Bioresource Technology 97 (2006) 1546–1553 Hydrogen sulfide removal by compost biofiltration: Effect of mixing the filter media on operational factors J.M. Morgan-Sagastume *, A. Noyola Department of Environmental Bioprocesses, Engineering Institute, National Autonomous University of México, Circuito Escolar, Ciudad Universitaria, Coyoacan, 04510 México City, DF, México Received 8 October 2004; received in revised form 24 May 2005; accepted 2 June 2005 Available online 26 July 2005 Abstract The overall goal of this work was to determine the effect of mixing the filter media of a compost biofilter on H2S removal efficiency. The behavior of important operational factors such as moisture of filter media, pressure drop and sulfate accumulation were evaluated, considering mixing the media. Additionally, tracer studies were performed in order to determine the effect of mixing the media on gas distribution. H2S removal capacity decreased over time, from 100% to 90%. When bed mixing was carried out, the removal capacity remained constant, close to 100%, and moisture content and sulfates accumulation were better controlled at 50% and at 12 mg S-SO4/g dry media respectively. In addition, under this operational pattern, an improvement in gas and particle size distribution was observed inside the filter media, fitting the axial dispersion model and the Ergun equation.  2005 Elsevier Ltd. All rights reserved. Keywords: Biofilter; Compost; H2S; Model; Odors; Packing particles 1. Introduction Compost biofiltration is one of the most important biological processes for waste gases treatment and for odor control (Van Groenestijn and Hesselink, 1993). This system is based on the interaction of gas phase pollutants with an organic packed media, such as compost. The degradation activity derives from microorganisms that live and develop in the filter media, in such a way that undesirable compounds in the gas are absorbed and removed. Three important general factors determine compost biofilter performance: (a) the type of the filter media (including void fraction, particle size, moisture content, * Corresponding author. Fax: +52 55 616 21 64. E-mail address: jmms@pumas.iingen.unam.mx (J.M. MorganSagastume). 0960-8524/$ - see front matter  2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2005.06.003 microbial diversity and nutrients), (b) the prevailing conditions of gas flow inside the biofiltration unit (including superficial velocity, gas distribution, temperature and inlet pressure) and (c) the substrate concentration, solubility and biodegradability. Research efforts are focused on the biofilter media in order to upgrade the performance of compost biofilters. Some use compost mixed with bulking agents in order to avoid high pressure drop, clogging and gas flow channels. Many materials have been used as bulking agents, such as activated carbon (Weber and Hartmans, 1995), polyurethane, polystyrene or glass particles (Zilli et al., 1996) as well as crushed oyster shells (Ergas et al., 1995). Other research efforts have been made on fluid distribution to overcome mass transfer problems associated with channeling and to increase substrate–microorganisms interaction using alternating flow direction (Ergas et al., 1994) or performing recycling streams (Ritchie and Hill, 1995). Special biofilter designs also have been developed such as the biorotor reactor (Buisman et al., 1990) and a J.M. Morgan-Sagastume, A. Noyola / Bioresource Technology 97 (2006) 1546–1553 1547 Nomenclature A C ti di dp Dg L cross-sectional area of the biofilter (m2) tracer concentration at a specified time i and Dti is the difference between ti and ti1 (mg/l) average mesh size of the wire sieves where a mass fraction xi (dimensionless) was caught (m) particle diameter (m) molecular diffusivity in gas phase (m2/s) biofilter height (m) modified biofilter with horizontal gas flow and baffles (Lee et al., 2001) to increase back-mixing. Additionally, bed mixing has been mentioned in literature as an important method to increase efficiency (Van Lith et al., 1997) but there are very few studies (Wubker et al., 1997) that have systematically explored this possibility. This work compares the performance of a benchscale compost biofilter for H2S elimination under two regimes of operation: (a) conventional and (b) with bed mixing. In the conventional operation, the polluted gas passes through a static packed media (without mixing), with water addition at the top of the biofilter for moisture control. In order to assess both operational regimes, some important parameters were monitored, such as moisture content, pressure drop, pH, sulfate accumulation and physical changes of the filter media (particle size, sphericity, void fraction, dead zones and gas distribution). 2. Methods gc U0 DP Q tr e q l r2h 1 kg m/s2 N air speed (m/s) pressure drop (cm of H2O) gas flow rate (l/min) average gas retention time (s) void fraction (dimensionless) air density, 0.927 kg/m3 air viscosity 1.83 · 105 kg/m/s normalized variance (dimensionless) 2.3. Biofiltration columns The biofiltration columns were built using PVC cylinders 0.10 m diameter and 1.2 m height (volume = 9.6 l). These columns were packed with compost to a height of 1.0 m. The filter media was retained in each column using a fine screening mesh. Each column had five gas and compost sampling ports spaced 20 cm along the column. 2.4. Packing procedure and bed mixing The columns were packed manually following the same procedure on each experimental run. The compost was taken using a spatula (approximately 300 g wet basis) and it was dropped freely into the column until obtaining a height of 1 m. Additional compaction of the media was avoided in order to allow only the natural compaction expected by the weight of the compost. Compost mixing was accomplished each 2 days by removing the entire bed from the column; manually homogenizing the media and then returning it into the biofilter column. 2.1. Filter media 2.5. Water addition The media used as biofilter packing was mature compost produced from food, and yard waste as well as horse manure. The compost was provided by the National University Compost Plant and was prepared in outdoor windows. The compost had a carbon/nitrogen ratio of 20:1, a moisture content of 65%, a pH of 7.48, an alkalinity of 357 mg CaCO3/L, a real and apparent density of 1.1 and 0.59 g/ml, respectively and a void fraction of 46%. 2.2. Air humidifying columns Two towers for air humidification were constructed. Both humidifiers were built using PVC cylinders of 0.15 m diameter and 1.2 m height. These towers were operated flooded, packed with Rashig rings (1/200 diameter) up to 0.90 m height. Water addition rate was based on a recommended water–air ratio between 1.5 and 3 ml water/m3 of gas (Leson and Winer, 1991). This resulted in a water addition rate of 57 ml tap water every 48 h considering a rate of 2 ml of water/m3. 2.6. Pressure drop A profile of pressure drop versus gas flow rate for the biofilters was fit to the Ergun equation in order to determine average particle diameter of the biofilter media as a function of height. Columns I, II and III were subjected to air flows from 10 to 70 l/min in 10 l/min increments to obtain plots of pressure drop versus gas flow rate. A water differential manometer was used for pressure drop measurements. The effect of the fine screen and other 1548 J.M. Morgan-Sagastume, A. Noyola / Bioresource Technology 97 (2006) 1546–1553 equipment at the bottom of each column was corrected by subtracting the pressure drop provided by those elements from each pressure drop measurement. 2.7. Physicochemical measurements The H2S concentration was measured along the length of the columns using electrochemical cells (SRII-U-100, BW Technologies). Sulfate concentration, moisture content and alkalinity were measured by removing small samples (approximately 1 g) of compost from each sample port. Sulfate (SO2 4 ) concentration in the media was measured by the photometric method using Merk Spectroquant equipment. Moisture content of the compost was determined gravimetrically (Parent and Caron, 1993). Alkalinity and pH of the biofilter media was measured using the method reported by Klute (1986). The granulometry of compost was determined using the sieve tray analysis method (Parent and Caron, 1993). Model 915B). A CO2 trap (KOH, 1 M) located between the gas sampler and the TOC analyzer was used to avoid interferences due to CO2 contained in the air. An automatic data recording system (Peaksimple II for SRI chromatographs) was connected to the TOC analyzer to reproduce the RTD curve on a screen and printer. Three tracer injections (1 ml each) per tracer study were performed and an average RTD curve was determined. Mathematical analysis was performed using an Excel spread-sheet program to determine average gas retention time. Each tracer study was carried out using airflow rate of 10 l/min. To assure minimum interaction between the tracer input and compost, the compost was saturated with butane prior to each tracer study. This was carried out using a constant butane input into the inlet air stream until a constant butane concentration of 1.5 mg/l was reached in the biofilter outlet. Mass balance calculations showed that this technique resulted in practically 100% recovery of the tracer. 2.9. Pilot plant 2.8. Tracer study Butane gas was used as tracer for determining the retention time distribution (RTD) curves because it has very low solubility, 1.26 mM at 298 K (Perry and Green, 1988) and can be easily measured by the monitoring system. The tracer was injected into the columns using a pulse injection technique (Levenspiel, 1972). A continuous sample was collected, using a gas pump, from the gas sampling ports of the biofilter to an infrared CO2 detector (Beckman Industrial TOC analyzer Compressed air (HAGEN-100 diaphragm compressor) was passed through two PVC humidification columns (Fig. 1). The humidification columns provided close to 100% relative humidity. A controlled flow of H2S from a gas cylinder was mixed with the main humidified air stream, which then was fed to the bottom of Column I resulting in a H2S concentration of 100 ppmv or 7 g H2S/m3/h. Air flow rate was maintained at 10 l/min, which provided a superficial loading rate of 74 m3/m2/d with an empty bed residence time (EBRT) Fig. 1. Schematic of experimental setup and tracer studies. 1549 J.M. Morgan-Sagastume, A. Noyola / Bioresource Technology 97 (2006) 1546–1553 of 50 s. The second and third columns (II and III) were used as controls; only a humidified air stream was fed to Column II and neither water or gas was fed to Column III. The biofilter columns were located on the roof of the Environmental Engineering Laboratory building at the barometric pressure of Mexico City (585 mm Hg) and at ambient temperature (20 ± 5 C). Columns I, II and III were analyzed at the end of the experiment. Columns I and II were operated for 206 days; the first 142 days using conventional operating criteria. Columns I and II continued their operation for an additional 65 days (from day 143 until day 206) with bed mixing every 2 days. Media moisture content was controlled using water addition at the top of the columns. Air supply for the columns was controlled using a Cole Parmer mass flow controller and calibrated rotameters. The columns were operated in upflow mode. 3. Results and discussion the other hand, for a mean input concentration of 100 ppm H2S (0.108 mg/l H2S or 0.101 mg S/l) the sulfur input to the media was 1.01 mg S/min. The H2S dissolved in the media can be calculated (0.39 mg H2S/l) using HenryÕs law (H = 483 atm at 20 C and 1 atm). To calculate the amounts of H2S and HS dissolved in water at a pH of 7.5, a species distribution graph reported by Sawyer et al. (1994) was used. It was possible to determine that total dissolved sulfur was 24% H2S and 76% HS, which means concentrations of 0.39 mg H2S/l and 1.19 mg HS/l (total sulfur of 1.52 mg S/l) in the compost media, supposing no biological reactions. Considering a water content of 3.12 l in the media and an homogenous water concentration of H2S, the sulfur content would be 4.74 mg of S in the compost. Under these conditions, total amount of sulfur at the gas inlet (1.01 mg S/min) and absorption capacity of the media (4.74 mg of S), the biofilter would be saturated in 4.7 min, neglecting adsorption. It is clear that, in the given conditions, removal of H2S can be explained by the activity of the microorganisms and not by the absorption only. 3.1. Start-up 3.2. H2S removal and sulfate control 120 110 100 90 80 70 60 50 40 30 20 10 0 H2S removal efficiency and compost media sulfate concentration for Column I are shown in Fig. 2. Initial H2S removal efficiency approached 100% and decreased to approximately 80% over 141 days of operation. However, after compost mixing was initiated, the H2S removal efficiency was maintained close to 100%. Sulfate concentration in the media stabilized at 21 mg SO2 4 -S=g dry compost. Sulfate accumulation in the media was due to H2S oxidation to sulfuric acid. As shown in Fig. 2, sulfate concentration was always lower than the inhibitory level of 25 mg SO2 4 -S=g of dry compost reported by Yang and Allen (1994). Routine compost washing and mixing could effectively mitigate sulfate accumulation. During the first 25 days of operation, the pH decreased from 7.5 to a stable value of 4.5. DegorceDumas et al. (1997) reported that this pH range is Conventional operation (without media mixing) Media mixing in column I 70 60 50 Column I (% removal) 40 Column I (Sulfate concentration) 30 20 mg Sulfate/g dry compost % removal During the start-up of the system, the biofilter (Column I) removed 63 ± 10% of the H2S in the first 40 cm of media This behavior shows the high H2S removal capacity that the biofiltration system had at the beginning of its operation. At the beginning of this period (0–25 days) the biofilter was capable of supporting high variations of H2S loading rates (±50%) produced accidentally due to control problems in the gas flow inlet system. Since the elimination capacity of the biofilter can be related to the water content of the biofilter, it was important to evaluate whether or not absorption was an important part of the H2S removal mechanism, especially at the beginning of operation when the media was fresh. The total weight of compost in the biofilter was 4.8 kg (calculated using a real density of 0.59 g of compost/ml) with a moisture content of 65%. This means a volume of 3.12 l of water is present in the biofilter. On 10 1 7 15 21 26 49 63 117 Time, d 148 162 175 191 0 206 Fig. 2. H2S removal efficiency and sulfate accumulation (mg SO2 4 -S=g dry compost) as a function of time with and without bed mixing. 1550 J.M. Morgan-Sagastume, A. Noyola / Bioresource Technology 97 (2006) 1546–1553 3 3 2 H2S Elimination capacity, % Biofilter I (4 gH2S/m /h: 74 m /m /h) Table 1 Experimental values for moisture content 100 90 80 Distance from gas inlet Compost moisture, % Initial compost sample – 65 Conventional operation over 141 days 20 40 60 80 100 31 35 40 51 71 Operation with media mixing over 65 days 20 40 60 80 100 51 50 51 50 51 Compost column with no flow 20 40 60 80 100 51 52 51 50 50 70 60 50 Standard deviation 40 Average 30 Standard deviation 20 10 0 0 5 10 15 20 25 30 35 40 45 50 55 Empty Bed Retention Time (s) Column I (Biofilter) Fig. 3. H2S removal capacity as a function of EBRT. adequate for high H2S removal efficiencies. Similar observations were made by Wada et al. (1986), who reported H2S removal close to 100% using a peat biofilter at a controlled pH of 3.5. Column III (Control) 3.3. Empty bed retention time and removal efficiency The biofilter had an empty bed retention time (EBRT) of 48.6 s. Each section of the biofilter had and EBRT of 9.7 s which in accumulative terms means 9.7, 19.4, 29.1, 38.8 and 48.6 s for sampling points 1–5, respectively. The average of H2S elimination capacity and standard deviation as a function of EBRT corresponding to the conventional operation is shown in Fig. 3. As expected, H2S elimination capacity increased as EBRT had a greater value (practically 100% for 40 s). The behavior of the standard deviation for higher EBRT confirms the stability in the operation of the biofilter. 3.4. Pressure drop, moisture content and particle size When comparing Columns I and II with III, a pressure drop increase may be appreciated from 4.5 to 9.3 cm of H2O when water was added at the start-up of Columns I and II. Excessive water content will lead to media compaction (Deshusses et al., 1995) and gas clogging that eventually will result in flow channels due to granulometry change. A very common practice to control the moisture of filter media following conventional criteria is adding water directly at the top of filter. However, high water loads in the upper part of the biofilter will flood the media, resulting in high pressure drops (Dullien, 1992) as water cannot be transported fast enough to the lower parts of the filter media. This behavior was noticed in Column I under conventional operation, as reported in Table 1. The homogeneity of the filter media was obtained when bed mixing was carried out every 2 days. Under this condition, the profile of pressure drop neither depends on moisture difference nor on change of the filter media granulometry. Under this type of operation, a proportional relation between filter height and pressure Note: The data for control II are practically identical to biofilter I. drop was obtained. For each 20 cm of biofilter height, it was determined an average pressure drop of 0.87 ± 0.11 cm of H2O. The Ergun equation (Ergun, 1952) correlates the pressure drop with the porosity and the filter media height. This equation was used to theoretically fit a representative experimental profile of pressure drop versus gas flow for Columns I, II and III in order to determine the particle diameters of the filter media as a function of its height. The Ergun equation is 2 DP 150ð1  eÞ lU 0 L 1.75ð1  eÞU 20 L ¼ þ ; gc e3 d p q gc e3 d 2p q ð1Þ where dp = particle diameter, m; gc = 1 kg m/s2 N; L = biofilter height, m; U0 = air speed, m/s; DP = pressure drop, cm of H2O; q = air density, 0.927 kg/m3; e = void fraction and l = air viscosity 1.83 · 105 kg/m/s. It is possible to observe in Fig. 4 that the homogenization of filter media produces a better particle size distribution throughout the biofilters. Without bed mixing, the largest particles are located at the top of the columns. This is due to the addition of water directly at the top of the columns as it was mentioned before. In contrast, dryer media could be found at the lower part of the columns, where particle sizes were smaller than those located at the top. 3.5. Gas distribution and dead zones The gas distribution in the biofilter media was quantified using tracer studies. The RTD curves showed plug flow behavior, which was analyzed using the axial dispersion model (Levenspiel, 1972): 1551 Particle size, mm J.M. Morgan-Sagastume, A. Noyola / Bioresource Technology 97 (2006) 1546–1553 2 With media mixing Conventional operation 1.5 1 Biofilter I (H2S: 50 ppm) 0.5 Biofilter II (control) Biofilter III (control) 0 20 40 60 80 100 20 Biofilter Height, cm 40 60 80 100 Fig. 4. Particle size as a function of biofilter height calculated using the Ergun equation.  !   2 1  ðD =U D D g g LÞ g 0 r2h ¼ 2 1e ; 2 U 0L U 0L  ð2Þ where Dg is the molecular diffusivity in the gas phase (m2/s) and r2h is the normalized variance (dimensionless). Retention time distribution curve for a step tracer input was determined with the following equation (Levenspiel, 1972): P ti C t Dti ð3Þ tr ¼ P i ; C ti Dti where tr is the average gas retention time (s); C ti is the tracer concentration (mg/l) at a specified time i and Dti is the between ti and ti1. Variance were calculated using Eq. (4) (Levenspiel, 1972): P 2 t C t Dti r2 ¼ P i i  t2r . ð4Þ C ti Dti The normalized variance (r2h ) was determined dividing the variance (Eq. (4)) by the square of the average time, tr(r2h ¼ r2 =t2r ) (Levenspiel, 1972). The term Dg/U0L in Eq. (2) is known as the axial dispersion number and is the inverse of the Peclet number. It can be used to assess dispersion within the system. Complete mixing can be expected as the axial dispersion number increases (Levenspiel, 1972). In Columns I and II (Table 2) a decrease in the axial dispersion number was observed at the end of conventional operation with respect to Column III (control), which implies reduced gas mixing. This can be explained by channeling and by the formation of dead zones. The average residence time for Columns I and III were 17.6 s and 19.1 s, respectively, indicating a dead zone in Column I of approximately 8%. The axial dispersion num- ber at the end of conventional operation in control III remained constant and close to the initial value. As it is evident in Table 2, when the compost bed was mixed, the gas distribution remained constant avoiding channel formation. Pressure drop as well as gas mixing and media moisture can be feasibly controlled when bed mixing is achieved. 3.6. Particle size, sphericity and void fraction Levenspiel (1998) reported an experimental method to estimate the sphericity, / using granulometry studies. The equation that relates / to the average particle diameter, dp (m) and the average wire sieve diameter, ds (m) is d p ffi /d s ; ð5Þ where ds is determined from the granulometry data using the following: " m X xi ds ¼ di i #1 ð6Þ ; where di (m) is the average mesh size of the wire sieves that retained a mass fraction xi (dimensionless). The average particle diameter, dp, can be estimated using pressure drop profiles and the Ergun equation as has been presented before. The void fraction, e, in a porous media can be estimated using tracer studies. Considering the definition of retention time, the void fraction can be calculated using the average time obtained from Eq. (3) as follows: e¼ tr Q ; AL ð7Þ Table 2 Tracer studies results in the biofilter I and controls II and III Biofilter (air + H2S) Control II (air) Control III (with no flow) Superficial loading rate, m3/m2/h Mass loading rate, g H2S/m3/h Conventional operation over 141 days Operation with media mixing over 65 days Dead zone, % Axial dispersion number, Dg/U0L · 103 (dimensionless) Dead zone, % Axial dispersion number, Dg/U0L · 103 (dimensionless) 74 74 0 7 0 0 7.8 7.8 0.1 8.5 8.4 10.7 0.1 0.1 0.1 10.8 10.7 10.7 1552 J.M. Morgan-Sagastume, A. Noyola / Bioresource Technology 97 (2006) 1546–1553 1 Coupling of particles; Brown, (1950) and Levenspiel, (1998) Sphericity, 0.95 0.9 Low Coupling (Cubic arrangement) High coupling (Rhombohedric arrangement) Sections 4 and 5 Column II (control at the end of conventional operation) 0.85 Column III (control) 0.8 Column I (Biofilter under conventional operation) 0.75 Column I (Biofilter with media mixing) Sections 1, 2 and 3 0.7 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 Void fraction, Fig. 5. Sphericity versus void fraction for compost particles compared with the extreme arrangements for incompressible particles. where Q is the gas flow rate (l/min) and A is the crosssectional area of the biofilter (m2). This method has the advantage over the water displacement technique in that it allows the estimation of the void fraction without affecting the physical structure of the bed. The sphericity of the particles remained nearly constant (approximately 0.82 ± 0.02) throughout the experiment in all of the columns. Neither particle agglutination nor particle disintegration changed the geometric form of the particles. The void fraction, e, and sphericity, /, of extreme arrays of incompressible spherical particles are plotted in Fig. 5. The tightest array is a rhombohedric arrangement and the looser one a cubic arrangement (Brown, 1950; Levenspiel, 1998). The points concerning the mixed compost as well as Sections 1–3 of the columns are located near the line of tightest arrangement. This means that the compost particles at the bottom of the biofilter are highly compacted having a rhombohedric arrangement, if compared with the particles situated at the top of the biofilter, which approach a cubic arrangement. This may be explained by dry conditions at the lower part of the biofilter (Table 1) that decreased the particle size. In addition, the bottom particles suffer higher compaction forces as they support the weight of the media. Also, there are some points (corresponding to Sections 1–3 of the columns) located at the left of the tightest array line that ideally should not be there since that line represents the maximum density arrangement possible for incompressible particles. This fact could be explained assuming a compression capacity of the compost particles, like sponge fuzzy balls, which could change their form under pressure. Under these conditions, a decrease of permeability can be expected in the lower sections of the biofilter, increasing the pressure drop. In fact, the average pressure drop in the first section of Column I was 46% of the total pressure drop obtained for conventional operation. Similar results have been obtained and may be explained by the well-known capillary phenomena in soils (Cernica, 1995; Gostomski and Liaw, 2001). Capillary forces filled the pores of compost particles up to the height to the average pore diameter. Thus, this region did not drain and subsequently could have low permeability. The hypothesis about particle compression mentioned before together with the capillary phenomena could explain the changes of media permeability over time, especially in the lower sections. 4. Conclusions In order to avoid flow channel formation, as well as to achieve constant media moisture, granulometry, void fraction, and in general, the homogeneity conditions of the original compost media, a bed mixing operation is recommended. Bed mixing also provides feasible conditions for the control of sulfates accumulation by means of a better media washout when H2S is treated by compost biofiltration. However, the investment and operational costs required to carry out bed mixing in full-scale biofilters, must be taken into account and evaluated before its application. The internal changes that the biofilters had over time were closely related to moisture conditions inside of the biofilter. The conventionally operated biofilters had significant internal changes, especially in the lower part of the column, where the gas inlet was located, due to compost drying. A hypothesis about compost particles compression is proposed in order to explain the deviation of some points from the tightest array plot in a sphericity versus void fraction graph. Acknowledgements The authors would like to acknowledge the financial support from the Mexican National Council for Science and Technology (CONACyT, Proy. 27776-B). J.M. Morgan-Sagastume, A. Noyola / Bioresource Technology 97 (2006) 1546–1553 References Buisman, C., Wit, B., Lettinga, G., 1990. Biotechnological sulphide removal in three polyurethane carrier reactors: stirred reactor biorotor reactor and upflow reactor. Water Research 24, 245–251. Brown, G., 1950. Unit Operations. John Wiley and Sons, New York. Cernica, J., 1995. Geotechnical Engineering. John Wiley and Sons, New York. Degorce-Dumas, J., Kowal, S., Le Cloirec, 1997. Microbial oxidation of hydrogen sulphide in a biofilter. Canadian Journal of Microbiology 43, 264–271. Deshusses, M., Hamer, G., Dunn, I., 1995. Behavior of biofilters for waste air biotreatment. 1. Dynamic model development. Environment Science Technology 29, 1048–1058. Dullien, F., 1992. Porous Media: Fluid Transport and Pore Structure. Academic, San Diego. Ergas, S., Kinney, K., Fuller, M., Scow, K., 1994. Characterization of a compost biofiltration system degrading dichloromethane. Biotechnology and Bioengineering 44, 1048–1054. Ergas, S., Shroeder, E., Chang, D., Mortan, R., 1995. Control of volatile organic compound emissions using a compost biofilter. Water Environment Resource 67, 816–821. Ergun, S., 1952. Fluid flow through packed columns. Chemical Engineering Progress 48, 89–94. Gostomski, P., Liaw, L., 2001. Air permeability of biofilter media. In: Proceedings of the 94th Annual Meeting of the Air and Waste Management Association, Orlando, FL, USA, June 24–28, in CD. Klute, A., 1986. Methods for Soil Analysis. American Society of Agronomy, Soil Science Society of America, Madison, WI, 122.6.5. Lee, D., Lau, A., Pinder, K., 2001. Development and performance of an alternative biofilter system. Journal of the Air and Waste Management Association 51, 78–85. Leson, G., Winer, A., 1991. Biofiltration: An innovative air pollution control technology for VOC emissions. Journal of the Air and Waste Management Association 41, 1045–1054. 1553 Levenspiel, O., 1972. Chemical Reaction Engineering. John Wiley and Sons, New York. Levenspiel, O., 1998. Engineering Flow and Heat Exchange. Plenum Press, New York and London. Parent, L., Caron, J., 1993. Physical properties of organic soils. In: Carter, M. (Ed.), Soil Sampling and Methods of Analysis. Canadian Society of Soil Science. Lewis Publishers, pp. 441–458. Perry, R., Green, D., 1988. PerryÕs Chemical Engineers Hand Book. McGraw-Hill International, Singapore. Ritchie, B., Hill, G., 1995. Biodegradation of phenol polluted air using an external loop airlift bioreactor. Chemical Technology and Biotechnology 62, 339–344. Sawyer, C., McCarty, P., Parkin, G., 1994. Chemistry for Environmental Engineering, fourth ed. McGraw Hill. Van Groenestijn, J., Hesselink, P., 1993. Biotechniques for air pollution control. Biodegradation 4, 283–301. Van Lith, C., Leson, G., Michelsen, R., 1997. Evaluating design options for biofilters. Journal of Air and Waste Management Association 47, 37–48. Wada, A., Shoda, M., Kubota, H., Kobayashi, T., KatayamaFujimura, Y., Kuraishi, H., 1986. Characteristics of H2S oxidizing bacteria inhibiting a peat biofilter. Journal of Fermentation Technology 64, 161–167. Weber, F., Hartmans, S., 1995. Use of activated carbon as a buffer in biofiltration of waste gases with fluctuating concentrations of toluene. Applied Microbiology and Biotechnology 43, 365–369. Wubker, S., Laurenzis, A., Werner, U., Friedrich, C., 1997. Controlled biomass formation and kinetics of toluene degradation in a bioscrubber and in a reactor with periodically moved trickle-bed. Biotechnology and Bioengineering 55, 686–692. Yang, Y., Allen, E., 1994. Biofiltration control of hydrogen sulfide I. Design and operational parameters. Journal of Air and Waste Management Association 44, 863–868. Zilli, M., Fabiano, B., Ferraiolo, A., Converti, A., 1996. Macrokinetic investigation on phenol uptake from air by biofiltration: influence of superficial gas flow rate and inlet pollutant concentration. Biotechnology and Bioengineering 49, 391–398.