Ss 4

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Building and Environment 115 (2017) 281e290

Contents lists available at ScienceDirect

Building and Environment


journal homepage: www.elsevier.com/locate/buildenv

An assessment of the atmospheric particle removal efficiency of an in-


room botanical biofilter system
P.J. Irga a, *, N.J. Paull a, P. Abdo a, b, F.R. Torpy a
a
Plants and Environmental Quality Research Group, Faculty of Science, University of Technology Sydney, Australia
b
School of Electrical, Mechanical and Mechatronic Systems, Faculty of Engineering and Information Technology, University of Technology Sydney, Australia

a r t i c l e i n f o a b s t r a c t

Article history: In addition to the growing requirement to reduce building energy needs, demand has arisen to find
Received 21 December 2016 sustainable methods of improving indoor air quality. Recent advances in green wall technology have led
Received in revised form to the development of activated systems, termed botanical biofilters, that move air through the plant
25 January 2017
growth substrate to increase the rate at which the interior atmospheric environment is exposed to the
Accepted 31 January 2017
Available online 2 February 2017
components of the plant-substrate system that are active in air pollutant removal. Development of this
technology is moving towards green wall integration within building air conditioning and ventilation
systems. The work presented here describes an evaluation of several parameters essential for deter-
Keywords:
Sustainable buildings
mining the functionality of a modular botanical biofilter, as well as experiments to systematically
Active living wall determine the filtration performance of the device, specifically the single-pass particulate rem filtration
Air filtration efficiency was evaluated and defined. The maximum filtration efficiency for total suspended particulate
Clean air delivery rate matter peaked at an air flow rate of 11.25 L s1 through the 0.25 m2 filter, with any increases in air flow
HVAC systems rate met with a reduction in efficiency. The system recorded removal efficiencies were 53.35 ± 9.73% for
Phytoremediation total suspend particles, 53.51 ± 15.99% for PM10, and 48.21 ± 14.71% for PM2.5. Comparisons were made
against the single pass efficiency of the system without the botanical component, as well as a common
in-duct pleated panel air filter, indicating that further development is required to enhance the filtration
capacity of the system if it is match current air filtration standards.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction particulates [6]. Fine particulate matter, however, may infiltrate


into indoor environments as it cannot be effectively filtered out of
Indoor airborne particulate matter (PM) is a growing concern as the incoming air, even in buildings with efficient mechanical air-
populations become more urbanised, with an increasing propor- conditioning systems [7]. Thus, the concentration of outdoor pol-
tion of individuals spending most of their time in indoor environ- lutants still affects the quality of indoor air, and a large proportion
ments [1]. Airborne PM range in size, with most epidemiological of indoor particles are of outdoor origin, despite existing filtration
studies focusing on PM10 which comprises particles less than or procedures [8]. Indoor sourced particles may also pose a health
equal to 10 mm in diameter, as these particles are able to penetrate issue for building occupants, as human activities such as printing
beyond the nasal passages and into the respiratory system, causing and human movement can lead to particle resuspension [9]. Irre-
adverse pulmonary and cardiovascular diseases [2e4]. Additional spective of origin, technology that mitigates and reduces inhalable
evidence suggests that smaller PM fractions such as PM2.5, are more particles and other air pollutants within the indoor airspace is
strongly associated with these cardiovascular health effects [5]. crucial for creating a healthy indoor environment.
The increase in air-tightness of modern buildings to gain greater The use of portable in-room air filters is a popular method for
control over the regulation and conditioning of indoor air, has led to the control of indoor suspended particulate matter [10]. These air
a decrease in the infiltration of large, filterable outdoor sourced filters mainly utilise a porous lattice of fibres, typically comprised of
fibreglass, although other filters that are available rely on electro-
static precipitators, ion generators, or composite filters containing
* Corresponding author. University of Technology Sydney, P.O. Box 123,
activated carbon [11]. The PM filtration efficiency of any indoor air
Broadway, Ultimo, NSW, 2007, Australia. filtration system is dependent on the air flow rate through the filter,
E-mail address: Peter.Irga@uts.edu.au (P.J. Irga). the single-pass filtration efficiency of the medium and the airflow

http://dx.doi.org/10.1016/j.buildenv.2017.01.035
0360-1323/© 2017 Elsevier Ltd. All rights reserved.
282 P.J. Irga et al. / Building and Environment 115 (2017) 281e290

pattern that the system generates in the room [12]. The airflow 2. Materials and methods
pattern in the room is influenced by furnishings and other interior
structures, duct or outlet characteristics, room shape and size and 2.1. Botanical biofilter system design
leak patterns [13]. Furthermore, particulate density, charge and
shape can affect particulate matter penetrability through filter The current study assessed a commercial active living wall
devices [11,14,15]. Regardless of the method of filtration, all filters biofilter (the Breathing Wall system, Junglefy Pty Ltd, Sydney),
are characterised by their ability to remove particulate aerosol which has been designed for use in commercial scale vertical plant-
materials from the air-stream and the rate at which the filter can walls (Fig. 8), previously described in Irga et al. [27]. Briefly, the
supply clean air, with removal efficiency and air flow used to system is comprised of two functional components; a biofilter
describe filter functionality and subsequently derive the ‘Clean Air through which ambient indoor air is drawn with a low pressure
Delivery Rate’ (CADR) of the system. mechanical ventilation system, which also acts as the plant growth
Botanical biofiltration systems are becoming an increasingly substrate, and a region of vertically grown plants which are located
popular method for indoor air pollution mitigation [16], in addition within circular compartments. Each module holds 16 plants which
to providing a positive workplace aesthetic. A component of the grow horizontally from circular compartments in the casing, with
functionality of these systems is the activity via which accumulated 36.3% of the front face allowing air to pass. For these experiments,
airborne pollutants act as energy, carbon or other nutritional Chlorophytum comosum (Spider plant) was utilised as the botanical
sources for the bioagents, which then convert the organic pollut- component in the system, as it has previously been assessed for PM
ants into less harmful compounds such as water, carbon dioxide accumulation [28]. Mechanical components include an air intake
and biomass [17], while inorganic pollutants simply accumulate with electric axial impeller forcing air through a 7.5 cm diameter
within or on the substrate. Passive botanical biofiltration can phy- port in the polyethylene case unit (50  50  13 cm with 0.75 cm
toremediate indoor air contaminants, such as volatile organic wall thickness). Air then enters a plenum (1.5 cm depth), where the
compounds (VOCs), which are primarily removed by the plants' air flow is evenly distributed across the system's packing medium.
rhizospheric bacterial community [18], and CO2 [19] which is The electric axial impeller draws air from the ambient indoor
removed through the action of plant photosynthesis. Compara- environment and provides a uniform flow across the biofilter,
tively, active systems that utilise mechanical or other forms of through the growth medium and past the botanical component
ventilation, force air through the substrate and planting bed root back into the indoor space.
system which may increase the filtration and air purification ability All planted modules had been nursery grown for 8 months
of the system. It has been suggested that the use of active botanical before being used in the trials. Throughout the duration of exper-
biofilters for indoor air pollution control is a more sustainable iments, all vegetated modules were maintained in a glasshouse,
alternative to standard mechanical methods which could poten- with an average temperature of 23.7 ± 3.6  C, relative humidity of
tially lead to a reduced city carbon footprint [20]. The majority of 68.1± 16.0% and a maximum mid-day light level of
research that has been conducted on botanical biofilters has 90 ± 10 mol m2 s1. All modules' moisture content was increased
focused on VOC removal. For example, Darlington et al. [21] to saturation once weekly.
observed a reduction in toluene, o-xylene and ethylbenzene from
indoor air when using their botanical biosystem. They noted that 2.2. Characterisation of the effect of varying air flow on PM
running their system at the highest air flux resulted in the most filtration efficiency within a closed environment (decay test method)
efficient VOC removal rates. Comparatively, Wang and Zhang's [22]
active botanical biofilter consists of an activated carbon hydro- Whilst proof of concept for PM removal by the system was
culture potted plant system, which was incorporated into a build- established by Torpy et al. [16], characterisation for the appropriate
ing heating. ventilation and air conditioning (HVAC) system. Their air flow through the system for maximum PM removal efficiency
system effectively removed formaldehyde and toluene, continuing has not previously been explored. All operations were conducted
to perform well over a 300-day testing period. The development of within a laboratory, with mean light intensity of
green wall technology for the reduction of high indoor CO2 con- 10 ± 2 mol m2 s1; average temperature 23.0 ± 0.1  C, and relative
centrations is an area of burgeoning research focus [19], especially humidity 55± 10% (means ± SD). Characterisation of the effects of
as a function of reduced building energy costs. Industry practi- air flow through the system on PM reduction was conducted in fully
tioners are now taking into account different plant choices with sealed, air-tight Perspex test chambers 0.6  0.6  0.6 m (0.216 m3
consideration into cultivars' ability to remove CO2. Researchers internally). To allow access into the chamber, they were fitted with
from the University of Seville [23] are exploring other applications a removable Perspex side on steel frames, sealed with adhesive
for botanical biofilters, including cooling potential and humidity foam-rubber and adjustable metal clamps. A silicone rubber septa
regulation, which they hypothesise can also reduce building was fitted on the removable side of the chamber to allow for the
ventilation requirements. There appears to be potential for these injection of test pollutants. High precision plunger-in-needle style
systems to be used as active evaporative coolers, and may assist syringes (SGE Australia) were used for all PM injections. Each
future building climate control as demonstrated in other studies chamber was fitted with a 2.4 W, 100 mm diameter fan to ensure
[24e26]. Currently, no botanical biofilters have been comprehen- adequate air mixing within the chamber and create a homogenous
sively examined for the removal of PMs. atmospheric environment. Lighting within the chamber was sup-
The research presented here describes experiments and plied with a bank of 18 W fluorescent tubes positioned directly
modelling conducted to determine the effectiveness of an active above each of the chambers (Wotan L 18/11 Maxilux daylight,
green wall type, botanical biofilter in reducing PM levels. The Ozram, Germany), producing a light intensity of 120 mol m2s1.
effectiveness of an in-room air cleaner under typical operating Single modules were tested at five air flow rates: axial impellers
settings depends on the principal characteristics of single-pass off, 3.75, 7.50, 11.25 and 15.0 L s1. To compare the efficiency of the
filtration efficiency and the airflow rate through the filter. Thus, packing media independently of the action of the botanical
the active botanical biofilter was further evaluated for its single- component of the system, the same experiment was conducted for
pass filtration efficiency for particles ranging in diameter from 2.5 the system without plant material, with the case containing only
to 10 mm along with total suspended particles. the packing material, utilising the same inlet air pressures. The air
flow rates were slightly different for these experiments due to
P.J. Irga et al. / Building and Environment 115 (2017) 281e290 283

altered flow resistance across the filter bed resulting from air flow the chamber (at 20 Pa pressure difference from ambient) and
permeability augmentation by the plant roots [29]; with the prevented unfiltered air from entering the chamber due to infil-
resulting air flow rates being: axial impellers off, 3.95, 7.90, 11.80 tration. Nonetheless, pressure differences were encountered across
and 15.8 L s1. To correct for the deposition rate of particles within various treatments, which were accounted for statistically (see
the chamber and on the exterior surface of the air filter; modules section 2.4).
were placed in the chamber without fan operation, and the PM Particles were generated in the chamber by burning a 1.0 cm2
deposition rate compared to that with no filter present in the ISO 536:2012 80 gsm square of paper [31], made to Standard
chamber. As no significant difference was observed between the AS4707:2014 [32] loaded with 4 mL of the diesel fuel as used in
two treatments (ANOVA: P ¼ 0.247), all further comparisons were section 2.2, utilising a 0.5e10 mL micropipette. This procedure
made against the non-operational module. created a pulse input response curve at the test instrument, with
Hybrid diesel fuel candles were created to facilitate the gener- particulate concentration rapidly generated and entering the
ation of PM. Briefly, the candles were 100 mL in volume with 40 mL feedstream in a short time, with effluent concentration measured
retail grade diesel fuel (Shell, Australia), 60 mL paraffin wax, and as a function of time.
housed within a glass housing (55 mm height x 65 mm diameter). Initially, the empty duct was exposed to the stream of sus-
The waxediesel fuel solution was heated and homogenised thor- pended particles with fixed concentration and flow conditions, and
oughly. A 60 mm  7 mm cotton wick was centred in the middle of sampling was done in triplicate at the sampling point as illustrated
the candle. The diesel fuel contained less than 10 parts per million in Fig. 1. The resulting concentrations for the three particle sizes
of sulphur, and meets the requirements of the Australian National measured (TSP, PM10, and PM2.5) were used as representative of the
Fuels Quality Standards Act 2000 [30]. Airborne PM were produced inlet stream. Calculations of fractional and overall efficiencies were
by burning the hybrid diesel candle and collection of effluent gas thus based on measurements at the same sampling point for the
into a 5 L volumetric flask. The volumetric flask was then rested for system with or without the biofiltration system present. Further
5 min to allow the deposition of larger particles, after which comparisons were made against the single pass efficiency of the
approximately 100 mL of air containing suspended particles was system without the botanical component, as well as a common
injected into the chamber by syringe. This procedure provided a 50 mm depth Minimum Efficiency Reporting Value (MERV) 11
chamber air concentration of ~700 ;g m3 of TSP, recorded with a pleated panel HVAC filter (Camfil Rydalmere, Australia). Whilst it is
DustTrack II 8532 optical laser nephelometer (TSI, Shoreview, recognised that 11.25 L s1 may not be the optimal test air flow rate
Minnesota). for the HVAC filter, the tests were conducted at that air flow rate to
enable comparison across equivalent volumetric air flow across
treatments, see section 2.5. A procedural control treatment was
2.3. Single pass efficiency of the biofilter also established, testing the draw down effect of aerosol deposition
on the system casing with no contents, and as no significant dif-
The purpose of these tests was to characterize the filtration ef- ference was observed (ANOVA: P ¼ 0.664), between this treatment
ficiencies of the system for a single pass of aerosol through the unit. and the empty duct, the procedural control data was utilised for all
Single-pass removal efficiency (SPRE) is defined as the fraction of SPRE calculations (see section 2.4). This approach was considered
the target pollutant that is captured by the cleaner. The goal was to reliable for evaluating the actual filter efficiency, as it takes into
characterize the efficiency of the system tested over a range of flow account any retention of inlet particles along the ducts between the
rates for aerosol diameters of 2.5 mm, 10 mm and total suspended generation point and the sample location in the duct that cannot be
particles (TSP). This was achieved by placing the biofilter within a attributed to the filter itself.
sealed flow duct and measuring the aerosol concentration down-
stream of the unit with the laser nephelometer (Fig. 1 and Fig. 2),
with effluent air flow controlled with a vacuum pump with a stable 2.4. Data analysis
volumetric flow.
Single-pass efficiency was determined at the flow rate deter- For each closed chamber draw down experiment, real-time
mined to be the most efficient; 11.25 L s1 (derived from section concentration data was obtained which was used to construct
2.2). For this purpose, a test flow-through chamber was designed, concentration versus time profiles. From this data, the total decay
compatible with the dimensions of the bioflter system, constructed constants (k) were calculated, which characterise the rate at which
from 10 mm thickness acrylic sheet. This system was designed particles are removed from chamber air, using Equation (1) [33].
specifically to allow the inlet chamber fan to positively pressurize Briefly, k was determined by fitting an exponential function to the

Fig. 1. Schematic of the single pass efficiency test apparatus set up: A ¼ Combustion chamber; B ¼ Digital pressure differential sensor; C ¼ Axial impeller; D ¼ Plenum within
system module; E ¼ Custom fit duct; F ¼ Biofilter packing medium; G ¼ Laser nephelometer; H ¼ Exhaust vacuum pump.
284 P.J. Irga et al. / Building and Environment 115 (2017) 281e290

Fig. 2. Photograph depicting experimental testing apparatus used to expose the modular botanical air biofilter to a single pulse event of particular matter.

PM concentration curves, between peak concentration, and the different treatments, as well as the residence time distribution
point where concentrations fell below ambient (15 g m3). (RTD) differences (which is the amount of time a pollutant that is
Equation 1: not filtered spends within the system until reaching the detector),
  the area under the concentration versus time data was calculated
0 Co 1 and subsequently used as the dependent variable. Comparisons
C were made utilising General Linear Model ANOVA after the data
k ¼ ln@ A
t was adjusted to conform to the normality and homogeneity of
variance the assumptions of the analysis, with pairwise post hoc
comparisons made using Tukey's test. Further, a priori power
where:
analysis was conducted, utilising pilot data results, which indicated
that sample sizes of 25 independent replicates per treatment were
C¼aerosol concentration at time t [mg.m3]
adequate to provide meaningful comparisons at alpha ¼ 0.05.
Co ¼peak aerosol concentration [mg.m3]
Single pass filtration efficiency (h) was calculated using Equa-
k¼overall decay rate constant
tion (2) [33]:
t¼time [h]
 
Cin  Cout
The constant k combines both the effect of the air filter and h¼  100%
Cin
deposition within the chamber. It was unknown a priori what the
effect of deposition was going to have on the decay constant, thus where:
the need for the procedural control. The natural decay constants
calculated for each test using the preceding formula were averaged Cin ¼ integration of the average particle concentration decay
to produce a single natural decay constant for each particle size curves in the duct with no application of treatment
class. All chamber tests were run in triplicate. To enable relative Cout ¼ integration of the average particle concentration decay
comparisons amongst air flow rates on efficiency and the curves in the duct with application of treatment
comparative effect of the botanical component of the system, two
factor Analysis of Variance was conducted utilising k as the data Particle resolved CADR derived from the single pass efficiency
variable. data CADR was calculated utilising Equation (3) [12,33]:
To assess the functional performance of the removal of partic-
ulates in the controlled test environment, clean air delivery rate CADR ¼ h  G
(CADR) was calculated. CADR is a metric that represents the volu-
metric output of ‘clean air’ an air filtration device provides per unit where:
time. Whilst CADR could be calculated from static chamber test
decay curves by taking the log loss function of particle concentra- h ¼the single pass efficiency of the system
tion corrected for the rate of natural decay, the chamber used in the G¼airflow rate through the system [L.s1]
current experiment was considered too small relative to normal
building sizes for the results to be representative of in situ func-
tional PM removal, and thus CADR was not calculated using this 3. Results
method.
Real-time concentration data from each single pass efficiency 3.1. Filtration decay rates of the biofilter
test was used to construct concentration versus time profiles. In
order to account for the differences in pressure created by the Total decay constants for TSP, PM10 and PM2.5 for both the active
P.J. Irga et al. / Building and Environment 115 (2017) 281e290 285

green wall system, and the system without the botanical compo- botanical component of the system was contributing to differences
nent of the biofilter are displayed in Figs. 3e5. With each stepwise in the packing media matrix, lessening the removal efficiency of the
increase in air flow through the system, particle removal from the system. Interestingly, the botanical treatment maximum filtration
chamber air was increased in most cases, irrespective of the pres- efficiency for TSP peaked at 11.25 L s1, with any increases in air
ence of the botanical component of the system. With increasing air flow rate met with a reduction in efficiency. It is likely that air flow
flow through the system, both treatments significantly increased rates above this speed cause contaminated air to be forced out of
the rate at which particles were removed from the chamber air the filter matrix faster than the time required for effective filtration
(overall within air flow treatment differences ANOVA: P ¼ 0.000). to occur. It cannot be determined from the available data how the
However, the botanical treatment maximum filtration efficiency for botanical component of the biofilter led to this effect. However, it
TSP peaked at 11.25 L s1, with any increases in air flow rate met appears likely that the root systems of the botanical component
with reduced removal efficiency, while the system without the have an effect on the air filled porosity and filtration dynamics of
botanical component of the biofilter maintained the same removal the system, although this observation is contradicted by the greater
efficiency with increased air flow rate (2-way ANOVA interaction air flow rates for given fan speeds detected here. The previous work
factor: P ¼ 0.000). Similar results were obtained for PM10 and PM2.5 conducted on this biofiltration system assessed pressure drop
for the system with the botanical component, with peak efficiency across the biofilter, air-flow distribution through it as well as flow
seen at air flow rates of 11.25 L s1 and any increases met with rate for various airstream pressures, indicating that a range of
equivalent or diminished efficiency (Tukey's; P  0.05 for the variables which include moisture content, plant type and associ-
comparison of PM removal at air flow rate 11.25 L s1 and air flow ated plant root volume and morphology do affect air flow and thus
rates less than and including 7.5 L s1). biofilter efficiency [29].
To date, limited work has investigated botanical biofilters for
3.2. Single-pass efficiency their capacity in reducing PM levels. Lee et al.'s study [34] assessed
active biofilter PM removal efficiency, focusing on the effect of the
Single-pass efficiency (h) represents the percentage of particu- botanical component; which at the time of writing, to this authors'
lates removed from the air stream as it passed through the device. knowledge, is the only published investigation testing an active
Particulate matter removal rates for each treatment are shown in green wall system for the reduction of PM. Their system is me-
Fig. 6, and a comparative summary of treatment CADR values chanically similar to the system presented here: it involves assisted
derived from h for the particle fractions measured is presented in aeration ducted to the packing media, and out through the botan-
Table 1. Removal was achieved in all treatments, and it can be seen ical component of the system, however prior to the air flowing
that there were strong similarities in response across the three through any biological components, the air travels through a hu-
particle sizes tested. As expected, the HVAC filter treatment showed midifying column, which has the effect of both converting the air
the highest removal rates, demonstrating removal efficiencies of pollutants to liquid phase as well as negating the need for irrigation
85.77 ± 8.70% for TSP, 78.44 ± 16.86% for PM10, and 77.76 ± 11.69% as the humidity supplies the botanical component with sufficient
for PM2.5. Comparatively, the system with the botanical component moisture. Lee et al. [34] demonstrated that their biofilter system
present recorded removal efficiencies of 53.35 ± 9.73% for TSP, had removal efficiencies for PM10 of 30% when the biofilter was
53.51 ± 15.99% for PM10, and 48.21 ± 14.71% for PM2.5, which were only comprised of packing media, compared to 40% when the
significantly lower than the removal efficiencies recorded for the biofilter also included the botanical component. Further, their
HVAC filter for TSP (Tukey's; P ¼ 0.022), PM10 (Tukey's; P ¼ 0.000), system had removal efficiencies for PM2.5 of only 2% when the
and PM2.5 (Tukey's; P ¼ 0.001). Additionally, when the system was biofilter was only comprised of packing media, and 4% when the
tested with no botanical component present, removal efficiencies biofilter also included the botanical component. Comparatively, the
were 64.66 ± 8.66% for TSP, 77.06 ± 14.43% for PM10, and removal efficiencies documented in the current study were
66.38 ± 12.12% for PM2.5. The PM10 removal efficiency was signifi- considerably higher, especially for PM2.5. Whilst PM10 reduction
cantly higher compared to the system with botanical component efficiency was higher in Lee et al.'s [34] system with the botanical
present (Tukey's; P ¼ 0.005), however no significant differences component present, these results contrast to those of the current
observed when comparing TSP and PM2.5 removal (Tukey's; study, where the system solely comprised of the packing media had
P ¼ 0.749 and P ¼ 0.066 respectively). Interestingly, the biofiltration a greater filtration efficiency. This could be due to plant species
system with no botanical components demonstrated statistically differences. Lee et al. [30] utilised Epipremnum. aureum which tends
equivalent efficiencies to the HVAC filter for PM10 and PM2.5 to have long fibrous roots, with a very high shootetoeroot ratio,
removal (Tukey's; P ¼ 0.340 and P ¼ 0.104 respectively). whereas the C. comosum we tested forms fleshy tuberous roots
(Fig. 7). The C. comosum roots may have altered the air filled
4. Discussion porosity of the packing media to a greater degree or in a different
way to those of E. aureum, altering the filtration matrix and thus the
This study provides data for the characterisation of the removal PM removal efficiency. This indicates that variability occurs
efficiency of atmospheric particles by an in-room botanical biofilter amongst the plant types that may be utilised in biofiltration sys-
system. We demonstrate that an enhanced removal of particulates tems, and that other plants may be identified that can produce
from air can be achieved by ventilation of the polluted airstream more efficient systems. Similarly, the system tested here was
through the biofilter tested. As expected, with increasing air flow modular, which may influence the pattern of root distribution.
rates, an increase in filtration efficiency in the static chamber was Plants will develop roots within the growth medium towards the
generally observed. With increasing flow rate, the air flow pene- lower end of the system due to gravitropic effects. Upwards root
trates a greater volume of the substrate, and thus removal effi- growth is limited, but will occur in lateral and secondary roots [35].
ciency increases as a greater proportion of the substrate contributes The likelihood of upwards root growth occurring may be increased
to the biofiltration process. as a reaction to waterlogging of the soil, which may occur if plant
The system's single pass removal efficiency was 53% for all wall casings units do not have appropriate drainage or are incor-
fractions of particles, 54% for PM10 and 48% for PM2.5. Differences rectly maintained. From a practical viewpoint, slow growing plants
were observed in filtration efficiency between the vegetated and may provide advantages in living walls as they require less main-
non-vegetated system at the higher air flow rates, indicating the tenance, resulting from reduced root migration, however slow
Fig. 3. Total suspended particle decay constants (k), across air flow rates for the system containing just packing media (diamond), and with the system botanical component present
(square).

Fig. 4. PM10 decay constants (k), across air flow rates for the system containing just packing media (diamond), and the system with the botanical component present (square).

Fig. 5. PM 2.5 decay constants (k), across air flow rates of the s system containing only packing media (diamond), and the system with the botanical component present (square).
P.J. Irga et al. / Building and Environment 115 (2017) 281e290 287

Fig. 6. Calculated single-pass removal efficiency (h) for TSP, PM10 and PM2.5. Data is expressed as mean and SE, n ¼ 25.

Table 1
CADR (m3.h1) calculated from single pass efficiency tests, when air flow is set to optimal.

Pollutant HVAC filter System without botanical component System with botanical component

TSP 46.01 34.68 28.61


PM10 42.01 41.33 28.70
PM2.5 41.71 35.60 25.86

initial root growth may result in difficulties establishing plants in tested in the current work). Consequently, whilst the PM removal
vertical growing media. and thus CADR are somewhat less efficient than to an in-duct HVAC
Clearly, targeted development of the packing media could filter, the sheer volume of modules in commercial systems will
further increase the PM removal efficiency of any biofiltration subsumes these efficiency differences, producing 1200 L s1 of
system, whilst sustaining concomitant improvements to the clean air.
botanical efficiency and distribution within the system. If the tor- Most botanical biofilter research to date has focussed on VOC
tuosity, surface area, effective pore size and electrostatic charac- and CO2 removal from indoor environments, including: Darlington
teristics of a medium can be tailored to entrap more particulate et al. [21] (VOCs: toluene, ethylbenzene and o-xylene.), Wang and
matter, or have a greater capacity for collecting PM, more effective Zhang [22] (VOCs: formaldehyde and toluene), and Xu et al. [41]
systems will result. The effectiveness of various biofilter media (i.e. (VOCs: formaldehyde). Thus CADR has only previously been
without plants) to remove VOC air pollutants has received calculated based on VOC draw down. Wang et al. [42] detecting
considerable research [36e38], and has identified materials that CADR values based on formaldehyde removal of
function effectively in specific environments. Green wall and at- 232e759 m3 s1m2 for their active green wall system. Addi-
mospheric phytoremediation technology is not yet at this stage of tionally, Torpy et al. [16] established CADR based on photosynthetic
development, and faces the challenge of maximising PM removal, CO2 removal, finding that their living wall produced 0.09 L s1m2
along with VOC and other pollutant removal, whilst maintaining green wall of clean air, estimating that a 1 m2 of green wall could
plant viability and minimising maintenance. Future work that fo- supply 1.67% of the ventilation requirements of a single occupant,
cuses on the development of medium characteristics for greater air and a 6 m2 could replace 10% of their ventilation requirements. As
pollutant removal would be percipient. the work presented here is the first to document SPRE and CADR for
The particle resolved CADRs for the botanical biofilter tested PM using a botanical biofilter, the current findings thus provide a
indicate that even at the lowest removal efficiency, each 0.25 m2 basis for comparison for future developments.
module produces 5.42 L s1 of particle-free air per hour. The US Research on the long term performance of the device tested
regulation for minimum ventilation rates in office space occupied at here to remove particles in in situ conditions is now required,
a density of 10 m2 per person is 5.5 L s1 per person of fresh air [39]. where realistic levels of particle loading within the biofilter and
The system investigated here, in its most effective setting produced other environmental parameters such as air exchange rates, relative
24 L s1m2 green wall of “fresh” air, i.e. air from which the PM has humidity and temperature are present. Further, energy perfor-
been removed. Based on these estimates, 1 m2 of green wall could mance implications are strongly tied to fan design and fan speed
supply the ventilation requirements of four occupants. control, with different axial impeller permutations behaving
A metric used to indicate the effectiveness of the biofilter in an differently under high static pressures. Development of blowers
in situ applications is the air cleaner effectiveness (G) [40]. This that maintain constant airflow rate despite changes in back pres-
value indicates the capacity of the biofilter to supplement the sure due to watering or root growth and particle accumulation over
ventilation for a conventional office. For a room volume of 50 m3, to long term use would be beneficial, although they may increase
supply a ventilation equivalent of 4 air changes per hour, the bio- power draw and affect the sustainable marketability of the device.
filter in its most effective operational setting would require 9.36 Passive convection technology may also play a future role in sus-
modules (2.34 m2) to fully replace mechanical ventilation. Typi- tainable ventilation systems that include biofiltration components
cally, current botanical biofilter dimensions are usually ~50 m2, to maintain indoor air quality. Finally, once operational parameters
incorporating 200 modules (see Fig. 8, an existing active green wall are firmly established, future work should focus on the adoption of
installation of 6 m  9.5 m, made up of 228 modules of the type this technology for indoor air cleaning, either as part of an HVAC
288 P.J. Irga et al. / Building and Environment 115 (2017) 281e290

Fig. 7. A comparison of the roots systems of C. comosum and E. aureum. C. comosum (left) displays large and tuberous roots with fine secondary root growth, whilst E. aureum (right)
displays lengthened tap roots, with occasional aerial roots used to cling onto vertical surfaces, with limited secondary root formation.

system or operated as a standalone unit to provide the required PM10, and 48.21 ± 14.71% for PM2.5. Whilst the system is currently
clean airflow rate. somewhat less efficient than a conventional HVAC filter, the system
shows potential, and further developments that improve its effi-
5. Summary and conclusion ciency to match those demonstrated by current air filtration
methodologies should be possible. Additionally, the system's ability
Although there have been many studies conducted on the per- to remove VOCs and CO2 and modulate temperature and humidity
formance of HVAC filters and portable in-room air cleaners, there is make the device superior to most non-biological systems as general
little research on the evaluation of botanical biofilters for particle air quality maintenance devices. Nonetheless, further controlled
removal. The work presented here quantifies the functionality of a laboratory experiments are needed to investigate the long term
modular botanical biofilter, with regards to the removal of partic- performance of the system, and to better describe the simultaneous
ulate matter from air. The botanical treatment maximum filtration removal of PMs, VOCs and CO2. These investigations will provide
efficiency for TSP peaked at an airflow rate of 11.25 L s1 per empirical data on which to develop a simulation model that can be
0.25 m2 modular unit, with any increases in air flow rate met with used to optimize the system's design, as well as to advance the
reduction in PM filtration efficiency. Similar results were obtained implementation of the device.
for PM10 and PM2.5, with peak efficiency also at air flow rates of
11.25 L s1. At its most efficient airflow rate, the system recorded
removal efficiencies of 53.35 ± 9.73% for TSP, 53.51 ± 15.99% for
P.J. Irga et al. / Building and Environment 115 (2017) 281e290 289

Fig. 8. Existing active green wall installation, dimensions 6 m  9.5 m, made up of 228 modules (Lend Lease head office, Barangaroo, Sydney Australia). (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article.)

Acknowledgments dust loading on long term portable air cleaner performance, Build. Environ.
112 (1) (2017) 261e269, http://dx.doi.org/10.1016/j.buildenv.2016.11.001.
[11] A. Luengas, A. Barona, C. Hort, G. Gallastegui, V. Platel, A. Elias, A review of
This research was funded by an Australian Government indoor air treatment technologies, Rev. Environ. Sci. Bio/Technology 14 (3)
Department of Industry Research Connections linkage project with (2015) 499e522.
Junglefy Pty Ltd. None of the authors have any commercial [12] ANSI/AHAM AC-1, Method for Measuring the Performance of Portable
Household Electrical Room Air Cleaner, American National Standard, 2006.
connection with either the funding body or Junglefy Pty Ltd. [13] O. Sepp€ anen, Ventilation strategies for good indoor air quality and energy
beyond the scope of this research project. We would like to thank efficiency, Int. J. Vent. 6 (4) (2008) 297e306.
Gemma Armstrong for her technical assistance, and Ashley Naomi [14] L. Boskovic, I.S. Altman, I.E. Agranovski, R.D. Braddock, T. Myojo, M. Choi, In-
fluence of particle shape on filtration processes, Aerosol Sci. Technol. 39 (12)
Jane Douglas and Iain Berry for their professional editing help with (2005) 1184e1190.
this work. [15] C. Chen, B. Zhao, Review of relationship between indoor and outdoor parti-
cles: I/O ratio, infiltration factor and penetration factor, Atmos. Environ. 45 (2)
(2011) 275e288.
References [16] F.R. Torpy, M. Zavattaro, P.J. Irga, M.D. Burchett, Assessing the Air Quality
Remediation Capacity of the JUNGLEFY Breathing Wall - Modular Plant Wall
[1] N.E. Klepeis, W.C. Nelson, W.R. Ott, J.P. Robinson, A.M. Tsang, P. Switzer, system., Research Report, School of Life Sciences, Faculty of Science, Univer-
J.V. Behar, S.C. Hern, W.H. Engelmann, The national human activity pattern sity of Technology, Sydney, 2015.
survey (NHAPS): a resource for assessing exposure to environmental pollut- [17] G. Soreanu, M. Dixon, A. Darlington, Botanical biofiltration of indoor gaseous
ants, J Expo anal Environ Epidemiol 11 (3) (2001) 231e252. pollutants e a mini-review, Chem. Eng. J. 229 (2013) 585e594.
[2] R. Silva, J. West, Y. Zhang, S. Anenber, J. Lamarque, D. Shindell, W. Collins, [18] M. Dela Cruz, R. Müller, B. Svensmark, J.S. Pedersen, J.H. Christensen,
S. Dalsoren, G. Falugeni, G. Folberth, Global premature mortality due to Assessment of volatile organic compound removal by indoor plantsda novel
anthropogenic outdoor air pollution and the contribution of past climate experimental setup, Environ. Sci. Pollut. Res. 21 (13) (2014) 7838e7846.
change, Environ. Res. Letters 8 (2013) 034005 (11pp). [19] F.R. Torpy, M. Zavattaro, P.J. Irga, Green wall technology for the phytor-
[3] Z. Li, Z. Yuan, Y. Li, A.K.H. Lau, P.K.K. Louie, Characterization and source emediation of indoor air: a system for the reduction of high CO2 concentra-
apportionment of health risks from ambient PM10 in Hong Kong over tions, Air Qual. Atmos. Health (2016) 1e11.
2000e2011, Atmos. Environ. 122 (2015) 892e899. [20] Z. Wang, J.S. Zhang, Characterization and performance evaluation of a full-
[4] M.A. Bari, W.B. Kindzierski, Fine particulate matter (PM2.5) in Edmonton, scale activated carbon-based dynamic botanical air filtration system for
Canada: source apportionment and potential risk for human health, Environ. improving indoor air quality, Build. Environ. 46 (3) (2011) 758e768.
Pollut. 218 (2016) 219e229. [21] A. Darlington, J. Dat, M. Dixon, The biofiltration of indoor air: air flux and
[5] Z.D. Ristovski, B. Miljevic, N.C. Surawski, L. Morawska, K.M. Fong, F. Goh, temperature influences the removal of toluene, ethylbenzene, and xylene,
I.A. Yang, Respiratory health effects of diesel particulate matter, Respirology Environ. Sci. Technol. 35 (2001) 240e246.
17 (2) (2012) 201e212. [22] Z. Wang, Dynamic Botanical Filtration System for Indoor Air Purification,
[6] L.C. Ng, A.K. Persily, S.J. Emmerich, IAQ and energy impacts of ventilation Syracuse University, Syracuse, NY, 2011.
[23] L. Perez-Urrestarazu, R. Ferna ndez-Can ~ ero, A. Franco, G. Egea, Influence of an
strategies and building envelope airtightness in a big box retail building, Build.
Environ. 92 (2015) 627e634. active living wall on indoor temperature and humidity conditions, Ecol. Eng.
[7] R.B. Mosley, D.J. Greenwell, L.E. Sparks, Z. Guo, W.G. Tucker, R. Fortmann, 90 (2016) 120e124.
[24] M.J.M. Davis, F. Ramirez, M.E. Pe rez, More than just a green façade: vertical
C. Whitfield, Penetration of ambient fine particles into the indoor environ-
ment, Aerosol Sci. Technol. 34 (1) (2001) 127e136. gardens as active air conditioning units, Procedia Eng. 145 (2016) 1250e1257.
[8] R. Meier, C. Schindler, M. Eeftens, I. Aguilera, R.E. Ducret-Stich, A. Ineichen, [25] M.M. Davis, S. Hirmer, The potential for vertical gardens as evaporative
M. Davey, H.C. Phuleria, N. Probst-Hensch, M.-Y. Tsai, N. Künzli, Modeling coolers: an adaptation of the ‘Penman Monteith Equation’, Build. Environ. 92
indoor air pollution of outdoor origin in homes of SAPALDIA subjects in (2015) 135e141.
Switzerland, Environ. Int. 82 (2015) 85e91. [26] M.J.M. Davis, F. Ramírez, A.L. Vallejo, Vertical gardens as swamp coolers,
[9] J.A. Rosati, J. Thornburg, C. Rodes, Resuspension of particulate matter from Procedia Eng. 118 (2015) 145e159.
carpet due to human activity, Aerosol Sci. Technol. 42 (6) (2008) 472e482. [27] P.J. Irga, P. Abdo, M. Zavattaro, F.R. Torpy, An assessment of the potential
[10] M.S. Zuraimi, M. Vuotari, G. Nilsson, R. Magee, B. Kemery, C. Alliston, Impact of fungal bioaerosol production from an active living wall, Build. Environ. 111
290 P.J. Irga et al. / Building and Environment 115 (2017) 281e290

(2017) 140e146. [36] A.K. Mathur, C.B. Majumder, S. Chatterjee, Combined removal of BTEX in air
[28] H. Gawron  ska, B. Bakera, Phytoremediation of particulate matter from indoor stream by using mixture of sugar cane bagasse, compost and GAC as biofilter
air by Chlorophytum comosum L. plants, Air Qual. Atmos. Health (2014) 1e8. media, J. Hazard. Mater. 148 (1e2) (2007) 64e74.
[29] P. Abdo, B.P. Huynh, V. Avakian, T. Nguyen, F.R. Torpy, P.J. Irga, Measurement [37] A.H. Wani, R.M.R. Branion, A.K. Lau, Biofiltration: a promising and cost-
of air flow through a green-wall module, in: Australasian Fluid Mechanics effective control technology for Odors, VOCs and air toxics, J. Environ. Sci.
Conference, 2016. Perth, Australia. Health. Part A Environ. Sci. Eng. Toxicol. 32 (7) (1997) 2027e2055.
[30] Australian Federal Government, Department of the Environment and Energy, [38] L. Malhautier, N. Khammar, S. Bayle, J.-L. Fanlo, Biofiltration of volatile organic
Fuel Quality Standards Act in. Energy (Ed.), 2000. compounds, Appl. Microbiol. Biotechnol. 68 (1) (2005) 16e22.
[31] International Organization of Standardization, Paper and board - determina- [39] ASHRAE, Standard, 62.1-2010 Ventilation for Acceptable Indoor Air Quality,
tion of grammage, ISO 536 (2012) 2012e2107. American Society of Heating, Refrigerating and Air-Conditioning Engineers,
[32] Australian Forestry Standard, ASF 4707, 2014, p. 2014. Atlanta, GA., 2010.
[33] U.S. EPA, Evaluation of In-room Particulate Matter Air Filtration Devices, US [40] M. Ward, J.A. Siegel, R.L. Corsi, The effectiveness of stand alone air cleaners for
EPA, Washington, DC, 2008. shelter-in-place, Indoor Air 15 (2) (2005) 127e134.
[34] C.H. Lee, B. Choi, M.Y. Chun, Stabilization of soil moisture and improvement of [41] Z. Xu, L. Wang, H. Hou, Formaldehyde removal by potted plantesoil systems,
indoor air quality by a plant-biofilter integration system, Kor. J. Hort. Sci. J. Hazard. Mater. 192 (1) (2011) 314e318.
Technol. 33 (5) (2015) 751e762. [42] Z. Wang, J. Pei, J.S. Zhang, Experimental investigation of the formaldehyde
[35] L. Jørgensen, D.B. Dresbøll, K. Thorup-Kristensen, Spatial root distribution of removal mechanisms in a dynamic botanical filtration system for indoor air
plants growing in vertical media for use in living walls, Plant Soil 380 (1) purification, J. Hazard. Mater. 280 (2014) 235e243.
(2014) 231e248.

You might also like