• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Correlating Emissions With Time and Temperature to Predict Worst- Case Emissions From Open Liquid Area Sources

Posted on: Tuesday, 16 August 2005, 03:00 CDT

ABSTRACT

The two primary factors influencing ambient air pollutant concentrations are emission rate and dispersion rate. Gaussian dispersion modeling studies for odors, and often other air pollutants, vary dispersion rates using hourly meteorological data. However, emission rates are typically held constant, based on one measured value. Using constant emission rates can be especially inaccurate for open liquid area sources, like wastewater treatment plant units, which have greater emissions during warmer weather, when volatilization and biological activity increase. If emission rates for a wastewater odor study are measured on a cooler day and input directly into a dispersion model as constant values, odor impact will likely be underestimated. Unfortunately, because of project schedules, not all emissions sampling from open liquid area sources can be conducted under worst-case summertime conditions. To address this problem, this paper presents a method of varying emission rates based on temperature and time of the day to predict worst-case emissions. Emissions are varied as a linear function of temperature, according to Henry's law, and a tenth order polynomial function of time. Equation coefficients are developed for a specific area source using concentration and temperature measurements, captured over a multiday period using a datalogging monitor. As a test case, time/temperature concentration correlation coefficients were estimated from field measurements of hydrogen sulfide (H^sub 2^S) at the Rowlett Creek Wastewater Treatment Plant in Garland, TX. The correlations were then used to scale a flux chamber emission rate measurement according to hourly readings of time and temperature, to create an hourly emission rate file for input to the dispersion model ISCST3. ISCST3 was then used to predict hourly atmospheric concentrations of H^sub 2^S. With emission rates varying hourly, ISCST3 predicted 384 acres of odor impact, compared with 103 acres for constant emissions. Because field sampling had been conducted on relatively cool days (85-90 F), the constant emission rate underestimated odor impact significantly (by 73%).

INTRODUCTION

The two primary factors that influence ambient air pollutant/ odor concentrations are emission rate and dispersion rate. Gaussian dispersion modeling studies for odors, and often for other air pollutants as well, use hourly meteorological data to vary dispersion rates by hour. However, emission rates are typically held constant over the course of a year: the emission rate value measured in the field on whatever day sampling occurred is used directly in the model as a constant value.

Use of constant emission rates can be especially inaccurate for area sources. Emissions from open liquid area sources, such as wastewater treatment plant units, are greater during summer, when temperatures are hotter and compounds, such as volatile organics, volatilize more readily. In addition, emissions that depend on biological activity will increase in the summer, because microbial growth rates roughly double with each 10 C increase in temperature until the optimum temperature is reached.1 For example, in the summer, activity of facultative microbes increases, which depletes oxygen and produces more anaerobic activity; anaerobic microorganisms, thus, produce more odorous sulfides at wastewater treatment plants in the summer.

Suppose that an odor dispersion modeling study is being conducted for a wastewater treatment plant. Field sampling of emission rates, to provide input to the dispersion model, is conducted on a day with hot (maximum annual) temperatures. Measured emission rates are then input directly into the dispersion model as constant values, whereas meteorological variables impacting dispersion are varied hourly over the course of a year. In this case, the number of exceedances of an odor detection or recognition threshold is likely to be overestimated, and the area of odor impact (concentration isopleths) will be overestimated as well.

On the other hand, suppose that field sampling for the wastewater treatment plan odor study is conducted on a day with cooler temperatures. Again, measured emission rates are input directly into the dispersion model as constant values, whereas meteorological variables impacting dispersion are varied hourly over the course of a year. In this case, the number of exceedances of an odor detection or recognition threshold may be underestimated, and the area of odor impact (concentration isopleths) may be underestimated as well.

This second case of underestimation of odor impact because of sampling on cooler days is often the more critical of the two, because an odor study is often attempting to capture worst-case conditions. Unfortunately, however, because of project schedules, not all emissions sampling from open liquid area sources can be conducted under worst-case summertime conditions. Sampling may need to be conducted in spring or fall or may be scheduled for a summer day when temperatures are relatively cool. Fortunately, correlations of concentrations with temperature can be used to scale flux chamber measurements to predict worst-case emission rates under hotter conditions.

Accordingly, this article presents a method of varying emission rates from open liquid area sources based on temperature and time of the day to predict worst-case emissions. As a test case, the method is used to vary emission rates from the Rowlett Creek Wastewater Treatment Plant in Garland, TX. The emission rates varied by hour are then used as inputs into a dispersion model to predict atmospheric concentrations of odors. Although the curve-fit coefficients presented in this article are specific to the Rowlett Creek plant, the methodology used to develop them could be applied to wastewater treatment units from other plants and, more generally, to any liquid area source that is open to the atmosphere.

This study is significant in that it is the first we know of to try to account for the impact of temperature on emission rate for area sources, rather than just assuming that the emission rate is constant. Because the method is innovative, no previous literature related to prediction of emission rates from area sources is available to be cited. The case study shows that this new method can be very significant in adjusting odor dispersion modeling results when area sources are involved.

Many references, however, document the significance of liquid area sources and particularly of odorproducing sources.2,3,4,5 In the United States, odor complaints account for over half of all complaints to air pollution regulatory agencies.6,7 In Japan, odor complaints to authorities rank second only to noise complaints.8 In Europe, 13-20% of the population reports being annoyed by environmental odors.9,10,11 Several entire conferences cosponsored by the Air & Waste Management Association and Water Environment Federation have been devoted to the topic of odor and volatile organic compound emissions, with odor emissions most often originating from area sources.12,13,14,15 In addition, the Air & Waste Management Association sponsors a session at its annual meeting dedicated to the topic of odor quantification and control. With increasing industrialization and shrinking space, more houses are being built near wastewater treatment plants; hence, the issue is becoming even bigger, and there is mounting pressure on wastewater treatment plants to control odorous emissions.16,17

Varying pollutant emissions from liquid area sources according to temperature and time of day draws on liquid-to-gas mass transfer theory, discussed in the next section.

Liquid-to-Gas Mass Transfer Theory

As the turbulence in a treatment unit increases, K^sub L^ increases, and the rate of pollutant stripping to the atmosphere increases. As the exposed surface area of a treatment unit increases, A increases and so does the rate of pollutant volatilization. Finally, the expression (p^sub p^/H C^sub L^) represents the "driving force" for mass transfer of pollutant from the liquid to the gas phase. When this concentration gradient is negative (i.e., C^sub L^ > p^sub p^/H), pollutant is transferred from the liquid to the gas phase (and not vice versa). Higher concentrations of pollutant in the liquid phase (higher values of C^sub L^) and lower concentrations in the gas phase (lower values of p^sub p^) will increase this "driving force" for transfer from the liquid to the gas.

Development of Correlations for Varying Emission Rates with Temperature and Time

Pollutant emissions can vary markedly depending on liquid temperature. Henry's law constants increase with temperature, meaning that gases are less soluble in water at higher temperatures; thus, more pollutant partitions into the air. The Henry's law constant for hydrogen sulfide (H^sub 2^S) in water, as a function of temperature, is shown in Figure 1.18 In addition, for the particular case of odor emissions from wastewater treatment plants, dissolved oxygen is less soluble in water at higher temperatures, meaning that anaerobic conditions are more likely to occur. Because H^sub 2^S and other odorous reduced sulfide\s are generated under anaerobic conditions, odor emissions are likely to increase. Biological activity, necessary for generating H^sub 2^S and other odorous reduced sulfides, also increases with increasing temperature.

In addition to varying with temperature, emission rates can exhibit a daily pattern based on regular variations in pollutant concentrations in the liquid. For example, if waste strength is routinely higher in the early morning, emissions will increase in early morning (not accounting for temperature).

Figure 1. Henry's law constant for hydrogen sulfide in water vs. temperature.

To measure how emissions vary with temperature and time of day, a pollutant monitor can be used to record pollutant concentrations and temperature for a period of days. It is assumed that pollutant concentrations measured directly over an area surface are proportional to pollutant emission rates. Since the pollutant monitor was typically blocked from the wind, this was likely a valid assumption. Curve-fitting methods can then be used to correlate measured concentrations with time and temperature. In the correlations, pollutant concentration is a dependent variable, and time and temperature are both independent variables.

By experimenting with various curve-fits, it was found in previous work that in most cases, a linear function of temperature and a tenth order polynomial function of time fit the data well (hydrogen sulfide emissions from a wastewater treatment plant).19 This particular curve-fit also has a rational basis, as described below.

If the variation in Henry's law constant dominates the temperature relationship, H^sub 2^S emissions would be expected to be a linear function of temperature, because the Henry's law constant varies linearly with temperature, as discussed above and as shown in Figure 1.

A sinusoidal-like curve would capture regular daily patterns in emissions. A tenth-order polynomial has enough degrees of freedom to capture these periodic variations in emissions.

Although lower order polynomial functions were tried in previous work, they did not provide enough degrees of freedom to satisfactorily capture periodic variations in emissions.

In using eq 2, t is allowed to vary from 0 to 24. This means that all of the data collected at a specific time, such as 9:00 a.m., is considered collectively in determining the curve-fit coefficients. For example, if emissions were measured every second for seven days, seven data points would be available at each value of t to determine the curve-fit coefficients a-1. It is assumed that what happened on those five days at 9:00 a.m. is representative of what happens year- round at 9:00 a.m., not including temperature impacts, which are accounted for separately.

Because the correlation is a linear function of temperature, it should be able to be used quite simply to predict emissions at warmer temperatures. In previous work, the correlation was tested with field data over a broader range of temperatures (60 F to 100 F), and the linear relationship with temperature was found to hold, as expected, because it has a scientific basis in the linear variation of Henry's Law Constant with temperature.19

Curve-fit software can be used to determine values for the coefficients in eq 2 from field data. Coefficients will differ for different treatment units because of differences in wastewater composition, turbulence and flow conditions, and surface area available for volatilization.

In our work, the pollutant monitor was typically placed in a location blocked from the wind, so that wind effects did not have to be considered in eq 2. This makes sense, given that eq 2 is being used to adjust flux chamber measured emission rates. A typical flux chamber blocks the liquid surface from wind effects, so it produces emission rates that are also independent of wind speed and direction. Jiang et al.20 and Jiang and Kaye21 have investigated methods of accounting for wind effects in flux chamber measurements.

Once coefficients are developed for eq 2 from field data for a particular treatment unit, eq 3 below can then be used to vary emissions rates, or emission fluxes, according to time and temperature over the course of a year for use in hourly dispersion modeling.

Eq 3 allows a single flux chamber measurement (in g/sec/m^sup 2^), which is difficult to collect, to be varied over various times and temperatures based on data-logging monitor data (in ppm), which is easy to collect, to generate hourly values for use in dispersion modeling. This is important, because dispersion modeling from area sources must be based on flux chamber data, which gives emission measurements in the required units of g/sec/m^sup 2^. A flux chamber, however, is labor-intensive and requires personnel present in the field for operation. A flux chamber generates only one emission flux value per 20 min and is difficult to operate in the field for extended periods. Flux chamber data is, thus, expensive to acquire. In contrast, data-logging monitors can be left in the field unsupervised for weeks, recording concentration values every second; data-logging monitor data is, thus, inexpensive to collect. Data- logging monitor data can be used to scale flux chamber data according to eq 3 but cannot be used directly as input to a dispersion model, because pollutant concentrations are measured in ppm or similar units. For area sources, dispersion models require emissions per time per unit surface area.

Case Study: Rowlett Creek Wastewater Treatment Plant, Garland, TX

The City of Garland owns and operates the Rowlett Creek Water Recycling Center, or wastewater treatment plant, which currently treats an average flow of 16 million gallons per day and is permitted up to 24 million gallons per day.22

Methodology

Field Data Collection. Surface hydrogen sulfide emission fluxes from area sources were measured using an Ac'Scent emission isolation flux hood, or flux chamber, from St. Croix Sensory. A vacuum chamber equipped with an airsampling pump was used in conjunction with the flux chamber to draw gases into a Tedlar bag. The flow rate from the surface to the Tedlar bag was recorded.

Table 1. Coefficients for correlations of emissions versus time/ temperature (eq 2) for various units at the Rowlett Creek Water Recycling Center.

An Arizona Instrument Model 631 Jerome meter was used to measure the gaseous concentrations in the Tedlar bags. The Jerome meter is a low-range hydrogen sulfide monitor capable of measuring concentrations from 1 to 50 ppm. Concentrations from point sources were measured directly from the stack using a Jerome 631X meter.

Continuous data-logging monitors, Apptek OdaLogs, were used to record hydrogen sulfide emissions every minute for a period of four to six days. OdaLogs can record hydrogen sulfide concentrations up to 1000 ppm.

Development of Coefficients for Emissions versus Time/ Temperature Correlation. The hydrogen sulfide emissions measured by the OdaLog were input to a curve fitting software, DataFit, a tool that simplifies the tasks of data plotting, regression analysis, and statistical analysis. Using DataFit, coefficients for eq 2 above were determined, so that hydrogen sulfide emissions could be expressed as a function of time and temperature. These coefficients for various treatment units are given in Table 1.

Varying Emissions with Time/Temperature. A customized database, which uses the OdaLog correlations to calculate emissions each hour based on hourly temperature data, was used to generate an hourly emissions file for use in ISCST3, as described below.

Dispersion Modeling Cases: Constant Emissions and Varying Emissions. Dispersion model simulations using constant and varying emissions were compared in an effort to determine the impact of time/ temperature variations.

The constant emissions case used the methodology typical in many odor-modeling studies. Constant emission rates, based directly on flux chamber measurements, are input into the model. The flux chamber measurements are effectively "grab samples," which represent emission "snapshots," but are unlikely to represent peak or average conditions. Measurements made during the summer may approach peak conditions, because emissions increase with temperature. However, emissions often vary with time of day, independently of temperature. Unless flux chamber measurements are taken at the peak time on the hottest day of the year, they will not represent absolute peak emissions.

In the varying emission case, dispersion modeling was conducted using time/temperature variations. The "hourly emissions file" option was chosen in ISCST3, and the hourly emission file generated by the customized database was used.

Dispersion Model Inputs/Assumptions. The dispersion model used in this study was U.S. EPA ISCST3 (Industrial Source Complex Short Term 3), with the following inputs/assumptions.

Meteorological Data. Hourly meteorological data from the Dallas- Fort Worth airport was obtained from the EPA Support Center for Regulatory Air Models for the year 1988.23 These data includes hourly values of temperature, wind speed, wind direction, and cloud cover. Upper air meteorological data was from the Stephenville weather station, which is the closest station to Dallas/Fort Worth with upper air data.

Receptor Grid. A 125 m 125 m grid with 25-m spacing was used to cover the treatment plant and surrounding area.

Downwash. Building Profile Input Program was used to calculate building downwash associated with point sources.

Tenain/Receptor Elevations. Terrain/receptor elevations were obtained from digital elevation model data from WebMet.24

Chemical Reaction/Dry Deposition. It was assumed that no H^sub 2^S removal occurs because of chemical reaction or dry deposition. Because H^sub 2^S reacts slowly in the atmosphere, the assumption of no chemical reaction is likely valid.

Other Model Options. Regulatory default mode, urban dispers\ion coefficient option, elevated terrain, and simple + complex terrain were used.

Odor Detection Threshold. An odor detection threshold of 0.5 ppb was used for H^sub 2^S to determine the hours per year with odor exceedances.

ISCST3 output was post-processed with Arc View 3.2.

Figure 2. Hydrogen sulfide concentrations measured by OdaLog.

RESULTS AND DISCUSSION

Field Data

Figure 2 provides an example of ambient hydrogen sulflde concentrations measured continuously over a multiday period using an OdaLog. The data shown was collected over a grit basin for 7 days. The dark thin line in the lower portion of the figure represents hydrogen sulfide emissions in ppm. The lighter color, thicker line in the upper portion of the figure represents the temperature variations recorded by the OdaLog during the same period.

Emissions versus Time/Temperature Correlations

Figure 3 shows emissions from one of the Rowlett Creek primary clarifiers, plotted as a function of time and temperature, along with the best-fit eq 2 curve. Coefficients for eq 2 for the primary clarifier are shown in Table 1, along with coefficient values for the other treatment units.

For a number of treatment units, the eq 2 curve fit had an R^sup 2^ value about equal to either the time or temperature curve-fit alone, meaning that the other variable did not improve the curve- fit. In a number of these cases, time served as a surrogate for temperature: emissions varied with time, because temperature varied with time but not because the waste strength varied with time. In these cases, time did not improve the emissions correlation over temperature alone.

It can be seen from Table 1 that emissions from some of the units, like anaerobic digester pressure relief valves, belt press room, and solids building scrubber, did not correlate with time and temperature variations. For such cases, hydrogen sulfide emissions were considered to be constant.

Figure 3. Hydrogen sulfide emissions vs. time and temperature = measured OdaLog values and eq 2 curve-fit.

Dispersion Modeling

The results of modeling using constant and varying emissions are shown in Figure 4. The inner line represents the 0.5-ppb isopleth (hydrogen sulfide detection threshold) for the constant emissions case. Because field sampling had been conducted on days when temperatures were relatively cool (85-90 F), the field sampling did not capture worst-case conditions. Hence, this case underestimated the area of odor impact.

Figure 4. Comparison of hydrogen sulfide 0.5-ppb concentration contours for constant and varying emission rates.

The outer line in the figure represents the 0.5 ppb isopleth for the varying emissions case. The meteorological data file contained days with hotter temperatures (up to 107 F) than those recorded on the field sampling day.

Contour areas were calculated for constant and varying emissions cases. For the varying emissions case, the area of the 0.5-ppb H^sub 2^S isopleth was 384 acres. For the constant emissions case, the area of 0.5-ppb H^sub 2^S isopleth was 103 acres, which is only 27% of the varying emissions case. For the varying emissions case, the time/temperature correlations enabled predictions of increased emissions on hotter days. Including emissions from these hotter days in dispersion modeling increased the area of odor impact by more than a factor of 3.

CONCLUSIONS

Emissions Can Be Correlated with Time and Temperature Using Field Data

In particular, in this study, H^sub 2^S emissions from a wastewater treatment plant were correlated with time and temperature using Odalog field data. In most cases, a linear correlation with temperature and a tenth-order polynomial function of time fit the data well. Correlations of emissions with time and temperature can be used with the ISCST3 "hourly emission file" option to vary emissions over the course of a year.

Emission Inputs to Dispersion Models From Open Liquid Area Sources Should Be Varied by Time and Temperature

Varying temperature and time in dispersion modeling reflect the actual emissions from open liquid area sources. Unlike the constant emissions case, which represents the emissions at a particular time and temperature, varying emissions create a realistic case. In the case study presented in this article, the 0.5-ppb H^sub 2^S odor detection isopleth when emission rates were varied covered an area over three times larger than for the case that used constant emission rates; hence, if emission variations with time and temperature had not been included, H^sub 2^S concentrations would have been significantly underestimated.

Correlation Coefficients Are Case Specific

A correlation predicting emissions as a linear function of temperature and a tenth-order polynomial function of time has been found to fit field data from various open liquid area sources well. Correlation coefficients (a, b, c, etc.) for eq 2, however, are case specific and must, thus, be determined on a case-by-case basis. The correlation coefficients can be expected to differ for different treatment units at the same plant because of differences in wastewater composition, turbulence and flow conditions, and surface area available for volatilization. For the same reasons, correlation coefficients would also differ for units at different plants.

IMPLICATIONS

The method presented in this paper is useful in predicting worst- case odor emissions, and subsequently atmospheric concentrations, from field data collected under conditions that are not worst case. The technique applies not only to odorous compounds from wastewater treatment plants but also to any air pollutant from an open liquid area source. This study is significant in that it is the first we know of to try to account for the impact of temperature on emission rates from area sources. The case study shows that this new method can be very significant in adjusting odor dispersion modeling results when area sources are involved.

REFERENCES

1. Metcalf & Eddy. Wastewater Engineering: Treatment, Disposal, and Reuse. 3rd Ed. Irwin/McGraw-Hill: Boston, MA, 1991.

2. Control of Odors & Emissions from Wastewater Treatment Plants; Manual of Practice 25. Water Environment Federation: Alexandria, VA, 2004.

3. Ruokojarvi, A.; Ruuskanen, J.; Martikainen, PJ. Oxidation of Gas Mixtures Containing Dimethyl Sulfide, Hydrogen Sulftde, and Methanethiol Using a Two-Stage Biotrickling Filter; J. Air & Waste Manage. Assoc. 2001, SJ, 11-16.

4. Jones, Kim D.; Martinez, A.; Maroo, K.; Dcshpande, S. Kinetic Evaluation of H^sub 2^S and NH^sub 3^ Biofiltration for Two Media Used for Wastewater Lift Station Emissions; J. Air & Waste Manage. Assoc. 2004, 54, 24-35.

5. Kim, H.; Kim, YJ.; Chung, J.S. Long-Term Operation of a Biofilter for Simultaneous Removal of H^sub 2^S and NH^sub 3^; J. Air & Waste Manage. Assoc. 2002, 52, 1389-1398.

6. O'Brien, M; Deffee, R.A.; Ostojic, N. Odors: Indoor and Environmental Air, McGinley, CN. and Swanson, J.R., Eds.; Air & Waste Management Association: Pittsburgh, PA, 1996; pp 261-267.

7. Hudon, G.; Guy, C.; Hermia, J. Measurement of Odor Intensity by an Electronic Nose; J. Air & Waste Manage. Assoc. 2000, 50, 1750- 1758.

8. Higuchi, T.; Konosuke, N. Odors: Indoor and Environmental Air; McGinley, C.N.; Swanson, J.R., Eds.; Air & Waste Management Association: Pittsburgh, PA, 1996; pp. 181-192.

9. Walpot, J.I. Characterization and Control of Odours and VOC in the Process Industries; Vigneron, S.; Hermia, J.; Chaouki, J., Eds.; Elsevier Science: Amsterdam, 1994; pp. 119-125.

10. Berglund, B.; Lindvall, T. The Nose; Proctor, D.F.; Andersen, L, Eds.; Elsevier Biomedical: Amsterdam, 1982; pp. 279-305

11. Hudon, G.; Guy, C.; Hermia, J. Measurement of Odor Intensity by an Electronic Nose; J. Air & Waste Manage. Assoc. 2000, 50, 1750- 1758.

12. Proceedings of the Odor and Toxic Air Emissions 1994; Water and Environment Federation, Jacksonville, FL, 1994.

13. Proceedings of the Odor and Toxic Air Emissions 1997; Water and Environment Federation, Houston, TX, 1997.

14. Proceedings of the Odors and VOC Emissions 2000; Water and Environment Federation, Cincinnati, OH, April 2000.

15. Proceedings of the Odors and VOC Emissions 2002; Water and Environment Federation, Albuquerque, NM, April 2002.

16. Serjak, T.E. Odor Modeling: An Effective Tool for Defining the Problem, Selecting the Solution, and Conversing with the Public. In Proceedings of the Odor and Toxic Emissions 1997; Water Environment Federation, Houston, TX, 1997.

17. Diosey, G. P. Atmospheric Dispersion Modeling Techniques for Odor Impact Assessment. In Proceedings of the Air & Waste Management Association's 90th Annual Meeting, Toronto, Ontario, Canada, June 1997; Paper 97-TA35.07.

18. Cooper, C.D.; Alley, F.C. Air Pollution Control: A Design Approach, 3rd ed; Waveland Press: Prospect Heights, IL, 2002.

19. Sattler, M.L.; McDonald, E.T. Time, Temperature, and Wind Speed: The Importance of Varying Emission Rates in Odor Modeling. In Proceedings of the Odors and VOC Emissions 2002; Water Environment Federation, Albuquerque, NM, April 2002; Paper 5d.

20. Jiang, K.; Bliss, PJ.; Schulz, TJ. The Development of a Sampling System for Determining Odor Emission Rates from Areal Surfaces: Part I. Aerodynamic Performance. J. Air & Waste Manage. Assoc. 1995, 45, 917.

21. Jiang, J.K.; Kaye, R. A Comparison of Odor Dispersion Modeling Results Using Emissions Data from Wind Tunnel and Flux Hood. In Proceedings of the Odors and VOC Emissions 2000; Water Environment Federation, Cincinnati, OH, 2000; Paper 4e.

22. Alan Plummer Assoc, Rowlett Creek Water Recycling Center Odor Abatement Evaluation and Master Plan, Draft report, December 2002.

23. U.S. EPA. Support Center for Regulatory Air Models (SCRAM); available at http://www.epa.gov/scram001/ (accessed May 20, 2005).

24. Webmet. Available at http://www.webmet.com (accessed May 20, 2005).

25. Diosey, P.G. Modeling for Odors at Wastewater Treatment Plants. In Proceedings of the Odor\s and VOC Emissions 1997; Water Environment Federation, Houston, TX, 1997; pp 11-33.

Archana Nagaraj

Weaver Boos Consultants, Fort Worth, TX

Melanie L. Sattler

Department of Civil and Environmental Engineering, University of Texas at Arlington, Arlington, TX

About the Authors

Archana Nagaraj is an environmental engineer with Weaver Boos Consultants. Melanie L. Sattler, is an assistant professor of Civil & Environmental Engineering at the University of Texas at Arlington. At the time this study was performed, Archana Nagaraj and Melanie Sattler were engineer/scientists with Alan Plummer Associates in Fort Worth, TX. Address correspondence to: Melanie Sattler, Civil & Environmental Engineering Department, University of Texas at Arlington, Box 19308, Arlington, TX 76016; phone: + 1-817-272-5410; fax: +1-817-272-2630; e-mail: sattler@ce.uta.edu.

Copyright Air and Waste Management Association Aug 2005

Source: Journal of the Air & Waste Management Association

More News in this Category