Deshusses, Johnson, and Leson

To date, biofilters have been used primarily to control dilute, usually odorous, off-gases with relatively low volatile organic compound (VOC) concentrations (<1 g m-3) and VOC loads (<50 g m-3 hr-1). Recently, however, U.S. industry has shown an interest in applying biofilters to higher concentrations of VOCs and hazardous air pollutants (HAPs). In this study, the behavior of biofilters under high loads of binary VOC mixtures was studied. Two benchscale biofilters were operated using a commercially available medium and a mixture of wood chips and compost. Both were exposed to varying mixtures of ethyl acetate and toluene. Concentration profiles and the corresponding removal efficiencies as a function of VOC loading were determined through frequent grab-sampling and GC analysis. Biofilter response to two frequently encountered operating problems—media dry-out and operating temperatures exceeding 40 °C—was also evaluated under controlled conditions. Microbial populations were IMPLICATIONS To date, biofilters have been used primarily to control dilute, usually odorous, off-gases with relatively low VOC concentrations. Yet there is a growing interest in expanding to higher concentrations the range of suitable applications for biological waste air treatment. However, this type of application has a high rate of system failure. Impediments for this application were reported in two recent case studies in which the treatment of high loads of ethanol from a foundry and a bakery, respectively, resulted in reduced percentage removal of the contaminants, formation of odorous acetic acid, and problems maintaining the proper moisture content because of the exothermic nature of the biodegradation process. The present study reports bench-scale operation of biofilters under high VOC loading conditions of mixed pollutant and the factors that lead to these adverse operating states also monitored to confirm the presence of organisms capable of degrading both major off-gas constituents. The results demonstrated several characteristics of biofilters operating under high VOC load conditions. • Maximum elimination capacities for ethyl acetate were typically in the range of 200 g m-3 hr-1. • Despite the presence of toluene degraders, the removal of toluene was inhibited by high loads of ethyl acetate. • Several byproducts, particularly ethanol, were formed. • Short-term dry-out and temperature excursions resulted in reduced performance. INTRODUCTION It is generally thought that biofilters are suitable only for the treatment of dilute emissions of odors, volatile organic compounds (VOCs) and hazardous air pollutants (HAPs). The large majority of biofilters installed to date treat off-gases containing organic carbon compounds at concentrations of less than 0.5 g m-3. Corresponding VOC loads and elimination capacities rarely exceed 50 grams of organic carbon per cubic meter of biofilter material per hour (g m-3 hr). Recently, however, industrial users, particularly in the United States, and vendors of biofilters have been attracted by the concept of treating higher concentrations and loadings of VOCs in biofilters to expand their range of applicability. The economic competitiveness of thermal and catalytic incineration systems, frequently the main competing technologies for complete pollutant destruction, has been a major obstacle to the biofilter treatment of higher concentrations of VOCs. However, additional fuel is needed for thermal or catalytic incineration of air streams containing less than about 20 to 50 g VOC m-3, increasing both the operating costs and dependency on fossil fuels. Further, incineration generates nitrogen oxides, which require expensive off-gas treatment. Deshusses, Johnson, and Leson 974 Journal of the Air & Waste Management Association Volume 49 August 1999 On the other hand, biological air pollution control does not require additional fuel and, under optimum conditions, complete degradation to carbon dioxide is achieved without the formation of secondary pollutants. Unfortunately, several technical obstacles have been reported when biofilters have been applied to highly polluted air streams. They include an elimination capacity that is limited to less than 200 g m-3 hr-1 for most VOCs,1,2 the formation of potentially odorous or hazardous degradation byproducts,3-5 and the increasingly difficult task of maintaining the proper moisture content of the biofilter medium.6 Previous laboratory and field work had also indicated that biofilters may experience additional performance problems when treating mixtures of VOCs at high loads.7-10 The goal of this study was to investigate and quantify, under controlled conditions, several of the phenomena encountered in full-scale biofilters when treating high loads of two commonly used industrial solvents: ethyl acetate and toluene. Objectives for the study included establishing maximum removal rates for two commonly used biofilter media, assessing the inhibition of toluene removal in the presence of ethyl acetate, and evaluating the formation of byproducts. Simulations of two potentially catastrophic events—dry-out and temperature excursions in the media—were also conducted. MATERIALS AND METHODS Two bench-scale biofilter columns treated mixtures of ethyl acetate and toluene at a fixed mass ratio (3:1) and empty bed residence time (EBRT = 3 minutes) but varying concentrations of total VOC. Biofilter A was filled with an 80/ 20-by-volume mixture (medium A) of wood chips and compost, respectively. Biofilter B was filled with a commercially available medium (medium B) mixed from compost and polystyrene spheres (Bioton). Calcium carbonate (limestone) for pH buffering (10% based on dry weight) and essentially insoluble inorganic nutrients (0.4% N, 0.4% P2O5, 0.2% K2O based on dry weight) from bone and blood meal were added to the wood chips medium mixture. To accelerate the acclimation process, both biofilters were inoculated with mixed bacterial cultures grown on ethyl acetate and/or toluene. All experiments described herein were performed over a two-month period. The biofilters were made of clear PVC pipes, with an internal diameter of 15.2 cm. The biofilter bed was split into three 50-cm sections. The top segment was referred to as the first segment and the bottom segment was referred to as the third segment (see Figure 1). This segmented setup made it easier to collect gas samples at intermediary heights. Biofilters were operated in a downflow mode. Using a metering pump (FMI, Inc., NJ), ethyl acetate and toluene were injected at specified rates into humidified air to produce the synthetic waste air stream. The air flow rate was controlled (mass flow controller, Porter Instruments, PA) to keep a constant empty-bed residence time of 3 minutes; the total VOC concentration ranged from 4 to 12.4 g m-3. Gas samples were collected and automatically injected into a Hewlett Packard 5890 Series II gas chromatograph fitted with a 30-m Supelcowax 10 column (0.53 mm, 1 μm film, Supelco, Bellefonte, PA), equipped with an flame ionization detector (FID) for the detection of ethyl acetate, toluene, and possible metabolites. During steady-state conditions, about one month after start-up, the population of heterotrophs on the biofilter media was estimated by plating serial dilutions of medium aqueous suspension on plate count agar (Difco, Detroit, MI). The relative numbers of ethyl acetate or toluene degraders were tracked in a similar manner on basal salt medium plates with ethyl acetate or toluene supplemented via the gas phase. Colony-forming unit (CFU) counts were made after 24 to 48 hours incubation at room temperature on the plate count agar, and 2 to 5 days on basal salt medium/VOC plates. RESULTS AND DISCUSSION Inhibition of Toluene Removal Concentration profiles of ethyl acetate, toluene and the byproduct ethanol at VOC inlet concentrations of 6.9 and 12.4 g m-3 using medium A are shown in Figures 2 and 3, respectively. Under both VOC inlet conditions, high elimination capacities for ethyl acetate, up to 180 g m-3 hr -1, were noted in the first two segments. Little or no toluene removal occurred throughout the biofilter with the 12.4 g m-3 VOC inlet concentration, or in the first two segments at the 6.9 g m-3 VOC inlet concentration. Only in the third segment of the latter biofilter, where concentrations of ethyl acetate had fallen off to less than 0.5 g m-3, did toluene removal commence. In this latter segment, Figure 1. Schematic of the experimental setup. Deshusses, Johnson, and Leson Volume 49 August 1999 Journal of the Air & Waste Management Association 975 the observed elimination capacity for toluene was about 20 g m-3 hr -1, comparable to values reported for other, compost-based, biofilter systems.7 Consistent with previously reported data,11 the concentration profiles illustrate that concentrations of ethyl acetate in excess of 0.5 to 2 g m-3, apparently inhibit the concurrent removal of toluene. To assess whether this lack of toluene removal is due to the absence of toluene-degrading microorganisms, microorganisms extracted from the first segments of each column were plated onto purified agar plates and exposed to ethyl acetate and toluene vapors, respectively, as the sole source of organic carbon. Results are summarized in Table 1. Even in the non-toluene-degrading first segment of the biofilter, toluene-degrading microorganisms were present in sufficiently high quantities (108–109 CFU/gram of moist medium) to be able to remove the toluene transferred from the gas phase into the biofilm. Further testing indicated that all of the isolated strains of toluene degraders were also capable of degrading ethyl acetate if supplied as the sole carbon source. However, these facultative toluene degraders were outnumbered by the microorganisms capable of degrading only ethyl acetate. These microbiological results indicate that the lack of toluene removal in segments 1 and 2 was not due to the absence of suitable microorganisms but rather to some inhibitory mechanism preventing them from metabolizing toluene in the presence of high ethyl acetate concentrations. At this time, the exact mechanisms leading to the inhibition of the toluene metabolism remain unknown. Similar inhibitions have been reported for simultaneous treatment of MEK and MIBK in biofilters,12 and butanol and toluene in a biotrickling filter.10 However, inhibition was not expected a priori in this case, since ethyl acetate and toluene are known to be degraded by very different pathways. One potential reason for inhibition of toluene degradation is failure of the toluene degraders to compete for oxygen, which limits total VOC removal at high concentrations.7 It is also conceivable that other mechanisms exist by which the presence of ethyl acetate or the accumulation of degradation byproducts may inhibit the expression of the inducible enzymes involved in the initial breakdown of toluene.13 Further research is required to determine the actual mechanisms of inhibition in this case and whether they can be controlled. Due to the minimal removal of toluene in the concentration range tested in our experiments, the following discussion will focus on the removal of ethyl acetate in both biofilters. Elimination Capacity Bulk ethyl acetate elimination capacities by mediums A and B were measured at various times during the project as a function of ethyl acetate influent concentration and load. Bulk loads, L, and elimination capacities, EC, for individual segments and for the entire reactor, singly, were calculated using the empty bed residence time in minutes (EBRT) as follows: Figure 2. Typical concentration profiles in the biofilter reactor. Here, the medium A biofilter is shown with a total inlet VOC concentration of 6.9 g m-3. Table 1. Log of the total number of heterotrophic and toluene-degrading microorganisms per gram of damp medium. Samples were taken at the top of the first segment. Total Count Toluene Degraders Log N Log N (% of total) Wood chips biofilter 9.36 8.45 (12.3%) Bioton biofilter 9.07 8.89 (66.0%) Figure 3. Concentration profiles in the medium A biofilter. Total inlet VOC concentration was 12.4 g m-3. Deshusses, Johnson, and Leson 976 Journal of the Air & Waste Management Association Volume 49 August 1999 L c