Binh Nguyen, Independent Researcher, Hanoi, Vietnam
Residents in modernizing cities such as Hanoi, Vietnam have been moving to newer and well-constructed high-rise buildings. Each appartment often equiped with sliding glass windows, which are better to isolate the environment inside and outside. In the meantime, air pollution imposes long-term health risks and short-term devaluation as a desirable residence and a place to work.
New buildings seem to adequtely isoluate for running air conditionner and is acceptable for sound. A new question comes with arising air pollution is if such new buildings provide a safer place against pollutants such as PM2.5 and PM10.
In this study, the author conducted a 5-day continuous monitoring during a long holiday between the April and May 2019 to evaluate the efficiency of those new buildings against PM(s) from the outside environment.
The apartment on the 10th floor of a over-20-floor building was constructed by Vinaconex, a national brand name for constructing important projects. The building is under 3 years old and located in the South of Hanoi city. The apartment faces a quiet cornor with relatively lower traffic on the ground floor.
One PMS7003 sensor was placed inside plastic box and particulate matters were pulled in for sampling by a constant rate fan. One box was placed inside the room about 70cm above the floor level and the other placed on the balcony with a similar height to the floor. The sampling interval is one minute. The PM2.5 and PM10 outputs are extracted directly from the sensor reading without any calibration.
Data analysis was carried out using Python language with an Jupyter Notebook editor. Graphs were produced using a popular library called Matplotlib. For a cleaner display, data is aggregated into a 10-minutes interval or one-hour interval. Effectiveness of the isoluation was evaluated using the ratios of PM2.5 and PM10 of the inside to those of the outside. The ratios to PM2.5 to PM10 inside and outside (I/O) were used to evaluate the particle profiles and the distribution of PM2.5 and PM10.
Fig.1 presents the concentration of PM2.5 of the inside and the outside of the apartment during 170-hours evaluation. The concentration is averaged into a 10-minute interval. The vertial lines in red marks the time when all windows and the main door are shut and the time they are opened after the residents returned. The number surrounded by the circle marks some episodes for later analysis.
Fig.2 presents the same data in the Fig. 2. The data is aggregated into 1-hour interval for a clear look.
Fig. 3 presents the concentration of PM10 similars to Fig. 1.
Similar to Fig. 2, Fig 4 presented aggregated PM10 centration over 1-hour period.
Basic ratios of PM2.5 and PM10 are presented in Table 2. The numbers are averaged by the entire period which is slightly different than if only between the period marked by two vertical lines in red. The authors decided to do so because of the convenience
|Std (in %)||5||4||14||15|
|Std (in %)||5||4||14||16|
The averaged ratios of PM2.5/PM10 in Table 1 and Table 2 were presented in Figs. 5 and 6 in a time series graph.
The apartment area is about 70m2 with clear space around the building. The entrance door and the main windows are in the North-South direction. When both the door and windows are open, strong wind imposes a heavy advective flow through the apartment. However, when either the door or the windows are closed, there is no distinct flow in the apartment. The author hypothesized, based on the observation of apparent advective flow, with the door and the windows are shut, the internal enviroment is relatively isolated from the outside environment.
The data presented in Fig. 1 supported the author's hypothesis partly. The internal environment contains consistently lower PM2.5 and PM10 concentration compared to the outside environment or the I/O < 1. However, the overall reduction is only marginal, about 13-17%. In the total PM10, PM2.5 is the dominant species inside and outside environments. Dominant fine particles (PM2.5) in the suspended particles supports the data that the sliding windows provide a little reduction of PM2.5 and PM10.
Examining closer the episodes (E) marked by the numbers in Fig. 1, E1 is the start-up period which contains an abnormal lower PM(s) concentration. The author did not recall where the fan was turned off or the intake was placed against the wall. E1 is discarded for future discussion.
E2 shows a typical pattern during the experiment period. The interval PM2.5 concentration responded to PM2.5 concentration with a delay. The magnitude of change was smaller for the interval concentration. When the outside concentration increased, the interval one also increased with delay and a smaller value. When the outside concentration decreased, the inside one reduced but the final value is slighly higher than the concentration in the external environment.
For E3, heavy rains with 30-60mm/24h [1, 2] during the night of April 29 and the morning of April 30 was marked by a sharp drop of PM2.5 in the outside environment. The internal value responses to the drop but only about one-third of the reduction and followed by the increment during the day of April 30. Two peaks during April 30 showed the correlation between two values. A slower drop in the early of May 1st followed by the drop of PM2.5concentration. In this instance, the PM2.5 inside remained lower than the outside one to E4.
E4 was when the residents returned to the apartment after the holiday. With the windows opened, the response of the concentration is almost instantly and the magnitude is close, in which the internal one is slightly smaller than the PM2.5 concentration outside.
Through these episodes, the data indicated with the door closed and with small spaces between the sliding windows and under the entrance door resulted in marginal isolation of PM2.5. The delay in the pattern of PM2.5 concentration is the evidence of active transport with the speed between advection and diffusion. The minimal isolation of the apartment suggested a more stringent construction if the residents prefer a better solution such as using sealants in fixed windows and positive pressure by filtered air flow to overcome advective flows from the outside environment.
The second significant finding is a clear diurnal cycle of ratios of PM(s) from the inside to the outside as presented on Fig. 6. The I/O of PM2.5 ratios are the lowest during early mornings and the highest in late afteroons. This pattern indicated a strong correlation of the temperature of the outside to the transport of PM(s) to and from the inside environment. Comparing Figs. 6 & 7 presents the correlation for most of the day, except the April 30 (E4). From this initial findings, the author recognized the correlation but not conclusively sugguests a direct effect of the temperature to the I/O ratios. The temperature also has a compound effect to the volume of air and making the same number of particles becomes more dense with a lower temperature and thus, presenting as a higher concentration.
The effectiveness of building isolation to PM(s) found in this study is less expected to the author but in aggrement with other studies. Challoner & Gill (2000) evaluated 10 buildings in Ireland and found the ratios of I/O of PM2.5 and NO2 close or above 1. Massey at el. (2008) evaluated I/O ratios of residental homes in central India concluded that I/O ratios closed to from 0.76 to 1.46. The I/O average of residental home by the roadside is close to 1, in the rural area is smaller to 1, and in the urbance is larger than 1 suggesting a different mix of emission sources. A study found a different outcome  in which the to I/O ratios are lower in the range 0.5-0.8 and the peak is during the daytime.
Using two Plantower PMS7003 sensors, one inside and the other outside a closed apartment, the author conducted a 5-day experiment to evaluate the isolation of PM2.5 and PM10. The preliminary findings indicated a minimal reduction of a closed environment of 13-17% of PM(s) with a distinct diurnal pattern of I/O ratios. The I/O ratios are consistently higher during late afternoons and lower during the early mornings. The PM(s) transport across the internal and external environments are not conclusive. The temperature shows a strong correlation with I/O ratios, in which a higher temperature in the inside environment coincides with a higher I/O ratio. The delay between PM(s) of the inside to the outside suggested the transport mechanism is slower than advection. The author concluded that with 2-3 years old buildings and considerd to be well-built only provides marginal insolation to PM(s).
The author would like to thank Dr. Han Huy-Dung from SPARC Lab (HUST) for lending PMS7003 sensors. The PMS7003 is the sensor used in AirSENSE kit for STEM education. The sensor was lent to the author for calibration and comparison of readings of other low-cost sensors and of reference air quality monitoring stations nearby.