J APPL POULT RES 2006. 15:394-405
© 2006 Poultry Science Association
Fine Particle Measurements Inside and Outside Tunnel-Ventilated Broiler Houses1
M. C. Visser*,
B. Fairchild
,
M. Czarick
,
M. Lacy,
J. Worley,
S. Thompson,
J. Kastner,
C. Ritz and
L. P. Naeher*,2
* Department of Environmental Health Science, College of Public Health, and Departments of Poultry Science and Biological and Agricultural Engineering, College of Agricultural and Environmental Sciences; The University of Georgia, Athens 30602
2 Corresponding author: LNaeher{at}uga.edu
 |
SUMMARY
|
|---|
Emissions from animal feeding operations have become a growing concern. Although many studies describe occupational exposures and exhaust concentrations associated with animal facilities, very little information has been provided about the ambient air around the houses. This study investigates real-time and primarily 24-h time-integrated levels of particulate matter 
2.5 µm in diameter inside and outside (up to 500 ft from the house) of commercial tunnel-ventilated broiler houses on a farm in northeast Georgia. None of the 24-h particulate matter measures collected when the houses were tunnel ventilated exceeded the Environmental Protection Agency’s 24-h National Ambient Air Quality Standard of 65 µg/m3.
Key Words: fine particle measurement • broiler house • poultry • emission
|
DESCRIPTION OF PROBLEM
|
|---|
As early as 1984, a study was published about the environmental exposures received in poultry houses [1]. In this study, houses were analyzed for concentrations of particulate matter (PM) and several gases. Concentrations of total dust averaged 4.4 mg/m3 and respirable dust averaged 0.24 mg/m3. Average ammonia levels were measured at 25 ppm. Other gases including carbon monoxide, hydrogen sulfides, nitrogen compounds, methane, mercaptan, formaldehydes, and hydrocarbons were tested for, but were too low for indicator tubes to measure. Immediately it became evident that ammonia and particulates were the major exposures in poultry houses.
More studies were completed to assess occupational exposures in different types of animal facilities. A study in Europe showed that the occupational exposures were higher at poultry farms than at swine facilities [2]. The median total dust was found to be 7.0 mg/m3 in chicken houses compared with 4.0 mg/m3 for swine houses (P < 0.001). Another European study from 1998 includes concentrations of particulates at cattle, swine, and poultry facilities [3]. Average concentrations reported in that study for all countries in which facilities were assessed were 3.60 mg/m3 for inhalable and 0.45 mg/m3 for respirable dust. Exposure levels for poultry facilities were generally the same or higher compared with the other facilities.
Today, as rural and suburban communities come into closer contact, homeowners near poultry houses are expressing concern about emissions from poultry houses. The objective of this study was to investigate concentrations of PM 2.5 µm in diameter (PM2.5) in the ambient air up to 500 ft away from tunnel-ventilated broiler houses.
|
MATERIALS AND METHODS
|
|---|
Study Site
A commercial broiler farm in northeast Georgia was selected for the study. The farm consisted of 7 broiler houses and no other livestock were present on the premises. Five of the broiler houses were 40 ft wide by 425 ft long and held approximately 26,200 birds each. One of the houses was 36 ft wide by 425 ft long and held approximately 24,400 birds. The last house was 50 ft wide by 500 ft long and held 38,500 birds. All houses were tunnel ventilated and had evaporative cooling pad systems. Environmental controllers maintained temperature, humidity, and air quality in the houses. The broilers were 24 d old at the start of the study on August 18, 2005, and 35 d old at the end of the study on August 29, 2005. The birds were grown to a live market weight of 3.8 to 3.9 lb and were marketed at 37 to 39 d of age. Older birds were selected so as to examine emission rates when they would tend to be the highest. The farm monitored was representative of broiler farms based on a survey of the nation’s largest poultry companies [4] and the house dimensions were within the national average [5]. Typical farm sizes in the United States range from 1 to 10 houses, whereas the average size is in the range of 4 to 6 houses. The average number of birds per house on a typical farm in the United States varies depending on the bird size. The bird density used on a typical farm in the United States meets the National Chicken Council animal welfare guidelines [6].
The study was designed to measure PM2.5 concentrations in the field when the houses were in tunnel-ventilation mode. The tunnel-ventilated broiler houses were chosen for several reasons. First, tunnel ventilation is the most popular method of ventilating broiler houses during hot weather, so it is the best test for the most common form of ventilation. Also, tunnel ventilation provides a more accurate method of measuring exhaust concentrations. The air enters at one end of a tunnel-ventilated house and gathers dust and gases before exiting at the opposite end. The dust and gases are concentrated and well mixed due to the high ventilation rates and provide a good measurement of the concentration leaving the house. In addition, complaints from neighbors are most likely to occur during the summer when the broiler houses are most likely going to be in tunnel-ventilation mode.
A map of the spatial layout of the houses and the air sampling sites is shown in Figure 1
. All the tunnel ventilation fans were on the east end of the houses. Therefore, samplers were set up in the field to the east of the houses to collect particulate matter emitted. An air sampling station was placed inside 4 of the houses 15 ft from the tunnel fans to measure the exhaust concentration. The samplers outside the houses were placed at distances of 100, 300, and 500 ft from the houses. The outside samplers were laid out to investigate collective PM2.5 emissions from all of the houses, not just from individual houses.
Air Sampling
Real-Time PM2.5.
Real-time PM2.5 measurements were taken with a DustTrak aerosol monitor with data logger [7]. Instruments were calibrated at the factory to standards of a respirable fraction of standard ISO 12103-1, A1 test dust (Arizona Test Dust; aerodynamic diameter of 4.4 µm). Particulate matter can have properties different from the test dust that can alter the instrument’s reading. For example, smaller particles scatter more light than larger ones and can increase the concentration reading. Based on previous studies, it was estimated that sampling of PM2.5 with this instrument might have presented the concentration levels at 2 to 3 times the actual value [8]. Therefore, PM2.5 real-time monitors were paired with gravimetric PM2.5 set-ups. The instruments were placed in and behind house no. 4 at the 100, 300, and 500 ft lines with an intake tube collecting samples at a height of 4 ft (Figure 1
). The DustTraks were zeroed at the beginning of the study and programmed to collect particulate matter data once per minute. One-hour averages were calculated for data analysis.
Time-Integrated PM2.5.
Pumps [9] with a Triplex Cyclone [10] containing a 3-mm Teflon filter [11] were used for gravimetric measurements of PM2.5. Filters were weighed in a climate-controlled room before being loaded into the cyclones. The pumps were set at a flow rate of 1.5 L/min. Pumps were placed in toolboxes on the ground and plastic Tygon tubing was run up the stand and attached to the cyclones at the top; site placement is depicted in Figure 1. At the beginning of the study, pumps operated for 48 h, but this time was shortened to 24 h for the second half of the study. Flow rates were measured at the end of the sampling period each day and used to obtain an average flow. If the flow differed more than 10% from the starting flow, samples were not considered to be valid and were not included in the data analysis.
Cyclones were returned to the lab and filters were removed and frozen at 0°C. Filters were thawed for 48 h before weighing in a climate-controlled room. Filters were weighed using a Cahn C-35 microbalance [12] with a sensitivity of ±1 µg and adjusted for buoyancy following standard methods [13]. Air densities during weighing sessions, nominal densities of calibration masses, and control filters were used to adjust the balance readings for the buoyancy effect of air [14]. Statistical analyses were performed with and without the PM2.5 values that were below 20 µg/m3, which is considered a very conservative limit of detection for the sampler setup used in this study. For reference, a more typical limit of detection for a 24-h sample using this setup would be in the range of 5 to 10 µg/m3. The results with or without the PM2.5 below 20 µg/m3 were unchanged, so all PM2.5 values, including those below 20 µg/m3, were included in the presented statistical analyses.
Statistical Analysis
Statistical analyses were performed using SPSS 13.0 [15]. One-way ANOVA with a Bonferroni test were performed for all the averages using house, 100, 300, and 500 ft sites as the main factors; t-tests were performed at each site to compare with and without birds present.
Weather
Temperature, wind, and precipitation measurements were obtained from the Athens (Georgia) airport located 15 miles from the study site. Wind roses (to illustrate wind direction) were generated using WRPLOT View version 5.2.1 [16].
|
RESULTS AND DISCUSSION
|
|---|
Time-Integrated PM2.5
Time-integrated PM2.5 samples were taken for 11 d when birds were on site. The houses were only in tunnel-ventilation mode for 9 d during this period. During the summer, temperatures are high and broiler houses have almost continuous tunnel fan operation. During the study the ambient temperature dropped below the normal summer temperatures on 2 d (August 25 and 26). When the temperature decreases, the environmental controller reduces the number of tunnel fans in operation, side-fan usage increases and the houses transition to a perimeter air inlet ventilation system. When this occurred, dust concentrations measured 15 ft from the tunnel fans did not provide an accurate measure of exhaust concentrations; therefore, data from these 2 d were excluded from the analysis.
The exhaust concentrations of all sites when the houses were in tunnel ventilation are reported in Table 1 and Figure 2. The first 3 columns of data represent 48-h samples whereas the remaining columns represent 24-h samples. The last column consists of data taken when no birds were present and is considered the control in this study; fans were not operating on this day. Table 1 also provides a summary of the averages at each sampling site (control, house, 100, 300, and 500 ft). The average in-house exhaust concentration (average ± SE; 58.6 ± 3.3 µg/m3, n = 20) was statistically higher than the control concentration (24.0 ± 5.9 µg/m3, n = 5) and all the field concentrations. However, the field concentrations (100 ft: 24.1 ± 1.7 µg/m3, n = 26; 300 ft: 24.9 ± 2.2 µg/m3, n = 26; 500 ft: 23.1 ± 6.4 µg/m3, n = 16) were not significantly different from each other or the control. Although the field concentrations are not statistically different there appears to be a trend with respect to distance from the house. On 4 of the 6 sampling days (August 18, 22, 27, and 28), the 100 ft concentration was higher than those at 300 and 500 ft.
View this table:
[in this window]
[in a new window]
|
Table 1. Fine particle (µg/m3) measurements over a 12-d period from a poultry farm using tunnel ventilation in northeast Georgia
|
|
View larger version (23K):
[in this window]
[in a new window]
|
Figure 2. Gravimetric particulate matter <2.5 µm (PM2.5) measurements (mean ± SE) from 18 sites within a broiler farm during tunnel ventilation at each site over a 9-d period.
|
|
As discussed in the Materials and Methods section, the outside samplers were positioned to investigate collective PM2.5 emissions from all of the houses, not just from individual houses. In general, the data presented in Table 1 and Figure 2 do not show a difference in PM2.5 concentrations between the sites directly behind the houses and the sites behind the houses but offset (not directly behind a house). However, PM2.5 at sites 6 and 11, directly behind house 7 (the largest broiler house with the greatest number of birds on the farm), are generally (but not statistically significantly) higher than those collected in the same period at other locations equidistant (i.e., 100, 300, and 500 ft) from the houses, whereas PM2.5 at sites 8, 13, and 17, directly behind house 4, are not statistically different than those collected in the same period at other locations equidistant from the houses.
Figure 3 displays this relationship and shows that the in-house exhaust concentrations were the only values statistically different from the rest (F = 28.2, P < 0.001). Figure 3 shows the difference between concentrations during the presence and absence of birds. The in-house exhaust concentration was the only one to be statistically higher when birds were present compared with absent (t = 7.8, P < 0.001). There was no statistical difference due to the presence of birds at the control or field levels.
View larger version (22K):
[in this window]
[in a new window]
|
Figure 3. Comparison of the average particulate matter <2.5 µm (PM2.5) gravimetric measurements of the control site with other sites on days with birds and the day without birds. ***P < 0.01.
|
|
It is difficult to compare these findings to other studies for several reasons. Most of the studies completed in Europe or other occupational studies measure inhalable and respirable dust, which cannot be directly compared with the values measured by this study [3, 17, 18]. Inhalable dust generally refers to total suspended particulates, whereas respirable dust generally refers to particulates with a diameter less than 5 µm. For the same reason, this makes it difficult to compare these studies to our regulatory limits, thus increasing the need for future studies that look at PM2.5.
The closest relevant, but not directly comparable, occupational limit for PM was set by the Occupational Safety and Health Administration (OSHA); they defined respirable dust as less than 5 µm and set the 8-h limit at 500 µg/m3 [19]. However, PM2.5 (those particles <2.5 µm) were collected in this study and cannot be directly compared with respirable dust. Therefore, the exhaust concentrations cannot be directly compared with the OSHA standards. The field concentrations measured in this study during tunnel operations ranged from 5.4 to 55.1 µg/ m3 and did not exceed the US National Ambient Air Quality Standards (NAAQS) PM2.5 24-h limit of 65 µg/m3 [20]. As recorded in the Federal Register on January 17, 2006, the Environmental Protection Agency recently proposed a reduction in the PM2.5 24-h NAAQS to 35 µg/m3 [21]. The concentrations measured during the study reported herein are within the current standard, but some of the daily concentrations would exceed the proposed (35 µg/m3) standard.
Real-Time PM2.5
Real-time data were obtained during the same period as the time-integrated samples. Representative data from all 4 locations for 2 d can be found in Figure 4. The real-time data show the time-related maximum and minimum concentrations. An increase in concentrations started at approximately 1200 h on both days and then dropped during the early evening hours (1700 to 2000 h). This increase was likely due to the increase in afternoon temperatures causing an increase in ventilation rate in the houses, thus increasing dust concentrations. Overall, the in-house exhaust concentrations were always higher than the field measurements, as expected. In the field, a pattern can be seen involving field concentration with respect to distance from the exhaust fans. Measurements at the 100 ft site were generally slightly higher than those at the 300 and 500 ft sites. These data support those trends observed with the time-integrated concentrations.
View larger version (24K):
[in this window]
[in a new window]
|
Figure 4. Real-time hourly averages for a 2-d measurement of particulate matter <2.5 µm (PM2.5) at 4 locations with birds present on the farm. Reported rainfall was 0.28 cm for August 19 and 20.
|
|
Figure 5 shows the PM2.5 concentrations during the control portion of the study on the day when no birds were present. It can be seen that house exhaust levels were slightly elevated compared with the outside, but generally lower than when the birds were present. At 100 and 300 ft from the exhaust fans, PM concentrations were almost identical and showed a slight increase in the afternoon.
View larger version (17K):
[in this window]
[in a new window]
|
Figure 5. Real-time hourly averages for a 1-d measurement of particulate matter <2.5 µm (PM2.5) at 4 locations when no birds were present on the farm.
|
|
Weather
A wind rose is a diagram for a given locality or area and shows the frequency or strength of the wind from various directions [11]. Wind roses are presented for each day of the study period (Figure 6). The temperature, precipitation, and average humidity for each day are also reported. The wind direction did not seem to have much effect on the control PM2.5 concentration. The wind was blowing predominantly toward the west (toward the control sampler; Figure 6) for 3 d of the study (August 24 to 25, 27 to 28, and 28 to 29). During these days the control samplers recorded the highest (46.3 µg/ m3) and the lowest (13.8 µg/m3) concentrations for the entire study. Either, the airport weather did not provide an accurate representation of what was occurring on the farm, or something else was affecting the background levels on the farm. For example, if the hay from a nearby field was being harvested, this might increase the background levels of PM. In future studies, an on-site meteorological station would provide more insight into the effect of wind on the concentrations in the field.
View larger version (75K):
[in this window]
[in a new window]
|
Figure 6. Wind roses representing the direction the wind was blowing from and weather data on each day of the study.
|
|
On 10 of the 12 d, humidity levels were above 70%. The high humidity caused some of the filters to be damp. However, damp filters should not have affected the size of the particulates collected. The cyclone prevents anything greater than PM2.5 from reaching the filter; therefore, the wetness of the filter would not alter the dust collected. Also, the samples were placed in a climate-controlled room for 48 h before they were weighed; therefore, the dampness of the filters during collection would not affect the weight during analysis. There was rainfall of 0.28 and 0.89 cm, respectively, on August 19 and 22. This precipitation may have reduced the distance that PM traveled from the house. Therefore, the concentrations in the field would be expected to be lower on those days. However, this phenomenon is difficult to confirm from the data presented in Table 1, because samples collected on those rainy days are part of the 2-d samples.
|
CONCLUSIONS AND APPLICATIONS
|
|---|
- Twenty-four-hour gravimetric time-integrated exhaust concentrations of PM2.5 outside the broiler houses on the farm used in this study in northeast Georgia on days when the houses were using tunnel ventilation ranged from 5.4 to 55.1 µg/m3 and did not exceed the NAAQS limit of 65 µg/m3.
- No significant differences were found between the control site (background levels) and the 100, 300, and 500 ft field concentrations of gravimetric time-integrated PM2.5. Real-time PM2.5 data from the 100, 300, and 500 ft sampling locations indicates that PM2.5 values are slightly greater at the 100 ft location vs. the 300 and 500 ft locations. A slight trend to this effect was also observed in the gravimetric time-integrated PM2.5 data.
|
FOOTNOTES
|
|---|
1 Funding provided by University of Georgia, Interdisciplinary Toxicology Program. 
|
REFERENCES AND NOTES
|
|---|
- Jones, W. 1984. Environmental study of poultry confinement buildings. Am. Ind. Hyg. Assoc. J. 45:760–766.[Web of Science][Medline]
- Radon, K., C. Weber, M. Iversen, B. Danuser, S. Pedersen, and D. Nowak. 2001. Exposure assessment and lung function in pig and poultry farmers. Occup. Environ. Med. 58:405–410.[Abstract/Free Full Text]
- Takai, H., S. Pedersen, J. O. Johnsen, J. H. M. Metz, P. W. G. Groot Koerhamp, G. H. Uenk, V. R. Phillips, M. R. Holden, R. W. Sneath, J. L. Short, R. P. White, J. Hartung, J. Seedorf, M. Schoder, K. H. Linkert, and C. M. Wathes. 1998. Concentrations and emissions of airborne dust in livestock buildings in northern Europe. J. Agric. Eng. Res. 70:59–77.
- Thornton, G., and T. O’Keefe. 2001. Housing and Equipment Survey. Watt Poultry USA (http://www.wattnet.com/).
- Watt Poultry USA. 2004. Live Production Survey/Expansion Plans (http://www.wattnet.com/) Accessed June 15, 2004.
- National Chicken Council. 2005. Animal Welfare Guidelines and Audit Checklist. National Chicken Council, Washington, DC. Accessed Feb 1, 2005.
- TSI Inc., model 8520. TSI Inc., Shoreview, MN.
- MacIntosh, D. L., P. L. Williams, and J. D. Yanosky. 2002. A comparison of two direct reading aerosol monitors with the federal reference method for PM2.5 in indoor air. Atmos. Environ. 36:107–113.
- SKC AirChek 2000, SKC Inc., Eighty Four, PA.
- Model SCC 1.062, BGI Inc., Waltham MA.
- Teflo 2.0 µm, Pall Life Sciences, Ann Arbor, MI.
- Thermo Electron Corporation, Waltham, MA.
- US Environmental Protection Agency. 2003. Chemical/Name Index to EPA Test Methods (April 2003), PM2.5, 40 CFR 50 App L. Available http://www.epa.gov/epahome/index/nameindx3.htm
- Koistinen, K., A. Kousa, T. Virpi, O. Hanninen, and M. Jantunen. 1999. Fine particulate measurement methodology, quality assurance procedures, and pilot results of the EXPOLIS study. J. Air Waste Manage. Assoc. 49:1212–1220.
- SPSS Inc., Chicago, IL.
- Lakes Environmental Software. 1998–2006. http://www.weblakes.com/lakewrpl.html
- Crook, B. 1991. Airborne dust, ammonia, microorganisms, and antigens in pig confinement houses and the respiratory health of exposed farm workers. Am. Ind. Hyg. Assoc. J. 52:271–279.[Web of Science][Medline]
- Wathes, C. M., M. R. Holden, R. W. Sneath, R. P. White, and V. R. Phillips. 1997. Concentrations and emission rates of aerial ammonia, nitrous oxide, methane, carbon dioxide, dust and endotoxin in UK broiler and layer houses. Br. Poult. Sci. 38:14–28.[Web of Science][Medline]
- Occupational Safety and Health Administration. 1999. Air Contaminants. 29CFR part 1910.1000.
- US Environmental Protection Agency. 1997. National Ambient Air Quality Standards for particulate matter, final rule. 40CFR part 50.
- US Environmental Protection Agency. 2006. National Ambient Air Quality Standards for particulate matter, proposed rule. Fed. Regist. 71 FR 2620. http://www.epa.gov/fedrgstr/EPA-AIR/2006/January/Day-17/a177.htm
TOP
SUMMARY
DESCRIPTION OF PROBLEM
= gravimetric sampling sites; 
= real-time sites.