BioCycle March 2005, Vol. 46, No. 3, p. 25
Analysis of alternative methods – agitated bed, aerated static pile and modified tunnel – show that properly applied facility configurations with biofiltration would reduce odors to acceptable levels at Colorado site.
Todd O. Williams and Stacey Servo
ODOR GENERATION and impact on surrounding communities is a significant issue facing existing and proposed composting facilities. Left unchecked, odor impacts can lead to a facility shut down. Odor modeling can be used to evaluate alternative operating methods, siting options and odor control technologies by calculating projected impacts on the surrounding community and demonstrating compliance with regulatory and community odor restrictions prior to facility siting and construction.
To demonstrate the odor modeling tools, this article uses the example of a municipality that is in the process of developing biosolids dewatering and composting facilities to generate a Class A biosolids product. Conceptual layouts for three alternative composting processes including agitated bed, aerated static pile, and modified tunnel were developed and evaluated. Each facility layout was sized to process 15 dry tons per day of dewatered biosolids on a five day per week basis. An odor evaluation was performed for a specific site using the United States Environmental Protection Agency (USEPA) recommended Industrial Source Complex Short Term (ISCST3) air dispersion model to quantify the potential odor impact on the surrounding community resulting from the alternative facility configurations. The analysis used a ten-minute duration odor peaking factor based on published values. In all three configurations, emissions from active mixing, composting, and curing processes were assumed to be collected and treated by biofiltration odor control systems before release to the atmosphere.
Five years of local meteorological data were used to cover a wide range of atmospheric conditions. The meteorological and local topographical data were used to determine the overall dispersion of the odor plume. In addition, specific receptor locations were selected to represent the nearest (most sensitive) residential and commercial receptors. A five square kilometer area surrounding the facility was modeled with 12 individual receptor locations examined for off-site odor impacts.
According to the example state’s odor emissions regulations, the odor standard for areas used predominantly for residential or commercial purposes states “…it is a violation if odors are detected after the odorous air has been diluted with seven or more volumes of odor free air” at the facility boundary. For purposes of this evaluation, an odor concentration at a receptor located outside the facility property line exceeding seven dilutions-to-threshold was considered an odor impact. The evaluation indicated that all three facility configurations were comparable resulting in minimal offsite odor impacts. Additional analysis concluded that modifying odor control system locations and layouts significantly affected the overall odor impact resulting from each of the facility configurations. The impact of layout modifications on predicted off-site odor impacts is presented.
CREATING CLASS A BIOSOLIDS
The planned biosolids recycling site is surrounded by public agency-owned property. Several residents reside within a half-mile of the recycling site property boundary. A high level of priority was placed on enclosing and treating the greatest odor generating potential parts of the dewatering and composting facilities. In order to assess the relative impact of one composting technology type versus the other, facility layouts were developed and odor modeling performed to determine the predicted performance of each technology/layout option. All were sized to process 15 dry tons per day of dewatered biosolids on a five day per week basis. All configurations include fully enclosed dewatering, mixing, composting, screening and curing operations with covered bulking agent and product storage buildings. Exhaust air from the fully enclosed buildings is captured and then treated through a conventional organic media based biofilter prior to discharge into the atmosphere. The evaluated composting alternatives are described below in detail.
Alternative A – Agitated Bed
This system consists of elongated bays running the length of a composting hall. An agitator is used to convey the compost contents down the length of concrete aeration bays, typically on a daily basis. At the end of the 21-day composting period, the material is discharged out the end of bays onto the floor where the material is then screened and loaded out to the 30-day aerated curing process.
Alternative B – Aerated Static Pile
The aerated static pile system is designed such that the compost mixture is placed by large equipment such as front-end loaders over an aeration plenum and then the mass is aerated for a period of 21 days. After screening, the material is loaded into the 30-day aerated curing process.
Alternative C – Modified Tunnel
The modified tunnel configuration consists of discrete buildings each typically containing one week’s quantity of material arranged in a static pile configuration. Concrete walls on three sides contain the material so little space is wasted due to pile sides tapering to the floor. A limited headspace exists above the pile thereby reducing the amount of exhaust air that must be treated. Enough tunnels are provided for 21 days of composting. In this analysis, these discrete structures are enclosed in one larger structure and off gases are again collected for treatment. Screening and curing follow composting, similar to the other two alternatives.
REGULATIONS AND MODELING METHODOLOGY
The USEPA recommended ISCST3 air dispersion model was used to quantify the potential odor impact on the surrounding community resulting from the three configurations. The ISCST3 utilizes hourly meteorological data to define the conditions of the plume contributing to pollutant transport and dispersion. The model estimates pollutant concentrations for each source and receptor for each hour of input meteorological data and calculates a user-specified short-term average concentration. The shortest averaging time for the ISCST3 is one hour; however, humans can detect odor at much shorter duration than one hour. An averaging time equal to 10 minutes was selected to represent a short duration odor concentration or peak odor concentration.
A conversion factor was applied to the model to convert the hourly odor concentrations (mean concentrations) to a 10-minute odor concentration (short duration concentration or peak concentration).
Using the -1/5 power law, the peak to mean concentration ratio can be calculated as follows:
C2/C1 = (t2/t1) -1/5
Where: C1 = Concentration estimate for
C2 = Concentration estimate for
t1 = Mean averaging time
t2 = Peak averaging time
Using this equation, peak to mean concentration ratio equals 1.43. This value was used as a modeling input for the model runs performed.
Odor concentrations resulting from previous sampling at existing biosolids composting facilities were used to determine odor emission rates for each odor source, including bulking agent stockpile, product storage stockpile and biofilter(s). Dilution to threshold values of 47 D/T, 31 D/T and 25 D/T respectively were assumed for these three odor sources. A surface flux velocity of 0.0021 feet per second was assumed for all static sources (bulking agent stockpile and product storage pile) based on flux chamber sampling rates used. A flow rate of 4 cubic feet per minute per square foot of biofilter surface was used for all biofilter sources. The emission rates for each of the odor sources for Alternative A, Alternative B and Alternative C were calculated and input into the model. Due to differences in the three composting configurations, varying odor emission rates resulted with Alternative B being the highest, Alternative A the second highest and Alternative C the lowest. Five years of local meteorological data was input into the model and used to cover a wide range of atmospheric conditions. The meteorological and local topographical data were used to determine the overall dispersion of the odor plume.
In addition, specific receptor locations were selected to represent the nearest (most sensitive) receptors. A five square kilometer area surrounding the facility was modeled with twelve individual receptor locations examined for off-site odor impacts. For purposes of this evaluation, an odor concentration at a receptor located outside the facility property line meeting or exceeding seven D/T was considered an odor impact.
IDENTIFYING ODOR SOURCES
The results of the model are shown in a plan view of the facility and the surrounding area with odor sources identified and curves (isopleths) indicating the maximum 10-minute odor concentrations expected to occur (with 7 D/T as the regulatory threshold odor concentration at the facility property line). Figure 1 shows the maximum 10-minute average odor concentrations expected to occur resulting from the Alternative A (Agitated Bed) facility configuration. Figure 2 shows the maximum 10-minute average odor concentrations expected to occur resulting from Alternative B (Aerated Static Pile) facility configuration. Figure 3 shows the maximum 10-minute average odor concentrations expected to occur resulting from Alternative C (Modified Tunnel) facility configuration. Figure 1 indicates that Alternative A is expected to generate odor concentrations greater than 7 D/T to the east of the facility property line. Figures 2 and 3 indicate that Alternative B and C are expected to generate odor concentrations less than 7 D/T at the property boundary.
After reviewing the results, a fourth model was generated using the Alternative A conceptual layout except the biofilter was split into two separate biofilters and moved approximately 250 feet apart. This was done in an effort to demonstrate that all three composting technology options could be shown to function without creating an offsite odor impact. The results are shown in plan view with the facility odor sources identified and curves (isopleths) indicating the maximum 10-minute odor concentrations expected to occur. Figure 4 shows the maximum 10-minute average odor concentrations expected to occur resulting from the Alternative A facility configuration with the biofilter split and moved 250 feet apart. The results indicated that splitting the biofilter decreased the overall number of impacts and the area of odor impact expected to occur from Alternative A. This is attributable to the impact of the site topography and structures on the site to disperse the plume more completely within the property boundary of the facility.
USING THE TOOL ELSEWHERE
As a result of this analysis, it was determined that splitting a larger area biofilter into several smaller area biofilters with equal cumulative flow being treated would reduce the area of odor impact predicted to occur. This understanding is now being applied to subsequent facility configurations for future analysis of the best overall configurations.
The result of this analysis showed that a well-designed and enclosed biosolids dewatering and composting facility using biofiltration, as the odor control technology, would reduce odors to acceptable levels at a specific site. Odor dispersion modeling, using site-specific meteorological and topographical information, is a powerful tool for predicting the impact these facilities will have on neighboring receptors. Once initially set up for a specific site, the modeling of multiple facility configurations can be done with relative ease. This approach can be used to provide meaningful data for agency and engineering evaluation of the multiple alternatives as well as to assist in public outreach and education.
Todd Williams is with Tetra Tech, Inc. based in Mechanicsville, Virginia. Stacey Servo is with Camp Dresser & McKee in Bellevue, Washington. Questions regarding this article can be directed to email@example.com.
March 28, 2005 | General
ODOR MODELING AT A BIOSOLIDS COMPOSTING FACILITY
BioCycle March 2005, Vol. 46, No. 3, p. 25