May 18, 2005 - essential oils and water (MEOW) from June 24 to July 21, 2003. Emission measurements in barn. 8 were temporarily discontinued between ...
This is not a peer-reviewed article. Livestock Environment VII, Proceedings of the Seventh International Symposium, 18-20 May 2005 (Beijing, China) Publication Date 18 May 2005, ASAE Publication Number 701P0205. Ed. T. Brown-Brandl. Copyright 2005 American Society of Agricultural Engineers, St. Joseph, Michigan USA.
AIR EMISSIONS FROM TWO SWINE FINISHING BUILDING WITH FLUSHING: AMMONIA CHARACTERISTICS Albert J. Heber, Pei-Chun Tao, Ji-Qin Ni, Teng T. Lim, and Amy M. Schmidt
ABSTRACT The goal of this 11-month study was to evaluate the characteristics of ammonia (NH3) emission during a test of 1) soybean oil sprinkling (SOS), 2) misting of essential oils, and 3) misting of essential oils and water. Measurements were recorded every 60 s from August 2002 to July 2003 at two tunnel-ventilated swine finishing barns that were flushed at least 16 times daily with lagoon effluent. Ammonia concentrations were measured with a chemiluminescence analyzer by time-sharing it between the barns and ambient air. The treated barn with SOS resulted in 40% less NH3 emission than the control barn. The mean (± st. dev.) NH3 concentration and emissions were: 17±8.5 ppm (n=184) and 62±22 g/d-AU (n=175), AU=animal unit=500 kg). KEYWORDS. Air quality, air pollution, air pollutants, pig barn, gas
INTRODUCTION AND OBJECTIVES Long-term measurements of air emissions from swine housing have shown that ammonia (NH3) emissions may be significant, especially at large sites (Ni et al., 2000; Heber et al., 2002). The average daily mean NH3 concentration at a mechanically-ventilated swine barn was 5.6±0.41 ppm (mean ± 95% confidence interval) and ranged from 2.8 to 10.6 ppm, and the mean emission rate was 145±10 g/d-AU (Ni et al., 2000). The authors concluded that barn ventilation significantly impacts NH3 concentration. Pig mass and indoor temperatures are directly related to NH3 emission due to greater manure production and a greater amount of manure degradation byproducts. Tao (2004) and Heber et al. (2004) described the basic results of a study in a swine finishing house using soybean oil sprinkling (two trials), and misting of essential oils with or without water as a carrier. The barns housed 1,100 pigs and were flushed several times daily. Ammonia and several other air pollutants were measured for eleven months. In the test of oil sprinkling, NH3 emission was 19% less in the treated barn. The NH3 emission was 16 to 20% less in the barn treated with essential oil spraying. The mean NH3 concentration and emission were 18 ppm for 132 d and 55 g/d-AU for 125 d (AU=animal unit=500 kg). The objective of this paper is to evaluate the concentration and emissions characteristics of NH3 measured at the two swine finishing barns (Heber et al., 2004).
METHODS AND PROCEDURES The fan-ventilated swine finishing barns (61.0 m x 13.2 m x 2.4 m) had two rows of 24 pens with a center alley and four shallow manure gutters under a slatted concrete floor (fig. 1). Each gutter was flushed four times daily with lagoon effluent. Ventilation air entered through ceiling inlets except during warm weather when it entered through curtains on the east end of the barns for tunnel ventilation. In warm weather, the barns were tunnel ventilated with four 1.22-m diameter belted exhaust fans on the west end wall along with one continuous 0.91-m direct-drive variable speed fan (Heber et al., 2002). The north and south barns were denoted B7 and B8, respectively, with B7 the control barn. New pigs arrived at about 25 kg and were harvested at about 123 kg. Odor, NH3, hydrogen sulfide, non-methane hydrocarbons, and particulate matter were monitored from both barns (Heber et al., 2004). The barns were compared to evaluate effects of abatement 436
methods applied to B8 for comparison with B7. First, a soybean oil sprinkling (SOS) system automatically applied soybean oil daily in the treated barn from August 28, 2002 to February 28, 2003. Second, misting of essential oils (MEO) was conducted in the treated barn (March 5 to April 10, 2003). The essential oils application was later modified by atomizing a mixture of essential oils and water (MEOW) from June 24 to July 21, 2003. Emission measurements in barn 8 were temporarily discontinued between May 2 and June 11, 2003, during which a test of another abatement method was conducted to treat the exhaust air of barn 7. Met Tower
Air Sampling Small Propeller Anemometer
∗
Relative Humidity/Temperature
∗
Lab
Static Pressure Port Barn 7
Temperature Sensor
Barn 8 Fan # 5
Barn 7
Summer Air Inlets
3
∗
Floor plan (61 m x 13.2 m)
1 2 4
S
Background air sampling
Instrument shelter
∗ Attic
Diffusers
Side View
Shallow pit with recycle flush
Pens
∗∗
1-5
Figure 1. Barn monitoring plan (Heber et al., 2002).
Automatic sequential gas sampling from the exhaust of each barn and ambient air at 4 L/min was conducted using filtered and heated Teflon tubes (Heber, et al., 2002). The exhaust sampling location was about 1 m away from the continuous winter ventilation fan (fig. 1). Air was sampled for one sampling period at each gas sampling location (GSL). A 10-min sampling period was used for the exhaust locations until Aug. 28 when it was increased to 60 min. The ambient sampling period was increased from 10 to 20 min on Sept. 23. Ammonia was monitored with a chemiluminescence NH3 analyzer (TEI Model 17C, Thermo Electron Co., Waltham, MA) with a range of 1 ppb to 100 ppm. Its full scale was set between 20 -100 ppm depending on the season. Assuming that NO and NO2 were negligible, the analyzer was operated in the total nitrogen mode. Periodically, the analyzer was calibrated using a programmable diluter and the analyzer’s zero and span response was checked twice a week. Fan status was recorded with propeller anemometers (SPA) (R.M. Young Company, Traverse City, MI) at each fan (Heber, 2003). A portable fan tester was utilized to measure in-field airflow rates of three fans in B7 and three fans in B8. Electronic transmitters (Model HMW61, Vaisala Inc., Woburn, MA) monitored temperature and humidity at barn fan locations. The HUMICAPTM sensor in these units had ±2% accuracy between 0 and 90% RH and ±3% between 90 and 100% RH. Barn static pressure was measured with a -100 to +100 Pa pressure sensor (Model 267MR, Setra Systems, Inc., Boxborough, MA). Wind was monitored with a cup anemometer (R.M. Young) about 10 m above the ground and solar radiation was monitored with a pyranometer (Li-COR). Three activity sensors (Model SRN-2000, Visonic Inc., Bloomfield, CT) monitored pig activity in each barn. Flushing and heating were also monitored (Heber et al, 2002). Mean pig mass was estimated from mean weights and animal numbers provided by the producer. The first several readings during each sampling period were invalid because of time required for the analyzer to equilibrate. To obtain continuous gas data to match continuous airflow, the minute-by-minute gas concentration data in long intervals between valid readings were estimated by linear interpolation between intermittent sampling periods. The anchor points for the interpolations were the mean of the last 10 min of the previous sampling period and the first 10 min of valid data from the next sampling period, each located at the midpoint of the 10-min interval. A “complete day” for a given variable was defined as a day with more than 80% of the data judged to be valid and a “full barn” was defined as a barn with at least 80% of full capacity. 437
Only data collected under the “complete days full barn” (CDFB) condition was analyzed.
RESULTS AND DISCUSSION Barn Environmental Control The basic statistics of several environmental variables are given in table 1 and figs. 2 to 4 show the daily mean temperatures, barn static pressure, airflow rates and relative humidity of exhaust air throughout the test. The ventilation, humidity, and temperature in the barns were quite similar. While the mean ambient temperature and relative humidity was 10°C and 66.1%, respectively, the mean exhaust temperatures in B7 and B8 were 23.2±3.6°C and 22.5±3.6°C, (fig. 2) respectively. The average annual temperature at the site was 10.4oC based on weather data collected from 1971 to 2000 (ggweather.com/normals/mo.htm), thus the test represented typical annual weather. Daily mean barn temperatures ranged from 15 to 30°C and typically exceeded the ambient temperature. Overall mean relative humidities were 54.2±6.1% and 56.3±6.6% in B7 and B8 (fig. 3), respectively. Daily relative humidity exceeded 70% on some occasions in the summer with large pigs and was as low as 40% in the spring with young pigs. The mean static pressures over 252 and 243 d were -16.0 and -19.5 Pa in B7 and B8, respectively. Daily means ranged from about -5 Pa in winter to about -40 Pa in summer (fig. 4). Overall mean airflows were 13.1±10.4 (n=231) and 12.5±11.4 m3/s (n=232) for B7 and B8, respectively. The observed similarities were expected because the barns were identical in construction, environmental control, and management. Each barn was subjected to the same weather disturbances. The barns were usually loaded simultaneously with similar pig numbers and ages. The cumulative frequency distribution of hourly mean ventilation rates (fig. 5) showed that the summer fans moved a relatively small fraction of the 318,045,000 m3 of fresh air that flowed through the barn during the 6661 h represented by the graph. Specifically, the amounts of ventilation air provided by less than 25, 50, 75 and 90% of the maximum barn capacity of 38.3 m3/s were 54.6, 79.1, 93.3 and 98.8% of time, respectively. Half the time, the ventilation rate was 8.6 m3/s (22.5% of total capacity) or less. The barn airflow rate was less than 4.3 m3/s (11.1% of total capacity) for 25% of the time, and less than 18.8 m3/s (49.1% of total capacity) for 75% of the time. These relationships indicate that air treatment systems (e.g. biofilters) should be installed first on the cold weather fans for the greatest cost/benefit ratio. Also, diminishing benefits occur as they are installed on a greater number of fans, which operate less and less. 40 30
o
Temperature, C
20 10 0 -10
Ambient T Barn 7 Barn 8
-20 -30 8/22
9/26
10/31
12/5
1/9
2/13 3/20 Date, mm/dd
4/24
5/29
Figure 2. Daily mean barn and ambient temperatures.
438
7/3
8/7
Exhaust Relative Humidity, %
100 90 80 70 60 50 40 30
Barn 7 Barn 8
20 10 0 8/22
9/26
10/31
12/5
1/9
2/13
3/20
4/24
5/29
7/3
8/7
Date, mm/dd
80
-10
70
-20
60
-30
50
-40
40
3
0
Ventilation Rate, m /s
Static pressure, Pa
Figure 3. Daily mean exhaust relative humidity in barns 7 and 8.
-50
dP B7
dP B8
Airflow B7
30
Airflow B8
-60
20
-70
10
-80 8/22
0 9/26
10/31
12/5
1/9
2/13
3/20
4/24
5/29
7/3
8/7
Date, mm/dd
Figure 4. Daily mean barn static pressures and airflow rates. 50
40
o
20
3
Airflow rate, m /s
Temperature
Temperature, C
30 40
30
10 0
20
-10 10 -20 Airflow 0
-30 0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Fraction of Time
Figure 5. Frequency distribution of hourly mean airflow rates and their corresponding temperatures in B7.
Ammonia Concentrations and Emissions The daily mean NH3 concentrations ranged from 1.2 to 37 ppm (table 1) and averaged 17 and 14 ppm in B7 (n=175) and B8 (n=129), respectively. The overall mean concentration was lower in B8 because it was the treated barn in several tests (Heber et al., 2004). The NH3 level was
439
influenced by ventilation as the highest concentrations were observed in January and the lowest were observed in July (fig. 6). Ventilation was controlled to maintain inside temperature, and higher ventilation in warm weather diluted gas concentrations. The cumulative frequency distribution (fig. 7) indicates that NH3 concentration in B7 was less than 9, 17, 20, 23, 25, and 29 ppm 25%, 50%, 70%, 75%, 83% and 90% of the time. The associated airflow rates show a very strong correlation with concentration (r = -0.78). Since airflow is coupled with temperature, the daily mean concentration also correlated well with daily mean outdoor temperature (r = -0.82) and indoor temperature (r = -0.71). The daily mean concentration was directly proportional to daily mean pig activity as indicated by the correlation coefficient of 0.57 (table 2). Seasonal trends in activity occurred, with higher and lower values in cold and warm weather, respectively. Table 1. Average daily means (St. Dev.) of NH3 and other variables. Parameter Control Barn (7) Treated Barn (8) Barn inventory of pigs 1120 1112 Pig mass, kg 63.5 64.4 88.3 88.9 Stocking density, kg/m2 Animal activity, mV 260 180 (output from activity sensors) Ambient temperature, oC 9.98(11.4) Ambient relative humidity, % 66.1(11.0) Wind speed, m/s 3.68(1.71) 23.2(3.64) 22.5(3.61) Exhaust temperature, oC Exhaust relative humidity, % 54.2(6.10) 56.3(6.57) 13.1(10.4) 12.5(11.4) Ventilation rate, m3/s Static pressure, Pa -16.0(8.58) -19.5(8.85) Test duration, d 290 80% full 252 243 80% valid data 203 151 Complete data full barn (CDFB) 175 125 Concentrations, ppm n=184 n=129 Mean 16.9(8.46) 14.2(8.03) Maximum 36.8 35.9 Minimum 1.75 1.20 Median 17.0 13.4 Emission rates n=175 n=125 Mean, kg/d 8.09(2.50) 6.50(2.61) Maximum, kg/d 14.1 12.0 Minimum, kg/d 1.48 0.63 Median, kg/d 8.15 6.88 62.0(21.6) 41.7(15.7) Mean, g/d⋅AU 121 82 Maximum, g/d⋅AU 11.6 10.1 Minimum, g/d⋅AU 59.0 39.0 Median, g/d⋅AU 7.27(2.39) 6.02(2.52) Mean, g/d⋅pig 13.3 11.3 Maximum, g/d⋅pig 1.29 0.55 Minimum, g/d⋅pig 7.30 6.16 Median, g/d⋅pig
440
300
130 Barn 7
Barn 8 Barn 8 B8 Pig Mass
Emission rate Barn 7 Concentration B7 Pig Mass
120 110
270
Emission Rate, g/d-AU
90
210
80
180
70 150 60 50
120
40
90
30
Concentration, ppm Total Live Mass, AU
240
100
60
20 30
10 0 8/22
0 9/26
10/31
12/5
1/9
2/13
3/20
4/24
5/29
7/3
8/7
Date, mm/dd
Figure 6. Daily mean ammonia concentration and emission rate.
The overall mean emission rates were 8.1 and 6.5 kg/d in B7 (n=175) and B8 (n=125), respectively, but only B7 represents baseline emissions since B8 was abated in various tests. The average daily mean normalized NH3 emission rates were 62 and 42 g/d-AU in B7 and B8, respectively (table 1). The B7 emission rate was correlated to outdoor temperature and ventilation rate with correlation coefficients of 0.23 and 0.26 (table 2). As expected, it was also directly proportional to total live mass (r=0.45). 3
Concentration, ppm; Airflow rate, m /s
50 45 40
Airflow
35
Concentration
30 25 20 15 10 5 0 0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Percent Time Less Than
Figure 7. Frequency distribution of NH3 concentration and its corresponding airflow rate. Table 2. Correlations between barn 7 daily NH3 concentrations and emissions (mg/s) and other variables. Variable Ammonia Concentration Emission rate Indoor temperature -0.71* 0.30* * Outdoor temperature -0.82 0.23* * Pig activity 0.57 -0.27* * Ventilation rate -0.78 0.26* * Total live mass -0.24 0.45* *p
