RCSWMA has focused on odor control provisions for the facility, which is approaching the end of its first year of operation without a single verified off-site odor complaint.
As part of the 12-week facility acceptance test, performance testing of the odor control system included odor concentration, ammonia, reduced sulfur compounds and VOCs. The odor control system has met acceptance criteria, but expectations for hydrogen sulfide and methyl mercaptan removal have been revised based on a new understanding of the overall degradation dynamics for reduced sulfur compounds.
The facility is adjacent to a closed landfill and borders state park land. The nearest residential development is approximately 1,000 feet away. RCSWMA recognized that the site buffers provided the opportunity for success, but only if a highly reliable odor control system were installed. New York state regulations require that the facility not cause objectionable odor impacts, but do not define an odor concentration that is considered objectionable. As part of the DBO contractor selection process, RCSWMA requested guaranteed performance for the odor control system with a goal of having no off-site odor impacts. The required removal rates could not be guaranteed with conventional open biofilter systems. Consequently, proposals for proprietary biofilter systems were evaluated where the required performance could be guaranteed.
RCSWMA carefully evaluated the costs versus benefits of several systems. It appeared that the life cycle cost of an open biofilter could approach that of proprietary systems if frequent media replacement (e.g. annual) were necessary to maintain high performance levels. After reviewing the cost and dispersion modeling results for a variety of sizing and guaranteed removal rate options, RCSWMA ultimately selected an odor control system supplied by Envirogen (formally CVT America, a division of Envirogen) with a guaranteed odor removal rate of 94 percent. The Envirogen package cost $1,670,000 and included supply and construction/installation of the exhaust fans, dual pretreatment scrubbers with chemical feed system, enclosed biofilter, and discharge stack (exclusive of the foundations). The design data for the odor control system are summarized in Table 1.
In addition to odor concentration, removal rate guarantees were provided for ammonia, hydrogen sulfide, and methyl mercaptan. For most biofilters, the maximum safe ammonia loading rate is 180 grams/cubic meter of media/day (typically in the range of 35 ppm in the inlet air), although some biofilters can operate successfully at up to 250 grams/cubic meter/day. Since the inlet ammonia levels could not be guaranteed to be within safe loading rate limits, RCSWMA included packed bed scrubbers for ammonia removal prior to the biofilter. The removal rate guarantee for ammonia was not driven directly by air quality concerns, but reflected reasonable performance expectations for the equipment to be supplied.
Similarly, the removal rate requirements for hydrogen sulfide and methyl mercaptan do not reflect any specific air quality concern with the compounds. The only air emission concern was the potential for objectionable odor impacts, which is best measured by odor concentration (dilutions-to-threshold, D/T). Hydrogen sulfide and methyl mercaptan were selected primarily because removal of these compounds has commonly been used as an indicator of odor control performance in wastewater treatment applications. As we shall see, the guaranteed removal rates for the two compounds proved problematic, but ultimately this choice has led to a better understanding of the degradation dynamics of reduced sulfur compounds.
SYSTEM DESCRIPTION, OPERATING PHILOSOPHY
The enclosed portion of the cocomposting facility includes biosolids receiving and feed mix preparation; composting via an agitated bin system supplied by U.S. Filter; and aerated static pile curing. A mixture of raw primary and secondary and anaerobically digested biosolids are composted with ground wood residuals. Ventilative air is fed into the building via makeup air units at either end of the building. The exhaust air (from the positively aerated bins) is collected in two ducts located in the center of the building above the agitated bin composting area. These ducts contain instrumentation to measure air flow, pressure and temperature.
The makeup air units and exhaust fans have variable frequency drives to control their speeds in order to maintain preset air flow rates. The design ventilation rate is a total of 82,000 cfm or 41,000 cfm/exhaust fan. The fans have the capacity to draw up to 50,000 cfm when only one unit is operating. The exhaust rate is set to exceed the makeup air rate to ensure that the building is maintained under negative pressure. The minimum temperature requirements for the biofilter are maintained by modulating heat from the makeup air units based on the temperature in the main exhaust ducts.
The ventilation rate is reduced from the maximum 82,000 cfm rate during occupied periods to approximately one-half during unoccupied periods. This allows the retention time in the biofilter to be increased, which was expected to improve odor removal during unoccupied periods. It also results in less air being discharged during the times when odor impacts are most likely during the early evening and early morning. Electric power requirements also are reduced at the lower flow rate.
Pretreatment scrubbers were provided for humidification and control of the ammonia and particulate loading to the biofilter. The pretreatment scrubbers are vertical packed bed towers with recirculation pumps. Sulfuric acid addition is used to minimize the blowdown rate (discharge to sewer) needed to maintain high ammonia removal rates. The need for ammonia removal in the scrubber is greater during cold weather when the lower temperatures slow down the biological activity. The removal rate can be varied by adjusting the pH set point. At lower pH set points, ammonia removal in the scrubber will be greater, but acid consumption will be higher.
The biofilter consists of three major sections: 1) a concrete inlet distribution tunnel; 2) the biofilter media section (made of organic materials and referred to as biomass); and 3) a concrete outlet collection tunnel and exhaust stack. The biofilter media section is divided into four independent cells or “chambers” that can be isolated via slide dampers (in the inlet/outlet tunnels) for inspection or maintenance. Each chamber contains a five-foot deep layer of the media and a 0.3-foot depth of a base layer of inert material. Air flows upward through the media from a distribution chamber below and out under the covers to the outlet collection tunnel. Treated air is then discharged through the 60-foot- high exhaust stack. Typical retention time in the biofilter is 20 seconds at 82,000 cfm design capacity (compared to conventional open biofilters with typical retention times of 60 seconds in this type of application).
Each chamber is provided with a dual irrigation system — one at the top of the media to maintain the moisture level, and the other between the media and the base layer. The lower layer ensures complete humidification of the air entering the biologically active area of the biofilter. A computer regulates the timing sequence of the irrigation systems. The watering equipment includes a nutrient feed system to periodically replenish the biomass.
Construction of the cocomposting facility was completed in January, 1999. A five-week shakedown period was followed by a 12-week acceptance test period. During the acceptance test period, three rounds of odor testing were scheduled to document compliance with the removal rate requirements as shown in Table 1. Based on results of the first test, however, a fourth was added.
The facility operated with all nine composting bins fully loaded at the design capacity during acceptance testing. During each testing event, samples were taken at the peak airflow rate of 82,000 cfm. All operations characteristic of peak emissions were carried out, including receiving of biosolids and processing into feed mix and working of the bins with the compost turners. The tests also included samples taken at conditions characteristic of unoccupied periods with the lower air flow rate of 50,000 cfm and no material handling activities in progress.
The odor system acceptance testing was carried out and/or coordinated by Odor Science & Engineering. Samples of inlet and outlet gases were collected and analyzed by standard sampling and lab procedures. Measurements of system inlet airflow, temperature, humidity and scrubber pH also were performed concurrently with sampling events.
Figure 1 shows the results of testing for odor concentration, ammonia, and reduced sulfur compounds at the 82,000 cfm peak emission rate. The removal rates are summarized in Figure 2 for both the 82,000 cfm peak emission rate and the 50,000 cfm average rate. Detailed reduced sulfur results are summarized in Table 2 for the testing at the 82,000 cfm peak emission rate.
Beyond the required acceptance testing outlined above, inlet VOCs levels were measured continuously during the 12-week period (Del Vecchio, 1999). Near the end of the acceptance test period, some outlet measurements also were taken.
PROBLEMS DURING INITIAL TEST
Results from the initial test on April 8 indicated that removal rates were below the required levels for all parameters as shown in Figure 2 and Table 2. An investigation was conducted to determine why the system had lower than expected performance. The inlet odor concentration was higher than had been expected, so the media was tested to ensure that nutrient levels were adequate for the overall contaminant loading. This testing showed a complete lack of phosphorus in the media. The reduced sulfur compound results were particularly enigmatic, since the outlet concentrations of methyl mercaptan were higher than the inlet. Unfortunately, the testing had been carried out only for the specific compounds. The laboratory was subsequently able to determine results for additional reduced sulfur compounds by checking the gas chromatograph output, but the testing had been curtailed before the full scan was completed.
Prior to the next test, the following plan was put into action: Nutrient addition was initiated to the media using monosodium phosphate; A fourth round of performance testing was added to the original three tests; and Complete reduced sulfur analyses would be conducted on future tests.
A subsequent media test was conducted following nutrient addition and showed a residual of phosphorus. The additional performance test then was scheduled to ascertain that factors limiting removal performance had been remedied. As shown in Figures 1 and 2, the odor removal rate met the required level of 94 percent during the second test (April 27); the ammonia removal also was within the required efficiency of 99 percent. However, both methyl mercaptan and hydrogen sulfide at times still exhibited the anomalies of higher outlet concentration than inlet. This was attributed to interactions with other compounds, particularly dimethyl disulfide (explained in more detail below). The remaining two rounds of testing (May 11 and June 1) showed similar, yet improving results as compared to the April 27 testing.
ODOR CONTROL SYSTEM PERFORMANCE
As shown in Figure 2, the odor removal efficiency was less than expected during the first test, but increased to greater than the required 94 percent for the final three tests. The inlet odor concentration closely followed the total reduced sulfur concentration and the VOC concentration. Ammonia, especially as seen during the last test, had much less of an impact on the actual inlet odor concentrations. Nevertheless, in terms of mass loading, ammonia is by far the most prevalent compound in the exhaust air.
All of the tests included a fence line and neighborhood odor survey. During these surveys, no odor from the cocomposting facility was detected at or beyond the property line. The use of a discharge stack appears to have been effective in dispersing residual odor in the outlet gas, thus reducing odor impacts in this application. A ground level area discharge would have almost certainly resulted in some off-site impacts in this situation.
During the first test on April 8, ammonia removal did not meet the required efficiency, primarily due to lack of removal in the biofilters. The scrubbers were operating at a pH of less than four, which resulted in a residual of one to 1.5 ppmv of ammonia entering the biofilter. This residual ammonia was not being removed in the biofilter, apparently due to the phosphorus deficiency noted above. As system performance improved following nutrient addition, the pH in the scrubbers was raised, thereby increasing the load that had to be removed by the biofilter. By the last test, the pH had been raised to 6.25, allowing 15 ppmv of ammonia into the biofilter while maintaining nondetectable levels at the outlet.
Reduced Sulfur Compounds
As noted above, results of the initial test for methyl mercaptan, and subsequent tests for hydrogen sulfide, not only did not show the required removal rates, but actually indicated production of these compounds. Table 2 shows the detailed breakdown of individual compound concentrations for the tests. The hydrogen sulfide and methyl mercaptan levels were increasing at times, even though overall removal rate for total reduced sulfur compounds was high. Envirogen completed a literature review that clarified the apparent mechanisms. Methyl mercaptan (MM) and hydrogen sulfide (H2S) are intermediates in the oxidative pathways for dimethyl disulfide (DMDS) and dimethyl sulfide (DMS), as shown in Figure 3 (Smith and Kelly, 1988a; Smith and Kelly, 1988b; Kanagawa and Kelly, 1986; De Bont et al., 1981; Suylen et al., 1986).
Hydrogen sulfide also can be formed as an intermediate product in the degradation of carbon disulfide (CS2) and carbonyl sulfide (COS) (Smith and Kelly, 1988c). However, as seen in Table 2, degradation of both carbon disulfide and carbonyl sulfide was negligible. Although the quantities of MM and H2S in the outlet air were relatively low, at least in some cases they were greater than the quantities in the inlet air to the biofilter, apparently due to production from the degradation of DMDS, DMS and/or MM. The removal rate requirements for these compounds were based on data from non-composting applications. In this application, there is no reasonable way to measure removal efficiency of these two compounds individually.
RCSWMA has concurred that the best approach to tracking sulfide removal is to measure the reduction in total reduced sulfur compounds, which encompasses all of the important target sulfur compounds and overcomes the issue of intermediate production of MM or H2S from DMDS, DMS, and/or MM. As previously noted, the emissions of reduced sulfur compounds were not directly an air quality concern, but merely an indicator that the odor control system is operating properly.
The reduced sulfur compounds showed most clearly the expected increase in removal rate when the retention time through the beds is increased as displayed in Figure 2. The removal rate increased from an overall average of 85 percent reduction at 82,000 cfm to 92 percent removal at 50,000 cfm. This corresponded to an increase in empty bed retention time from 20 seconds at 82,000 cfm to 32 seconds at 50,000 cfm.
Volatile Organic Compounds
VOC concentrations in the inlet averaged in the 20-ppmv range with peaks exceeding 200 ppmv as propane. The concentration of reduced sulfur compounds present had a large effect on the total VOC concentration. A test to determine what compounds made up these VOCs found the following present in addition to the reduced sulfur compounds noted in Table 2: Methane; Acetone; Formaldehyde; Acetaldehyde; Methyl Ethyl Ketone; Isopentanal; Ethanol; N,N-Dimethyl Methenamine; (-Pinene); 4-Methylene-1-(1-Methylethyl Cyclohexane); and Dimethylamine.
Testing was carried out at the end of the acceptance test period to determine the reduction efficiency of the system on VOCs. Based on the data collected, VOCs were reduced from an average of 15 ppmv in the inlet to less than 0.5 ppmv in the outlet, or a removal rate of greater than 95 percent.
The proprietary biofilter has performed well overall and met acceptance criteria. However, there were some interesting findings from the system shakedown and performance testing.
• The composting exhaust air does not provide adequate phosphorus to maintain removal performance. The media will require the addition of phosphorus, approximately every three to six months. At the Rockland County facility, this need has been met through the addition of monosodium phosphate.
• Removal of individual reduced sulfur compounds is complicated by the fact that some compounds, including methyl mercaptan and hydrogen sulfide, are formed as intermediates in the oxidative pathways of more complex compounds such as dimethyl disulfide and dimethyl sulfide. Consequently, the best available method is to track the reduction of total reduced sulfur compounds.
John Goodwin is with Wheelabrator-BioGro, a division of Waste Management, and the facility operator. Waste Management of New York was the selected design-build-operate entity. Sebastian Amenta is with the Maguire Group, the lead engineer for the cocomposting facility. Ronald Delo is with the Rockland County Solid Waste Management Authority. Michael Del Vecchio is with Envirogen, supplier of the odor control system. Jeffrey Pinnette is with Wright-Pierce, the subcontractor (to Maguire Group) responsible for process design. Theodore Pytlar is with Dvirka & Bartilucci, the Authority’s engineer from the inception of the project. The facility was constructed by Helmer-Cronin