1. UNIVERSITY OF SOUTH FLORIDA COLLEGE OF PUBLIC HEALTH
Figure 10. ISC Modeled NH3+ Concentration Gradients
ABSTRACT
The Pasco County, Florida, Shady Hills Sludge Processing Facility [Fig. 1] converts waste sludge into fertilizer. The treatment process results in a significant release of ammonia to the atmosphere. The purposes of this study were to measure the ambient air concentrations of the ammonia plume emanating from the facility, estimate the source emission rate from the observed concentration gradient, and refine ambient air ammonia sampling techniques for future sampling in the downtown Tampa Bay Area. The study had three steps. The first step was the determination of ammonia concentrations at varying distances from the facility. This step involved collecting samples from passive sampling devices (PSDs) strategically positioned to represent the ammonia dispersion from the facility’s downwind side. The second step was to sample the source strength at its generation point using a wet rotating annular denuder (WRAD) and collocated PSDs. The third step was to input source, facility, and meteorological data into an air pollution modeling program with an initial guess of source strength. The model was run with successive guesses until predicted ammonia concentrations reasonably agreed with measured concentrations.
ESTIMATION OF THE AMMONIA EMISSION RATE FROM A SLUDGE PROCESSING FACILITY IN PASCO COUNTY, FLORIDA
S. Mower, D. Anderson, J. Taylor, C. Mulhern, S. Berge, A. Garcia, and N. Poor
UNIVERSITY OF SOUTH FLORIDA COLLEGE OF PUBLIC HEALTH
Figure 11. Measured NH3+ Concentration Gradients
WRAD RESULTS
The ammonia concentration measured from WRAD sample differed by 33% from the collocated PSD sample results. The PSDs measured an ammonia concentration of 6,100 g m-3, whereas the WRAD measured 4,100 g m-3. This was not unexpected since the 0.85% phosphoric acid coating solution completely evaporated between the first and second hour of the two hour WRAD sample collection period. This evaporation was the result of the incoming air flow vaporizing the liquid coating solution. Without the coating solution no more ammonia gas could be captured. The sample was extracted by adding more solution to the denuder and rotating to recover any phosphoric acid/ammonia that was still adhering to the sides of the annular space.
SAMPLE ANALYSIS
Following sampling, all PSD filters were removed in the lab. These filters were sonicated for 15 min in 5 mL of deionized water. Each PSD and the one WRAD sample were run twice through a Dionex® Model DX-500 Ion Chromatographic Workstation. The sample results at each station were averaged to obtain the mean ambient air ammonia concentration at that location.
PSD SAMPLING
The Ogawa® Passive Sampling Device (PSD) [Fig. 2 & 3] was used to sample for ambient ammonia air concentrations. Each PSD is equipped with two disposable 13-mm glass fiber filters for ammonia gas collection. Filter cleaning, preparation, and extraction procedures identified in ManTech (1999) were utilized. Cleaned filters were coated with an 0.85% phosphoric acid solution prior to insertion in the PSDs. Assembled PSDs were sealed in polyethylene bags, stowed in airtight orange styrene containers, and stored under refrigeration. A cooler with ice packs maintained the cold chain during transport.
The PSD sampling scheme employed was designed to determine ammonia concentrations at varying distances from the facility. This scheme required positioning PSDs in three sample lanes from the facility’s downwind side. These sampling lanes were orientated at 225°, 180°, and 125° angles from the facility and labeled as Lanes 1, 2, and 3; respectively. Each lane was comprised of four sampling stations with two PSDs at each station. These stations were placed at 1 m, 16 m, 31 m, and 46 m distances from the facility (distances A, B, C, and D; respectively). Blank PSD samples were set at the 1 meter stations [Fig. 4]. Lane 2 had an additional sampling station at 61 meters (distance E). This sampling scheme is shown in Figure 5. The sampling period was set at nine hours to coincide with normal facility operating hours, since past sampling had indicated that measurable ammonia releases only occur during facility operation. The sampling was conducted from on 5 October.
Figure 2. Ogawa® PSD.
Figure 1. Sludge Treatment Processing Facility
ISC MODELING
The final step of the experiment was to input source, facility, and meteorological data into the Industrial Source Complex (ISC) model with an initial guess of the source emission rate. For modeling purposes, the source was characterized as an area source. Meteorological variables were obtained from surface weather observation data at the Tampa International Airport. The model was run with successive guesses until predicted ammonia concentrations reasonably agreed with measured concentrations. The model results and the measured data values were plotted for a polar grid coordinate system. Surfer 7® software was then used to produce contour graphs of the downwind concentration gradients for both the modeled and measured data.
RECOMMENDATIONS
Further PSD sampling at distances greater than 61 meters should be performed to determine the distance where detectable levels of ammonia are no longer present. A minimum of three PSDs rather than two should be collocated at the sample stations nearest the facility. This should compensate for the greater variability in ammonia concentrations found at these stations and permit the derivation of a more representative average concentration. WRAD sampling should be repeated at the site with a 1 hour sample cycle time to prevent the complete evaporation of the coating solution. This sample data should be compared with the concentrations determined from collocated PSDs. The ISC modeling should be repeated to incorporate actual on-site meteorological data. The building’s effects on the concentration gradient should also be factored into the model. This may require taking air flow measurements at various locations along the perimeter of the semi-enclosed facility.
ISC MODEL RESULTS
The Industrial Source Complex model predicted an ammonia release rate to the atmosphere of 4.5 mg s-1 m-2. This corresponds to an annual release of 30 metric tons assuming a 9 hour work day and 5 day work week. The modeled and measured concentration gradients were dissimilar. The modeled results did not diminish as rapidly at increased distances. In addition, the model concentrations predicted at Lane 1 were too low while the Lane 3 values were too high. Possible explanations include the meteorological data for Tampa International Airport was not representative of the actual weather conditions at the Pasco County Sludge Facility, the building had a significant influence on the concentration gradient, and the source was incorrectly characterized in the modeling program. The modeled and measured downwind ammonia concentration gradient contour graphs plotted by the Surfer 7® software are shown in Figures 10 and 11. The contour graph from the modeled data more vividly defines the source and depicts less downwind ammonia dispersion than the corresponding measured data contour graph.
Figure 3. Disassembled PSD.
NH3 Conc.(µg/m3)
CONCLUSIONS
The Pasco County Sludge Facility releases a significant amount of ammonia gas to the atmosphere. This emission was estimated to be 30 metric tons per year under normal facility operating conditions and an average ambient temperature of 25° Celsius.
Ambient air ammonia concentrations diminished with an increased distance from the facility.
Detectable quantities of ammonia were measured at 61 meters from the facility. No PSD samples were collected past this distance.
The variability between collocated PSDs were within reasonable limits. The greatest variability occurred at the sample stations closest to the facility.
The WRAD sample collection cycle time of two hours resulted in the complete evaporation of the coating solution. This resulted in the WRAD sample concentration being 33% less than the mean ammonia concentration from collocated PSDs.
The model downwind concentration gradients derived from the ISC Model were lower than the measured gradient.
Figure 5. Sludge Processing Facility Sampling Scheme
Figure 4. Sampling Station
with Blank Sample.
PSD RESULTS
Ammonia concentrations were detected at all the sampling points. These concentrations diminished the further from the facility sampled. The Lanes 1(225°) and 2 (180°) ammonia concentrations were similar and tended to be higher than the corresponding Lane 3 concentrations. This could be attributed to the wind which blew the ammonia plume from the facility away from the outlying Lane 3 sample stations. The only exception to this was the concentration at station 3A (Lane 3, 1 m distance). This station had the highest ammonia concentration measured. This could be attributed to a large processed sludge pile located near this station. The other two stations were much further away from the sludge piles. The ammonia concentrations also showed more variance at the closer sampling stations with the 1m sampling stations demonstrating the highest variability between collocated PSDs. A possible cause was air flow turbulence from the 2 meter high concrete wall surrounding the semi-enclosed facility and an adjacent fully enclosed building. This adjacent building was located to the facility’s immediate rear and housed the sludge processing equipment. The concentration gradients measured for Lanes 1, 2, and 3 are graphed in Figures 7, 8, and 9.
Figure 7.
NH3 Conc.(µg/m3)
WRAD SAMPLING
A wet rotating annular denuder (WRAD) was also used to sample ambient air ammonia concentrations. The WRAD [Fig.6] was a prototype model on loan from the Florida Department of the Environment. The WRAD was positioned inside the facility next to a conveyor belt, where the final treated sludge product exited the process stream. An 0.85% phosphoric acid solution was used to capture the ammonia gas. One WRAD sample was collected during a 2-hour sampling period on 19 October. Two PSDs were collocated with the WRAD for comparison purposes. These PSDs sampled for 3 hours to insure that a detectable quantity of ammonia was collected. A blank sample was also employed.
REFERENCES
• ManTech Environmental Technology, Inc., Standard Operating Procedure for Analyzing Anions and Cations with the Model DX-500 Ion Chromatographic Workstation, the Ogawa TCI-NOx 1000 Analyzer, and the Orion 911 pH Meter, May 1999.
• Pacific Enviromental Services, Inc., Industrial Source Complex Users Guide, Volume II, January 1996.
Figure 8.
Figure 9.
NH3 Conc.(µg/m3)
NH3 Conc.(µg/m3)
Figure 6. Wet Rotating Annular Denuder.
ACKNOWLEDGEMENTS
Archie Ellwood, University of South Florida College of Public Health, Tampa Bay Estuary Program, and Florida Department of the Environment Protection.