There was no statistically significant difference in the intensity of use of people.

Comparison of CO2 Concentration with and without ILWS
The daily average value of CO2 concentration is shown in Figure 6. The outdoor CO2 concentration fluctuates from 380 ppm to 450 ppm most of the time. However, in some scenarios, for example in measurement 1 in the heating season, it can even reach 550 ppm. It can be found that the CO2 concentration at the west corridor (with ILWS) is lower than that of the east corridor (without ILWS), and sometimes it is even lower than the outdoor value. Specifically, as shown in Table 6, in the corridor using ILWS during the heating season, the CO2 concentrations of measurement 1, measurement 2, and measurement 3 are 76.5 ppm, 63.6 ppm, and 66.2 ppm lower than those without ILWS. In the cooling season, the remaining three measured concentrations are 53.1 ppm, 49.3 ppm, and 47.3 ppm lower, respectively. Moreover, according to an independent sample t-test on ∁ and (with ILWS and without ILWS), the difference is statistically significant in most of the cases. Since the differences in air change rate and the number of people passing through the two corridors are not statistically significant (considered as equal), and there are no other objects that generate or absorb CO2, the difference in CO2 concentration in the two corridors were considered to be caused by the presence of ILWS. On average, the CO2 concentration decreased by 16.7% in the heating season and 11.7% in the cooling season, respectively. In this study, parameters of the indoor air quality on both sides of the corridors with and without ILWS were measured, including CO 2 concentration, PM concentration, temperature, and relative humidity. The obtained data of each parameter were then analyzed, using Mean, Mean Relative Error, and Independent Sample t-test, to examine whether there was a statistically significant difference. It can be inferred that the differences in parameters measured in the two corridors were caused by the ILWS, for the following reasons: (1) The form of two corridors are completely the same and they are symmetrical in the floor plan; (2) There was no statistically significant difference in the air change rate; (3) There was no statistically significant difference in the intensity of use of people.

Comparison of CO 2 Concentration with and without ILWS
The daily average value of CO 2 concentration is shown in Figure 6. The outdoor CO 2 concentration fluctuates from 380 ppm to 450 ppm most of the time. However, in some scenarios, for example in measurement 1 in the heating season, it can even reach 550 ppm. It can be found that the CO 2 concentration at the west corridor (with ILWS) is lower than that of the east corridor (without ILWS), and sometimes it is even lower than the outdoor value. Specifically, as shown in Table 6, in the corridor using ILWS during the heating season, the CO 2 concentrations of measurement 1, measurement 2, and measurement 3 are 76.5 ppm, 63.6 ppm, and 66.2 ppm lower than those without ILWS. In the cooling season, the remaining three measured concentrations are 53.1 ppm, 49.3 ppm, and 47.3 ppm lower, respectively. Moreover, according to an independent sample t-test on C W and C E (with ILWS and without ILWS), the difference is statistically significant in most of the cases. Since the differences in air change rate and the number of people passing through the two corridors are not statistically significant (considered as equal), and there are no other objects that generate or absorb CO 2 , the difference in CO 2 concentration in the two corridors were considered to be caused by the presence of ILWS. On average, the CO 2 concentration decreased by 16.7% in the heating season and 11.7% in the cooling season, respectively.

Comparison of Particulate Matter Concentration with and without ILWS
The daily average value in PM 0.3-1 , PM 1-2.5 , and PM 2.5-10 concentration is shown in Figures 7-9. In all cases, the PMs concentrations are much higher in the heating season than in the cooling season. This trend is consistent with the national historical data [44,45]. The reasons for this trend are complicated. The main attributes that influence the concentration of PM are pollution source emissions and meteorological conditions. Coal-fired heating and incomplete combustion in vehicle engines at low temperatures may increase emissions in the heating season. The temperature inversion effect and less precipitation in the heating season may also become adverse meteorological conditions for pollutants' diffusion and washout [46][47][48][49]. Table 6. Comparisons on indoor CO 2 concentration between the corridors.

Measurements
Average CO 2 (ppm) Difference MRE t-test Sig.

Comparison of Indoor Air Temperature with and without ILWS
During the experimental period, the average indoor air temperature is approximately 12.3 • C with the relative humidity of 54% in the heating season whereas the mean indoor temperature reaches about 30.0 • C with the relative humidity of 65% in the cooling season.
As shown in Figure 10, the outdoor temperature fluctuates sharply, while the indoor temperature is relatively stable. The temperature difference between indoor and outdoor is 3-6 • C, but the temperature difference between corridors with or without ILWS is much smaller. Data analysis (in Table 8) shows that in most of the cases, the difference in air temperature between the corridors with and without ILWS is statistically significant. The average air temperature in the corridor with ILWS is 0.13 • C higher than that without ILWS in the heating season, and 0.33 • C higher in the cooling season. Given the accuracy of the experimental instrument (±0.2 • C), the ILWS just has a quite limited effect on indoor air temperature. However, further research is needed to rectify its effect on indoor air temperature.

Comparison of Indoor Air Temperature with and without ILWS
During the experimental period, the average indoor air temperature is approximately 12.3 °C with the relative humidity of 54% in the heating season whereas the mean indoor temperature reaches about 30.0 °C with the relative humidity of 65% in the cooling season.
As shown in Figure 10, the outdoor temperature fluctuates sharply, while the indoor temperature is relatively stable. The temperature difference between indoor and outdoor is 3-6 °C, but the temperature difference between corridors with or without ILWS is much smaller. Data analysis (in Table 8) shows that in most of the cases, the difference in air temperature between the corridors with and without ILWS is statistically significant. The average air temperature in the corridor with ILWS is 0.13 °C higher than that without ILWS in the heating season, and 0.33 °C higher in the cooling season. Given the accuracy

95% Confidence Interval of the Difference
With

Comparison of Indoor Relative Humidity with and without ILWS
The daily average relative humidity is plotted in Figure 11. It can be seen that in most of the days, whether in the heating season or in the cooling season, outdoor relative humidity is 10-30% higher than indoors. Relative humidity in the cases with the ILWS is slightly higher than that without ILWS. The data in Table 9 show that in the three monitored periods in the heating season, the average relative humidity in the corridor with ILWS is 3.3%, 7.5%, and 8.3% higher than that without ILWS (6.4% higher on average). In the cooling season, the values are 3.1%, 2.6%, and 3.6% higher (3.1% higher on average). The increase in relative humidity should be related to the transpiration of plants. Table 9. Comparisons on relative humidity between the corridors.

95% Confidence Interval of the Difference
With

Comparison of Indoor Relative Humidity with and without ILWS
The daily average relative humidity is plotted in Figure 11. It can be seen that in most of the days, whether in the heating season or in the cooling season, outdoor relative humidity is 10-30% higher than indoors. Relative humidity in the cases with the ILWS is slightly higher than that without ILWS. The data in Table 9 show that in the three monitored periods in the heating season, the average relative humidity in the corridor with ILWS is 3.3%, 7.5%, and 8.3% higher than that without ILWS (6.4% higher on average). In the cooling season, the values are 3.1%, 2.6%, and 3.6% higher (3.1% higher on average). The increase in relative humidity should be related to the transpiration of plants.

Discussions
In this study, the method of field monitoring was adopted. Since the confounding variables were not perfectly controlled, the causal relationship between the living wall system and indoor air quality cannot be directly revealed. However, through the monitoring of the actual environment and statistical analysis of the data, it has been found that there are statistically significant differences in the CO 2 concentration, PM concentration, and relative humidity between the corridors with and without ILWS. However, since the form of the two corridors are the same, and the difference in air change rate, as well as intensity of use, between the two corridors, were not statistically significant, it can be inferred that the differences were caused by the ILWS. Indoor CO 2 and PMs concentration reduced, while temperature and relative humidity slightly increased. Results from this study were compared with other studies (Table 10). Currently, the number of studies on ILWS is quite small, and the most studied climatic zone is Cfa in the Köppen classification. The impact of ILWS on CO 2 concentration reduction has been confirmed in studies, although the performance varies from case to case. Only two studies (this study and [39]), within the same climatic zone and similar space volume, investigated the impact of ILWS on PMs. Qualitatively, they draw the same conclusion that ILWS could reduce PMs concentration. However, quantitatively, the discrepancies in effectiveness are quite large. The influence of ILWS on temperature shows no obvious trend. In all the studies, relative humidity increases with the presence of ILWS. The increase in dry climatic zones is greater than that in humid regions. The results from this study could be used as a reference for other research and development in a similar climatic zone. In this study, the ILWS was considered as a whole system to influence the monitored environmental parameters, which provided an understanding of the overall effect of a typical ILWS in the particular climatic zone, although the weight of each attribute within the system was still unknown. For example, as the results indicated that there was a statistically significant reduction in PMs caused by the presence of the ILWS, the main attributes could not be quantitatively or even qualitatively identified. Whether the pollutants were adsorbed by the plants' leaves or by the water in the growing pods or by the water vapor produced by plants' transpiration could not be separated out and given weight. In order to figure out the causal relationship, a number of strictly controlled experiments in sealed chambers are needed. Of course, this is very time-consuming. However, from the perspective of the built environment, more attention should be paid to the overall performance of a typical system to evaluate its effectiveness and feasibility. In addition, studying the overall impact could achieve even more accurate prediction results than studying the effects of individual attributes and then adding them up. Since, under real-world conditions, it is almost impossible to control all influencing factors, extract the weight of each one from a whole system, and eliminate the possibility of mutual influence between attributes. Real-world tests can be used as references for other projects or studies, where an analogy method could be used with a large database to predict the potential impact of a new ILWS. A number of tests of different ILWS projects from different climatic zones with different combinations of plants are needed to build such a database. When an analogy method is used, the higher the matching degree of parameters, the higher the accuracy of prediction results will be. Area of ILWS to space volume ratio could be used as a key parameter for such an analogy evaluation, along with other parameters, such as plant type, the proportion of plants' combination, daily PAR on the ILWS, and so forth. In order to match the parameters from a new ILWS to those from a database, a training sample dataset is needed for machine learning, using a computer algorithm to determine the weight of each parameter. This study is just the first step, as one of the important case studies in many, towards this target.
The idea of introducing ILWS to the indoor environment starts from improving indoor air quality and environmental aesthetic value for human well-being [33]. However, the benefits do not stop here. The potential of using ILWS to reduce building energy consumption in heating and cooling seasons cannot be ignored [28]. The direct impact of ILWS on indoor air temperature is still inconclusive. It may be also influenced by some exterior factors, such as the orientation of the building, the location of openings on the building, or even the function of space. Further studies on these factors are needed. The indirect effect of ILWS on building energy saving could be mainly reflected in ventilation energy consumption. Currently, the main method of removing indoor air pollutants still relies on diluting the indoor pollutants by ventilating the fresh air from the outside [55]. During the spring or autumn seasons, when the outside temperature fluctuates within the temperature comfort zone, usually 18-26 • C, natural ventilation can be used directly [56]. However, in cooling or heating seasons when the temperature is beyond the comfort zone, the fresh air needs to be pre-heated or pre-cooled before it is ventilated mechanically to the indoor space in order to reach the indoor designed temperature. Therefore, in such scenarios, the ventilation rate, determined by the fresh air requirement, has a proportional relationship with the building energy consumption in ventilating sector. Introducing ILWS into buildings may improve the air quality, thus equivalently reduces the fresh air requirement and the ventilation rate when the outside temperature is beyond the comfort zone for natural ventilation. Apart from indoor and outdoor temperature and humidity differences, the performance of energy consumption reduction in ventilation is mainly determined by the CO 2 absorption rate of ILWS instead of the adsorption rate of PMs or VOC. It is because PMs and VOC can be adsorbed by various artificial porous materials, but there is currently no low-cost practical method for massive CO 2 conversion [57]. Plants' photosynthesis is still the most extensive process of carbon dioxide to oxygen conversion on earth. Consequently, if working with other artificial filters, ILWS may reduce the building energy consumption through ventilation heat loss/gain in heating or cooling seasons. The effectiveness of this mechanism still needs further study.

Conclusions
In this study, a field monitoring test of the ILWS was implemented in order to evaluate the system performance for the Cfa climate zone (Köppen climate classification) from November 2018 to August 2019. The influences of the ILWS on indoor air quality with and without ILWS between two parallel corridors were studied in an office building based on different parameters such as CO 2 concentration, PM, atmospheric temperature, and relative humidity. Some key findings are concluded as follows: • ILWS could be a promising opportunity in the eco-design of the buildings considering its aesthetic value and purification function of indoor air. • Based on the results from the statistical analysis, the differences in CO 2 and PM concentration in the two corridors are statistically significant-this indicates the positive effect of ILWS on improving indoor air quality.

•
In terms of the CO 2 concentration, the average indoor air temperature of the corridors with ILWS can be reduced to the same level of outdoor condition, or even slightly lower than outdoor levels.

•
The purification performance of CO 2 and PMs in the heating season is better in comparison with that in the cooling season due to the local seasonal climate condition and pollutant emission.

•
The effect of ILWS on indoor air temperature is quite limited. However, the relative humidity in the corridor with ILWS is slightly higher (3.1-6.4%) than that without ILWS.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest. Mean Relative Error PM 0.3-1 The particle size is between 0.3 µm and 1 µm PM 1-2. 5 The particle size is between 1 µm and 2.5 µm PM 2.5-10 The particle size is between 2.5 µm and 10 µm c average value of each measured parameter C W 15 days' mean in the corridor with ILWS C E 15 days' mean in the corridor without ILWS C s 15 days' mean of outdoor environment C W,i concentration with ILWS by minute C E,i concentration without ILWS by minute C S,i outdoor concentration by minute m daily reduction of the corresponding pollutants per square meter R r pollutant reduction ratio