The Canadian Society for Bioengineering The Canadian society for engineering in agricultural, food, environmental, and biological systems.
La Société Canadienne de Génie Agroalimentaire et de Bioingénierie La société canadienne de génie agroalimentaire, de la bioingénierie et de l’environnement
Paper No. 06-175S 1
2
3
4
5
6
7 8 9
10
11
12 13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
EMISSION FACTOR DEVELOPMENT FOR PARTICULATE MATTER FROM A BROILER HOUSE
Taylor Roumeliotis; Bill Van Heyst School of Engineering, University of Guelph, 50 Stone Rd. Guelph, ON. N1G 2W1
Written for presentation at the CSBE/SCGAB 2006 Annual Conference
Edmonton Alberta July 16 - 19, 2006
Abstract The recent intensification of the livestock industry has raised concerns regarding the air pollution generated from
the various animal housing operations. One pollutant of concern is particulate matter (PM), which is capable of
lodging itself deep in the respiratory tract and causing serious detrimental respiratory effects to the workers and
livestock. Little information is available in the literature that characterizes the emissions of PM from various types
of livestock houses that also incorporate daily and seasonal variances typical of Canadian climates.
A reproducible and reliable methodology for characterizing the emissions of PM has been developed based on
various operational settings in a broiler house. Three DustTrak® aerosol monitors measured PM concentrations for
various particle size ranges and an Electronic Balancing Tool (EBT) with capture hood measured exhaust rates.
Continuous PM monitoring from several bird production cycles has verified the reliability of the methodologies and
established a procedure for characterizing the emissions from the poultry house. The measured results from the
broiler house indicate that the estimates in the literature are not representative of the colder climate in south western
Ontario. Findings also indicate that the bird’s age and their activity levels strongly influence the emissions from the
broiler house suggesting that the use of constant emission rates do not accurately represent the trend in PM
emissions. Emission factors and emission inventories have been developed to describe the variations in daily and
seasonal emissions.
Papers presented before CSBE/SCGAB meetings are considered the property of the Society. In general, the Society reserves the right of first publication of such papers, in complete form; however, CSBE/SCGAB has no objections to publication, in condensed form, with credit to the Society and the author, in other publications prior to use in Society publications. Permission to publish a paper in full may be requested from the CSBE/SCGAB Secretary, PO Box 23101, RPO McGillivray, Winnipeg MB R3T 5S3 or contact [email protected]. The Society is not responsible for statements or opinions advanced in papers or discussions at its meetings.
INTRODUCTION 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
The recent intensification of livestock housing operations in Canada has led to the initiation of
concentrated animal feeding operations (CAFOs). CAFOs provide automated maintenance of the
facility, which allows livestock operators to output considerably more product from one building.
However, with this increase in production comes a decrease in indoor air quality. Of growing
concern is the animal housing operations contribution to the local air quality in the form of
particulate matter (PM). PM produced from livestock housing operations can contain viable
particles (bioaerosols), such as pathogenic bacteria, viruses, and endotoxins, which are capable of
lodging deep within the lungs and have the potential to cause serious detrimental respiratory
effects (Pillai and Ricke 2002). It has been established that bioaerosols can survive for a
considerable length of time and distance, which raises concerns for the quantity of PM emissions
from livestock facilities.
It is believed that poultry operations produce the highest concentrations of bioaerosols of all
livestock housing operations (Pallai and Ricke 2002; CAC 2000). The amount of PM emitted
from poultry houses is dependent on several operational and maintenance conditions. In an
attempt to characterize the emissions from several poultry operations, averaged yearly emission
factors for some of the common size classifications have been developed based on limited
experimental data from various geographic regions (Van Heyst 2005; Lacey et al. 2002; Takai et
al. 1998; Wathes et al. 1998). However, variations in emissions exist from a range of seasonal
operating conditions and animal activity levels which may bias the emission factors. Since the
yearly emission factors were averaged from limited data, their use in accurately representing a
livestock emission inventory outside of Canada is questionable. Also, the collected data from
several geographic regions likely do not accurately represent the unique Canadian climate. Since
little information is available for PM and no threshold levels have been established, experiments
1
are needed to assess the current emission levels in order to understand the future benefits of
introducing best management practices and/or PM controls in the agricultural sector.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Study Objectives
One aspect of livestock emissions that is not well characterized is the lack of PM size
classifications data. In order to advance the current knowledge regarding livestock emissions, an
intensive study at a broiler house was initiated to develop consistent emission factors by
continuously monitoring PM emissions for three size classifications.
The objectives of this study were threefold:
1. Standardization of instrumentation and measurement techniques for the characterization of
airborne particulate matter emissions from a livestock building. This objective
encompasses the measurements of both the airborne PM concentrations and the volumetric
exhaust rates from the livestock building;
2. Quantification of emissions rates characterized by the changes in daily and seasonal activity
in a livestock house; and
3. Development of PM emission factors from a livestock house which can be used to estimate
an emission inventory for a portion of the Canadian agricultural sector. Emission factors
will be developed for several common size classifications: specifically PM10, PM2.5, and
PM1.
If rational and fair standards for environmental air quality are going to be introduced for the
agricultural sector, there must first be improvements on data quantity and quality. The
completion of these objectives should provide the first steps required to initiate the development
of appropriate emission inventories for the poultry industry.
2
MATERIALS AND METHODS 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Since an overwhelming quantity of data is required to describe the emissions from the entire
agricultural industry, it was decided to conduct the initial experiments at a broiler house with the
anticipation that this methodology could be transferred to other livestock housing operations at a
future date. The selected facility was a commercial broiler house constructed in 2003, that is
located off of county road 32 at the Southwestern corner of Guelph, Ontario, Canada. The broiler
house is a single storey, litter floor facility that is 150.5 m long by 18.75 m wide by 3.26 m tall
(refer to Fig. 1 for the broiler house schematic). The house was carefully selected to be
mechanically ventilated with no natural ventilation. Twenty-eight ducts with exhaust fans run
along the length of the southern side of the building. There are 10 fans with a diameter of 1.3 m,
12 fans with a diameter of 0.65 m, and 6 fans with a diameter of 0.56 m. Of these 28 fans, four
of the smallest fans are variable speed fans, while the remainder are single speed fans with on/off
controls. Automated adjustable louvers are located along the northern side of the house.
Ventilation rates and three pipe heaters control the internal temperature of the house.
Temperature is maintained at 33.5oC and reduced by an average of 0.25oC/day until the
temperature reaches 21.5oC. Feed and water supply to the birds is also fully automated.
This commercial house can accommodate 31,000 to 37,000 birds during one production cycle.
The average stocking density of birds in the house is 11.5 birds/m2 with an average residence
time of approximately 45 days (6½ weeks). The broilers are free to roam throughout the entire
house for the entire duration of the production cycle. During the 6½ weeks, the birds will mature
from 50 g to an average weight of 2.5 kg. After each production cycle, a 2 week period is used to
clean the house for biosecurity reasons and reapply a new litter floor. During this time, the
louvers are fully opened to replenish the house with clean ambient air.
3
Instrumentation 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Indoor PM concentrations were measured with optical particle counting instruments, namely
three DustTrak® aerosol monitors. Three locations along the length of the house were used to
verify the uniformity of PM concentrations throughout the house. It was established that PM
concentrations are relatively uniform along the length of the house and hence, measurements
were taken 25 m from the control room to minimize broiler disturbances. Measurements were
taken from an environmental enclosure with a protruding multipoint sampling line. An
adjustable stand allowed the sample inlet to be reposition to different heights. At all three
locations along the length of the building, the multipoint sample inlet was located 2.5 m from the
exhaust side of the broiler house. It was assumed that the concentration measured at this distance
would characterize the concentrations being exhausted from the fans.
A PIXE cascade impactor (PCI) was also used to verify the size fraction of the PM. The PCI
provided the precise particle size distribution in the broiler house. The impactor consisted of 10
stages including an after-filter (AF). The PCI separates particles into the following size
fractionations: greater than 16, 16-8, 8-4, 4-2, 2-1, 1-0.5, 0.5-0.25, 0.25-0.12, and 0.12-0.06 µm
(PIXE International Corporation 2003). With these size fractionations, the relative mass fractions
of each of the common size classifications could be used to evaluate the size fractionations from
the particle counting instruments by manipulating the results from the PCI.
The Alnor® balometer kit, consisting of an electronic balancing tool (EBT) and capture hood,
was used to measure the exhaust rates from the exhaust ducts. The instrument’s capture hood
was connected directly to an exhaust duct, downstream of the fans. A customized skirt isolated
the flow sensors from influences on flow rate measurements from the surrounding air currents.
Pressure differentials were measured across a 16 point velocity matrix to determine the average
velocity over a given area. A complete range of flow rates from each fan type were fully
4
quantified during the bird production cycles. The broiler house’s control room dictated which
fans were used at any given time, and this information was used in conjunction with the
balometer’s measurements for each specific fan to estimate the overall ventilation rate for the
house.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Estimates of PM Emissions
Sampling was conducted throughout four complete bird production cycles from October 28th,
2005 to June 23rd, 2006. The four production cycles characterized the environmental conditions
typically encountered in southern Ontario. The instruments were programmed to collect running
averages of PM concentrations and flow rates every 10 minutes. Hence, an estimate of the PM
emissions could be calculated each 10 minutes for the duration of each bird production cycle.
The emissions were estimated with the equation:
CQE vent= (1)
Where E is the emission rate of PM (g/hr), Qvent is the total house ventilation rate from all the
exhaust ducts (m3/hr), and C is the mean concentration of PM (g/m3). The emission factors
developed from this study were expressed in units of kg of PM per 1000 birds from a single
production cycle.
RESULTS
Ventilation Rates
The rate at which air was exhausted from the house varied with the age of the birds and
ambient temperatures. During the initial days of a production cycle, the four small variable speed
fans were the only fans used to ventilate the house. Their speed was managed to preserve the
5
internal temperature, which was highest at the beginning of a production cycle and consequently
ventilation rates were lowest to prevent cool incoming air from reducing the internal temperature.
As the birds aged, the internal temperature setting was gradually decreased and concurrently, the
bird weight increased, resulting in the production of more body heat. Consequently, higher
ventilation rates were needed to cool the internal temperature. The four variable speed fans were
used until they reach their maximum ventilation rate. After this time, additional on/off fans were
used to continue increasing the house ventilation rate. The computer control panel for the house
dictated the number of fans used as well as the rate of the variable speed fans. Since the control
panel records the percentage of the house’s ventilation capacity, it was possible to estimate the
house ventilation if the flow rate for each size of fan was fully quantified and the number of fans
used was correlated to a particular ventilation percentage. With this data, an accurate real time
house ventilation rate was predicted for use in emission estimates. Hence, the balometer was
used to estimate the range of flow rates for the variable speed fans and a mean ventilation rate for
the on/off fans. Two small variable speed fans were quantified; one fan was used continually
through a production cycle while the other was used less frequently. The measurements for the
entire range of flow rates from both of these small fans are shown in Fig. 2.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
The medium and large sized fans were quantified in a similar manner. The use of the larger
fans varied greatly since their operation was dependent on the temperature difference between the
inside and outside the house. During the winter months, several medium and large sized fans
were not used at all. Moreover, even in the summer months, several fans were only used in the
latter stages of a production cycle. For this reason, the flow rates from three medium sized fans
and two large exhaust fans were measured to quantify any differences in flow rates due to fan
wear. Figures 3 and 4 give the flow rate measurements as well as the calculated mean flow rate
for the medium and large exhaust fans, respectively.
6
It was evident from the flow measurements from several fans that there is some extent of wear
on the fan: flow rates were reduced up to 17% for frequently used fans compared to infrequently
used fans. For this reason, the correlation between the percentage of house ventilation capacity
and the actual flow rate from the broiler house is not linear. Since there is greater wear on
frequently used fans, their flow rate will be overestimated by the control panel. Figure 5 displays
the correlation between the percentage of house ventilation capacity and the estimated flow rate.
The correlation was determined to be:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
13.37372.5152.0 2 ++= ppvent VVQ (2)
Where Vp is the percentage of house ventilation (%). This correlation was used for estimating
Qvent with Eq. 1 for real time emission estimates. The correlation between the percentage of
house ventilation capacity and estimated flow rate has a coefficient of determination (R2) of
0.976. Once the correlation was established, the total flow rate for three bird production cycles
was quantified. Figure 6 displays the complete real time ventilation rates for three bird
production cycles. Since the ventilation rates are dependent on the difference in temperature
inside and outside the house, the ventilation is highest in the summer and lowest in the winter.
Particle Concentrations
Diurnal patterns were apparent during all seasons. There was a prompt reduction in airborne
particulates during the night followed by a rapid increase early in the morning. Due to the
promptness of the changing PM concentrations, it was established that this pattern was not a
result of environmental deviations but rather a result of the change in bird activity from the
lighting schedule. The house control panel switched the lights off for one hour per day for the
first five days and roughly the last 21 days of a production cycle while six hours per day was
7
delegated for the days in between. Since the diurnal pattern was relatively brief during the first
five days and last 21 days, it was confirmed that the diurnal variations in PM concentrations were
connected to the lighting schedule. Figure 7 displays two plots of daily concentrations that
emphasize the changes in the diurnal patterns for one hour and six hours of dark.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
From these plots, it is clear that diurnal patterns are strongly correlated to the lighting
schedule. Environmental parameters may cause the PM concentration in the broiler house to
decrease at night slightly as evident from the 1 hour of dark plot but its effect is minimal in
comparison to effects from lighting. This diurnal trend must be characterized in the emission
factors if they are going to accurately describe the behaviour in emissions from the broiler house.
In all four production cycles, airborne PM concentrations increased as the birds aged. The
initial concentrations observed during all seasons were consistent at roughly 0.25, 0.10, and 0.08
mg/m3 for PM10, PM2.5, and PM1, respectively. The concentrations of all particle size ranges
increased steadily for the first 21 days of the production cycle. After this time, the respective PM
concentrations increased up to their peak concentration. This concentration was essentially
sustained for the remainder of the production cycle. However, there was a significant
discrepancy between the winter and summer concentrations especially towards the end of the
production cycle. This was likely a result of the increased ventilation rates in the summer that
exhausted more PM from the broiler house. PM10 concentrations for winter and summer
production cycles are displayed in Fig. 8.
In order to verify the accuracy of the particle counter instruments, the PCI was used to confirm
the precise particle size distribution. The results from the PCI were used in combination with the
particle counter instrumentation to estimate the contribution of PM2.5 and PM1 from the PM10
emission estimates during the summer production cycle. Three PCI tests were performed during
two bird production cycles. The results from each of the tests were consistent with the
8
measurements obtained from the DustTrak aerosol monitor. Figure 9 displays the particle size
distribution represented as a cumulative percent less than the particle size (d
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
p) obtained from the
three PCI results as well as the DustTrak measurements during the same sampling period. Since
the DustTrak’s do not measure the total suspended particulates (TSP), PM10 is normalized to the
mean PCI results for 10 µm particles using the equation in Fig. 9. From the results, it was
established that the amount of PM2.5 relative to PM10 was approximately 56.3% on a mass basis.
Similarly, the amount of PM1 relative to PM10 was approximately 36.7% on a mass basis. With
this mass fractionation, it was possible to approximate all the common size classifications from a
single size range.
PM Emission Rates and Emission Factors
The real time concentrations from each bird production cycle were suitable measurements for
the concentration values (C) in Eq. 1. These real time measurements were combined with the
estimates of total house ventilation (Qvent) to determine the real time estimates of PM emissions
for the various common particle size classifications.
The PM emission rates were estimated from three production cycles. The winter and spring
emissions were estimated from measurements from the DustTrak aerosol monitor for all three
size classifications. The only measured size classification for summer emissions was PM10 and
as such, the mass fractionations established from the PCI were used to estimate the remaining
size classification.
Emission rates for all seasons follow similar patterns throughout the bird production cycle.
During all bird production cycles, PM emissions increased as the bird’s aged. This was result of
an increase in both the PM concentrations and ventilation rates as the bird’s aged. Furthermore,
emission rates were far greater in the day compared to the night because of increased bird
9
activity. Since the PM emissions are interrelated to the indoor PM concentrations, they both
follow a comparable diurnal pattern. Figure 10 displays the PM
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
10 emissions from three different
seasons.
From Fig. 10, it was evident that PM10 emission rates were generally the same for all seasons
throughout the entire bird production cycle. In some instances, summer emissions were
occasionally higher at the end of the production cycle due to the extreme ventilation rates used
during this period. Since there was a large diurnal variation in emissions, the developed emission
factors must reflect this trend. The diurnal emission factors were generated from a six hour
lighting schedule but are reported on an hourly basis which will allow them to alter the total
particulate mass emitted per production cycle for any lighting schedule at any broiler house. The
seasonal and diurnal emission factors for all common particulate size classifications are reported
in Tables 1 and 2. The emission factors were developed from the number of birds and length of
time for each specific production cycle as a total mass of particulates per production cycle per
1000 birds: an average of 33,828 birds were raised for 47 days in the winter, 35,537 birds were
raised for 45 days in the spring, and 36,158 birds were raised for 44 days in the summer. The
average emission factors were determined to be 0.137, 0.190, and 0.515 kgPM 1000 birds-1
production cycle-1 for PM1, PM2.5, and PM10, respectively.
These emission factors provide a means to estimate an emission inventory of PM for an entire
year of broiler production at a single commercial facility. Based on the length of the studied
production cycles, this broiler house should be able to output 6 batches of broilers per year. The
average number of birds in a production cycle was 33,925 birds. From these averages, the
emission inventory of PM for a single facility would be 27.9, 38.8, and 104.8 kgPM/year for PM1,
PM2.5, and PM10, respectively.
10
CONCLUSIONS 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
An emission inventory of PM for a single commercial broiler house was characterized. The
methodology for estimating PM emissions involved obtaining measurements of continuous real
time indoor PM concentrations from optical particle counting instruments located near the inlet of
the broiler house’s exhaust fan as well as real time ventilation estimates calculated from flow
instrumentation and a capture hood. This methodology was applied to four bird production
cycles, which provided estimates of emission rates that were representative of a southern Ontario
climate during three seasons. It was established that seasonal emissions only slightly varied
towards the end of a bird production cycle. Increasing emission rates with bird age were
observed during all bird production cycles. From these emission rates, averaged PM emission
factors were developed and reported as the mass of PM emitted per 1000 birds per production
cycle. The emission factors were found to be 0.137, 0.190, and 0.515 kgPM 1000birds-1
production cycle-1 for PM1, PM2.5, and PM10, respectively. Based on these emission factors and
the average number of broilers produced per year, it was possible to generate a complete
emission inventory of PM for a single commercial broiler house. The emission inventory was
27.9, 38.8, and 104.8 kgPM/year for PM1, PM2.5, and PM10, respectively.
REFERENCES
Canadian Agricultural Census (CAC). 2000. Area source agriculture: section O5: agriculture
(animals). Not published.
Lacey, R.E., J.S. Redwine, and C. B. Jr. Parnell. 2002. Emission factors for broiler production
operations: a stochastic modeling approach. The Society for Engineering in Agricultural, Food,
and Biological Systems. Paper #: 024212. Presented at 2002 ASAE Annual International
Meeting.
11
1
2
3
4
5
Pillai, S.D. and S.C. Ricke. 2002. Bioaerosols from municipal and animal wastes: background
and contemporary issues. Canadian Journal of Microbiology 48: 681-696.
PIXE International Corporation. 2003. Inertial Impactor Models I-L and I-1L. Tallahassee, FL.
http://pixeintl.com 6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Takai, H., S. Pedersen, J.O. Johnsen, J.H.M. Metz, P.W.G. Groot Koerkamp, G.H. Uenk, V.R.
Phillips, M.R. Holden, R.W. Sneath, J.L. Short, R.P. White, J. Hartung, J. Seedorf, M. Schroder,
K.H. Linkert, and C.M. Wathes. 1998. Concentrations and emissions of airborne dust in livestock
building in northern Europe. Journal of Agricultural Engineering Resources 70: 59-77.
Van Heyst, B.J. 2005. Final report: Evaluation of emission factors for the improvement of the
estimation methodology for particulate matter from agricultural poultry industry. University of
Guelph. Contract: K2361-04-0116. Not published.
Wathes, C.M., V.R. Phillips, M.R. Holden, R.W. Sneath, J.L. Short, R.P. White, J. Hartung, J.
Seedorf, M. Schroder, K.H. Linkert, S. Pedersen, H. Takai, J.O. Johnsen, P.W.G. Groot
Koerkamp, G.H. Uenk, J.H.M. Metz, T. Hinz, V. Caspary, and S. Linke. 1998. Emissions of
aerial pollutants in livestock building in northern Europe: overview of a multinational project.
Journal of Agricultural Engineering Resources 70: 3-9.
ACKNOWLEDGEMENTS
This work was assisted by funds provided from the Poultry Industry Council. The authors
would also like to thank the support from the School of Engineering at the University of Guelph.
12
FIGURES 1
2
3
4
5
Figure 1 – A schematic of the commercial broiler house used in this study, showing the location
of the exhaust ducts and control room
6
7
8
Figure 2 – Complete range of flow rates for the small variable speed fans
13
1
2
3
Figure 3 – Measured flow rates from three medium sized exhaust fans
4
5
6
Figure 4 – Measured flow rates from two large sized exhaust fans
14
1
2
3
4
Figure 5 – Correlation describing the complete range of house ventilation capacity as a
percentage to the exhaust flow rate
5
6
7
Figure 6 – Estimated flow rate for three bird production cycles during three different seasons
15
1
2
3
4
Figure 7 – A comparison of the effects of the two lighting schedules (1 hour and 6 hours of dark)
used at the broiler house on PM10 concentration
5
6
7
Figure 8 – Complete measurements of winter and summer indoor PM10 concentrations
16
1
2
3
Figure 9 – Mass fractions of PM measured from the DustTrak aerosol monitor and the PCI
4
5
6
7
Figure 10 – PM10 Emissions estimated for the duration of three entire bird production cycles
during three seasons
17
TABLES 1
2
3
Table 1 – Seasonal emission factors reported as total mass of particulates emitted per production
cycle per 1000 birds
Emission Factors PM1 PM2.5 PM10Season
kgPM 1000 birds-1 production cycle-1
kgPM 1000 birds-1 production cycle-1
kgPM 1000 birds-1 production cycle-1
Winter 0.117 0.149 0.443 Spring 0.095 0.117 0.560
Summer 0.199 0.305 0.542 4
5
6
Table 2 – Hourly emission factors estimated during the hours of light and dark in the broiler
house
Hourly Emission Factors PM1 PM2.5 PM10Averages from all
seasons kgPM hr-1 1000 birds-1
kgPM hr-1 1000 birds-1
kgPM hr-1 1000 birds-1
Lights off 7.36x10-6 9.29x10-6 39.7x10-6
Lights On 0.102x10-3 0.127x10-3 0.537x10-3
7
18
Top Related