Performance evaluation of a scoria-compost biofilter treating xylene vapors
- 1.7k Downloads
The removal of xylene vapors was studied in a biofilter packed with a new hybrid (scoria/compost) packing material at various inlet loads (IL) and empty bed residence times (EBRT) of 90, 60, and 40s. The best performance was observed for EBRT of 90s, where a removal efficiency of 98% was obtained under steady state condition for inlet xylene concentration of 1.34 g m−3, while a maximum elimination capacity of 97.5 g m−3 h−1 was observed for IL of 199.5 g m−3 h−1. Carbon dioxide production rates and the microbial counts for xylene-degraders followed xylene elimination capacities. Overall look to the results of this study indicates that the scoria/compost mixture could be considered as a potential biofilter carrier, with low pressure drop (here <4 mm H2O), to treat air streams containing VOCs.
KeywordsXylene Biofilter Scoria Compost Elimination capacity
Xylene, together with benzene, toluene, and ethyl benzene, constitute the volatile organic compounds (VOCs) group BTEX. Xylene isomers namely p-xylene, m-xylene, and o-xylene are listed as hazardous and toxic atmospheric contaminants under Clean Air Act Amendments (CAAA) . Chronic exposure to xylene is associated with adverse effects on the liver, kidneys and the central nervous system . Xylene is a major constituent of gasoline and is used as a solvent in many production industries including printing, rubber, leather, painting and varnishing industries. In general, over 60% of the total emissions of xylene into the atmosphere originate from industrial facilities, especially petrochemical plants ,.
Among the air pollution control technologies, the biological treatment of VOCs provides an environmentally-friendly and low-cost alternative to other physicochemical treatment technologies -. The most widely used biological air treatment process is biofiltration . The performance of biofilter significantly depends on the packing material properties . Some of the desirable media properties include balanced chemical composition, relatively high water holding capacity, high microbial population density, and structural integrity ,. These properties, except the later one, could be found in organic materials such as compost and peat, which allow the biofilter to have a quick startup and high elimination capacity. However, organic materials have a relatively low durability and tend to settle and compact, which in turn result in increased pressure drop and channeling . On the other hand, the use of an inert material with rigid structure and large pores as filter bed allows a better gas distribution inside the reactor than organic carriers and minimizes the pressure drop build ups ,. Consequently, using a mixture of organic and inert material inoculated with acclimated microbial consortium not only might improve the performance of biofilters, but also provides low pressure drops.
Many studies have been conducted on the biofiltration of xylene from waste air streams mainly in mixture with other VOCs. Most of them demonstrated that the removal of xylene isomers is always less efficient among the BTEX compounds -. In other hand, the biofiltration of xylene as the sole pollutant has been less considered. Saravanan and Rajamohan  studied the biofiltration of xylene vapors on a laboratory scale biofilter packed with press mud as filter material inoculated with activated sludge and reported a maximum elimination capacity of 67 g m−3 h−1 for inlet load of 103 g m−3 h−1 and EBRT of 42 s.
To the best of our knowledge, no attempt has been made on the removal of VOCs in biofilters packed with a mixture of scoria and compost. Scoria is a vesicular pyroclastic rock with basaltic composition, which is light in weight, porous, and cheap. It is abundant in many places worldwide . In this study a mixture of scoria (pre-soaked in an adapted inoculum) and compost was used for the biofiltration of xylene. The aim of this study was to evaluate the performance of the biofilter under various operating conditions in terms of xylene removal efficiency and elimination capacity. The carbon dioxide evolution and microbial population distribution along the bed length were also monitored.
Materials and methods
Inoculum and packing media preparation
Biofilter media characteristics
Particle size (mm)
Bulk density (kg m−3)
Initial moisture (%)
Water holding capacity (kg kg−1 dry weight)
C/N ratio (−)
Special surface area (m2 g−1)
Compressed air was continuously passed through a granular activated carbon column to remove moisture, oil and particulate matter. The purified air was divided into two streams; the major one was passed through a water column in order to increase the relative humidity of inlet air and the minor one was bubbled through a glass bubbler containing xylene solution (Merck, Germany). The bubbler was held in a thermostatically-controlled water bath. Afterward the streams were mixed and fed to the bottom of the biofilter column in a counter-current flow mode. Xylene concentration was maintained at the desired value by adjusting the flow rate of air passed through the bubbler, temperature of water bath, and the level of pollutant in bubbler.
Gas phase xylene was measured by a gas chromatograph equipped with flame ionization detector (GC-FID) (Agilent GC, 7890A). The column used was an Agilent 19091S-433 capillary column, 30 m × 250 μm × 0.25 μm. High purity helium gas (99.995%) was the carrier gas and supplied at a flow rate of 1.11 ml min−1. The temperatures of column oven, injector and detector were 150, 230 and 250°C, respectively. The method detection limit (MDL) for xylene analysis was estimated as 11.1 μg l−1. The calibration curve was prepared by injecting known amounts of the xylene into a glass bottle sealed with Teflon septum according to the standard procedure . Air samples of 100 μl were drawn from the bottle and various sampling ports along the biofilter column with a 1 mL gas tight syringe (Hamilton, USA), and directly injected into GC. A Guardian Plus Model D100 IR CO2 analyzer (Edinburgh Sensors Ltd., Munchen, Germany) (0–3000 ppm) was used to analyze carbon dioxide concentration along the biofilter column. Total organic compounds (TOC) in the leachate were measured using a TOC analyzer (Beckman 915A, USA). In order to break microbial flocs into small pieces and achieve a homogeneous TOC concentration, the leachate sample was surged by ultrasonic wave. The moisture content of packing material was determined as weight loss after drying at 105°C for 24 hr.
The temperatures at midlevel of each section and at gas sampling ports were measured by online K-type thermocouple and a digital thermometer, respectively. The pressure drop across the biofilter was measured using a water column manometer. Scanning electron micrographs of the scoria were carried out using a Vega/Tescan scanning electron microscope (SEM). For the sample taken at the end of experiment, the microorganisms were fixed with 3% glutaraldehyde aqueous solution overnight. The fixed sample was washed with phosphate buffer, dehydrated by placing in 30%, 50%, 70% and 100% ethyl alcohol, dried and then covered with a gold layer.
Microbial cell counts
Almost one gram of the packing material was taken from each section of the biofilter and mixed with 9 mL sterile saline solution (0.9% w/v NaCl). The samples were then vortexed for 3 min and serially diluted up to 10−10 in sterile saline solution. Aliquots of 0.1 ml were spread over the surfaces of agar plates. The Nutrient Agar amended with Nistatin (to inhibit fungal growth) and the Potato Dextrose Agar (PDA) amended with chloramphenicol (to inhibit bacterial growth) were used for culturing bacteria and fungi, respectively. Furthermore, a sterile mineral medium containing 1.5% (w/v) agar was used for the enumeration of xylene-degraders. Xylene vapor as sole carbon source was supplied through placing plates in a sealed container. The cultured plates were incubated at 28–30°C for 5–7 days. The counts were reported as colony forming unit (CFU) g−1 of packing material (dry basis).
where Q is the air flow rate (m3 min−1), V is the volume of packed bed (m3), Cin and Cout are the inlet and outlet xylene concentrations (g m−3), and CO2,in and CO2,out are the inlet and outlet carbon dioxide concentrations (g m−3), respectively.
Results and discussion
Overall performance of biofilter
Biofilter operating conditions
Days of operation
Inlet concentration range (g m−3)
Inlet loading rate range (g m−3 h−1)
During phase I, the inlet concentration was increased slowly in order to prevent shock loading of biofilter and to reach a maximum performance (ECmax). From day 10 to 25, the inlet concentration was varied from 1.32 to 1.58 g m−3 and the removal efficiency was maintained over 92.6%. The removal efficiency decreased significantly to 53.6% as the xylene concentration increased to 4.47 g m−3 by day 44. From days 46 to 56, the inlet concentration was raised and maintained at highest levels during the whole experiment, i.e. 4.7 to 5.3 g m−3 with corresponding loads of 187.8 to 212.8 g m−3 h−1. In this period a pseudo steady state was achieved, since the xylene removal efficiencies varied from 45.3 to 47.9%. According to Rahul et al.  the pseudo steady state was presumed when the changes in the xylene removal efficiency were within 5% for three successive days.
The next two phases of operation were aimed at investigating the biofilter performance under lower EBRTs. In phase II, the EBRT was reduced to 60 s and the biofilter could provide over 80% removal of xylene for the inlet concentrations below 1.14 g m−3. The removal efficiency sharply declined to an average of 33.4 ± 4.1% during the days 68–76, when the inlet xylene concentration was increased to a range of 3.22 to 3.68 g m−3, with loading rates of 193.4-220.9 g m−3 h−1. The last phase of the experiment commenced on day 77; the EBRT was further decreased to 40 s and a removal efficiency of 81.0% was observed for the xylene concentration of 0.26 g m−3 (IL = 23.3 g m−3 h−1). With the sudden increase in the loading rate (84.9 g m−3 h−1), the removal efficiency of xylene decreased to 53.6%, but it recovered to 68.2% after 3 days. Similar to the previous phases, the biofilter experienced high xylene loading rates ranged from 192.7 to 204.6 g m−3 h−1 during days 90–95 and in accordance with this operating condition, an average xylene removal efficiency of 30.3 ± 2.2% was obtained. At the end of this phase, the xylene concentration was lowered and the removal efficiency recovered to 89% for loading rate of 22.2 g m−3 h−1.
In this study, a removal efficiency of 98% was obtained under steady state condition for the inlet concentration of 1.34 g m−3 at EBRT of 90s. Comparatively, Saravanan and Rajamohan  observed a maximum xylene removal efficiency of 50% for the inlet concentration of 1.2 g m−3 at an EBRT of 88.2 s in a biofilter packed with press mud.
The results obtained during this study indicate that the xylene removal efficiency decreased either by an increase in xylene concentration in the entering air or by a decrease in EBRT. At lower EBRTs, the microbial population on the surface of media gets less contact time for the degradation of xylene. Besides, the increase in xylene loads towards the highest levels during the last two phases, especially phase II, occurred in a more rapid manner than that of the phase I. Then, there was more time for microorganisms to adapt the increased xylene at phase I.
In case of EBRT 90 s; (i) under DLR, the elimination capacity increased to 94.3 g m−3 h−1 when the inlet loading rate was increased to 137.8 g m−3 h−1. In DLR, an increase in the inlet concentration of xylene enhances the transfer rate of this pollutant from the gas-phase to the biofilm and more microorganisms actively participate in the biodegradation process. (ii) under RLR, the EC stabilized around 93.9 gm−3 h−1 for inlet loadings beyond 178.8 g m−3 h−1 and up to 212.8 gm−3 h−1. Presumably, the RLR occurs when the amount of active xylene-specific microorganisms is insufficient to degrade all the gas-phase xylene that could possibly be transferred to the biofilm. Under this circumstance, the biofilm would be fully saturated with xylene and the metabolic activity of the cells in the biofilm would be the rate limiting step . The maximum elimination capacity of the biofilter was 97.5 g m−3 h−1 for inlet xylene load of 199.5 g m−3 h−1 at EBRT of 90 s. The biofilter performance at EBRTs of 60 s and 40 s had similar trends to that of the EBRT of 90 s, with lower EC values. The maximum ECs of 79.8 and 65.5 g m−3 h−1 were achieved at EBRTs of 60 and 40 s, respectively. Jorio et al.  obtained maximum xylene elimination capacities of 67, 52, and 41 g m−3 h−1 at EBRTs of 158.9, 90.8, and 63.6 s, respectively, in a conventional biofilter packed with peat mixed with structuring and conditioning agents and initially inoculated with a microbial consortium. In other hand, Gallastegui et al.  reached a p-xylene elimination capacity as high as 130 g m−3 h−1 at high EBRTs ranging from 180 to 270 s in a biofilter packed with pelletised sawdust and pig manure.
Higher elimination capacities at EBRT of 90 s compared to EBRTs of 60 and 40 s (for the same loading rate) may be attributed to the higher residence time of carrier gas, and availability of higher concentration of xylene at higher EBRTs in a given inlet loading rate.
Xylene profile along the biofilter length
At day 95, the inlet xylene concentration was about that of the day 31 (Cin = 2.43 g m−3, IL = 97.1 g m−3 h−1), while xylene load was about that of the day 52. As shown in Figure 5, the xylene removal profile at day 95 was very similar to that of the day 52. This indicates that the xylene removal profile along biofilter length is more dependent on the inlet load than on the inlet concentration.
Carbon dioxide production
Accordingly, the value of the slope in equation  shows that about 60% of the eliminated xylene was converted to CO2. Wu et al.  obtained a mass ratio of PCO2/EC of 1.65 when they treated p-xylene in a hybrid biofilter with an added nutrient solution containing ammonium salts. Li et al.  reported that 62% of removed xylene in a bacterial and fungal biofilter was converted to CO2.
The discrepancy observed in CO2 production in comparison with the case of complete chemical oxidation of xylene can be mainly explained by the biomass production. In addition, some of the CO2 produced may partly accumulate in the liquid-phase in the form of .
Therefore, 0.71 g of dry biomass was produced per g of xylene consumed, which corresponds to a biomass yield coefficient value of 0.42 gC dry mass synthesized per gC xylene degraded.
At day 3, the counts of xylene-degraders decreased around two orders of magnitude from bottom to the top of biofilter, which can be related to the complete removal of xylene at the first two sections. The xylene degraders increased by 10-fold to an average count of 2.87 × 109 ± 4.8 × 108 CFU g−1 at day 54 of operation and then decreased by 5-fold to 5.82 × 108 CFU g−1 at day 93. These values are higher than those reported for p-xylene degraders in the biofilters packed with food waste compost (1.28 × 108 g−1 of dry compost) and pig manure compost (2.58 × 107 g−1 of dry compost) . The microbial counts for xylene degraders follow xylene elimination capacities, which were 33.0, 93.7, and 50.2 g m−3 h−1 at days 3, 54, and 93, respectively. In addition, the number of xylene degraders in all four sections of the biofilter were within the range of the same order of magnitude at days 53 (9 log CFU g−1) and 93 (8 log CFU g−1), which are related to approximately similar performance of the sections at high inlet xylene loads.
The comparison of the number of microbial cells (average ratio of bacterial to fungal CFUs =2.4 × 104:1) implies that the bacteria were the dominant microorganisms responsible for the degradation of xylene in the biofilter. Irrespective of the day of sampling and biofilter section, the average total bacterial count was 4.8 × 1010 CFU (gram of dry mass)−1 containing 10.5% xylene degraders.
Temperature and pressure drop
The pressure drop values were less than 4 mm H2O during the entire experiments. In addition, negligible compaction and deterioration of the bed was observed, indicating a good mechanical strength of the media. The low pressure drop, which is the prime prerequisite for biofilter packing material, could be related to low velocity of carrier gas, the filter media characteristics, irrigation strategy and the nature of microbial community within biofilter. By volume, about 75% of the filter material was scoria (6–13 mm particle size), which is an inert media. One major advantage of inert material on organic ones is the low pressure drop due to minimum compaction over time and good air distribution. Excess biomass accumulation is a major factor causing an increase in the pressure drop . Nevertheless, adequate irrigation of the filter bed made the removal of the excess of biomass possible . Here the filter bed was irrigated with 1 liter nutrient solution, where, 728 and 1620 mg TOC were discharged from the system at days 59 and 94, respectively. The filter bed was dominated by bacteria. Usually pressure drop in bacterial biofilters is lower than that of observed in fungal biofilters. This is because of the occupation of the free space by the mycelia. Estrada et al.  reported that the bacterial biofiltration treating a VOC mixture exhibited a final pressure drop of 60% lower than that of the fungal biofilter due to mycelial growth.
A biofilter packed with scoria/compost was employed for the removal of xylene vapors. The best performance of biofilter (ECmax = 97.5 g m−3 h−1) was observed at the highest EBRT, that may be due to higher residence time of carrier gas. The CO2 production rate and the distribution of microbial populations in the biofilter were well correlated with the xylene removal rates, indicating the biodegradation of xylene in the biofilter. The packing media provided low pressure drop and long term stability, demonstrating its potential as media in the biofiltration of VOCs.
This article is the result of Ph.D thesis approved in the Isfahan University of Medical Sciences (IUMS). The authors wish to acknowledge to Vice Chancellery of Research of IUMS for the financial support, Research Project, # 391249.
- 2.ATSDR:Toxicological Profile Information Sheet. Agency for Toxic Substances and Disease Registry, Department of Health and Human Services, Public Health Service, USA; 2003.Google Scholar
- 5.Dehghanzadeh R, Aslani H, Torkian A, Asadi M: Interaction of acrylonitrile vapors on a bench scale biofilter treating styrene-polluted waste gas streams. Iran J Environ Health Sci Eng 2011, 8: 159–168.Google Scholar
- 18.Lodge JP: Methods of Air Sampling and Analysis. Lewis Publishing Inc, New York; 1989.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.