- Email: [email protected]
Measurements of emission factors from a naturally ventilated commercial barn for dairy cows in a cold climate
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/issn/15375110
Research Paper
Measurements of emission factors from a naturally ventilated commercial barn for dairy cows in a cold climate Ngwa M. Ngwabie a,*, Andrew Vanderzaag b, Susantha Jayasundara c, Claudia Wagner-Riddle c a
Department of Environmental Science, Faculty of Science, University of Buea, Box 63, Buea, South West Region, Cameroon b Agriculture and Agri-Food Canada, Ottawa, Ontario, Canada c School of Environmental Science, Ontario Agricultural College, University of Guelph, Guelph, Ontario N1G 2W1, Canada
article info
Emission rates of CH4, N2O and NH3 were measured in a commercial free-stall barn that
Article history:
housed 141 lactating dairy cows, and 75 dry cows and replacement heifers. Animal activity,
Received 3 February 2014
measured using the ALPRO™ dairy herd management system was used together with the
Received in revised form
CO2 balance method to calculate the ventilation rate. Methane emission was also modelled
16 August 2014
using the IPCC Tier 2 method. Animal activity variations similar to reported patterns
Accepted 26 August 2014
indicated that the activity monitoring system provided high resolution measurements
Published online 16 September 2014
since all cows were considered. Diurnal variations were observed in the emissions with
Keywords: Animal activity Emission factor Dairy cows Greenhouse gases Enteric fermentation
1.
mean values of 12.2e13.9 g CH4 LU1 h1, 0.43e0.64 g NH3 LU1 h1 and 29.4 e41.3 mg N2O LU1 h1. Modelled enteric CH4 emission was 312 g CH4 head1 d1 (10.58 g CH4 LU1 h1). It was estimated that indoor manure emitted 73 g CH4 head1 d1 (2.5 g CH4 LU1 h1), with enteric fermentation representing 81% of the total barn CH4 emission. Lactating cows emitted about 363 g CH4 head1 d1 (11.42 g CH4 LU1 h1) while non-lactating cows emitted 241 g CH4 head1 d1 (9.67 g CH4 LU1 h1). Crown Copyright © 2014 Published by Elsevier Ltd on behalf of IAgrE. All rights reserved.
Introduction
The agricultural sector is a significant contributor to greenhouse gas (GHG) emissions, accounting for 8% of the total 2010 GHG emissions in Canada, indicating a 19% increase from the 1990 level (Environment-Canada, 2012). In particular, agriculture accounts for 24% and 72% of the Canadian CH4 and N2O
emissions, respectively, with the main sources coming from cattle and pig raising as well as the use of synthetic nitrogen fertilisers. In total, Canadian livestock contribute about 60% of the total agricultural GHG emissions. Increasing trends in estimated national emission factors have been observed in some livestock categories, e.g. CH4 emission from enteric fermentation in dairy cows increased
* Corresponding author. Tel.: þ237 71643209. E-mail address: [email protected] (N.M. Ngwabie). http://dx.doi.org/10.1016/j.biosystemseng.2014.08.016 1537-5110/Crown Copyright © 2014 Published by Elsevier Ltd on behalf of IAgrE. All rights reserved.
104
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Nomenclature ADF AU CF
acid detergent fibre animal unit (500 kg animal mass) correction factor for the heat produced at any temperature Ti ( C) CO2indoors CO2 indoor concentration (ppmv) CO2outdoors CO2 outdoor concentration (ppmv) CP crude protein DMI dry matter intake (kg head1 d1) E average metabolisable feed energy content (MJ kg1 dry matter) EE ether extract G daily gain in weight of a heifer (kg d1) GEI gross energy intake (MJ1 head1 d1) GHG greenhouse gas HPU heat produced unit IPCC Intergovernmental Panel on Climate Change LU livestock unit (500 kg animal mass) M average animal mass (kg) NDF neutral detergent fibre NE no emissions assumed NFC non-fibre carbohydrate (g kg1) PTFE polytetrafluoroethylene RA relative animal activity measured temperature ( C) Ti TDN total digestible nutrients (%) ventilation rate per heat producing unit VRHPU (m3 h1 HPU1) Y average daily milk production (kg d1) methane conversion factor Ym total heat produced by dairy cows and heifers Ftot (W) Fdairy cows heat produced by dairy cows (W) Fheifers heat produced by heifers (W) from 109.4 kg head1 yr1 in 1990 to 127.1 kg head1 yr1 in 2010, with an associated increase in milk production (Environment-Canada, 2012). However, significant differences still exist between emission factors at the barn, provincial and national levels due to the simplified approach needed for national scale inventories and the diversified nature of dairy management and climatic conditions (Environment-Canada, 2012; IPCC, 2006). Canadian dairy cows are mostly kept indoors in naturally ventilated barns but in some cases they are sent outdoor for a few hours during the day in warm weather (Sheppard, Bittman, Swift, Beaulieu, & Sheppard, 2011). On average, Canadian dairy cattle manure storage is evenly distributed among solid and liquid forms (~40% each), with ~20% being deposited on pastures; however in certain provinces, the proportion of dairy manure handled as liquid can be as high as 89% or as low as 20% (Environment-Canada, 2012; Statistics-Canada, 2003). Although several measurements have been conducted to quantify and study the variation patterns of NH3 and GHG emissions from dairy cow barns in the cold regions of North America and Europe (Harper et al., 2009; Ngwabie, Jeppsson, Gustafsson, & Nimmermark, 2011; Zhu, Dong, & Zhou, 2012), only a few measurements have been conducted in Canada (Bluteau, Masse, & Leduc, 2009;
McGinn & Beauchemin, 2012; McGinn, Flesch, Harper, & Beauchemin, 2006). More measurements are therefore needed to fully understand and quantify emissions from commercial dairy cow barns in Canada. However, direct measurements can be expensive and the application of models (IPCC, 2006; Li et al., 2012) to estimate emission factors may provide a more reliable and cheaper alternative especially when input data from specific barns is used. Estimates from such models for Canadian dairy cows at the provincial level have carried out (Jayasundara & Wagner-Riddle, 2014) and they need to be validated with direct measurements. The determination of emission factors from livestock buildings requires measurements of gas concentrations and ventilation rates. While several techniques and instruments to measure gas concentrations in livestock buildings have been used and extensively reviewed (Ni & Heber, 2008; Ni et al., 2009; Ogink, Mosquera, Calvet, & Zhang, 2013; Wheeler, Weiss, & Weidenboerner, 2000), measurements of ventilation rates, especially in naturally ventilated buildings (popular in the dairy industry) remains highly difficult. The CO2 balance method is recommended for ventilation rate determination in naturally ventilated buildings but it has a temporal resolution of 24 h, which can be improved if the animal activity is known (CIGR, 2002). Several methods have been applied to obtain values for the animal activity: modelling (CIGR, 2002; Cornou & Lundbye-Christensen, 2012), infrared detectors (Pedersen & Pedersen, 1995) and cameras (Costa, Borgonovo, Leroy, Berckmans, & Guarino, 2009). There is a need for easily used methods that can also improve the spatial resolution of the measured activity since most detectors and sensors have a limited field of view. A system that is increasingly used in highly automated barns is the ALPRO™ herd management system which is incorporated into the DeLaval system (Tumba, Sweden). Amongst its features is an activity monitoring system that is used to determine when animals are in oestrus so as to identify optimal insemination times and for monitoring animal health deviations. It is a promising tool for emission research which may provide high spatial resolution data for animal activity to be used in ventilation rate determination by the CO2 balance method. This is because it measures the activity of each animal that is fitted with a collar. Given the considerations mentioned above, measurements were conducted in a naturally ventilated barn for dairy cows during the spring and the fall transitional seasons with the aim of (a) studying diurnal variations (variations within a day) in the emissions the CH4, N2O and NH3; (b) determining the emission factors of these gases through direct measurements; and (c) comparing CH4 emission factors obtained through direct measurements and through modelling using the IPCC (2006) Tier 2 method with local barn data.
2.
Materials and methods
2.1.
Dairy cow barn
Measurements were carried out during the transitional seasons in the spring (FebruaryeApril) and in the fall (September and October) of 2012 in a dairy cow barn located
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
near Drayton (Latitude 43 450 18.2200 N and Longitude 80 400 16.6000 W) in Southern Ontario, Canada (Fig. 1). The barn had an automatic milking system with open access to the milking machine. The barn was divided into five sections based on the animal category. The main barn had sections for high milk producing cows, low milk producing cows and for replacement heifers. A corridor connected the main barn to a side barn. The corridor housed transition cows while the side barn contained dry cows. All sections in the barn had cubicles on an elevated platform where the cows rested. The bedding material had a rubber mat base which was covered with a layer of straw during the spring measurements. During the autumn measurements, the rubber mat was covered with a layer of solid digestate as bedding material. The solid digestate was obtained by mechanically compressing digestate obtained from a biodigester which became operational during the autumn. The main barn had a solid concrete floor which was mechanically scraped six times a day (12 a.m., 4 a.m., 8 a.m., 12 p.m., 4 p.m., 7 p.m.), with each cleaning session lasting 1 h. The slurry was dumped into a covered pit at one end of the barn. The pit had an outdoor extension from which the manure held for 2e3 weeks was pumped into a nearby storage tank in the spring or into a biodigester in the fall. The dry and transition cows were on slated floors. The slats were over a manure channel with an outdoor extension for temporary manure storage before being pumped into the storage tank or the biodigester. The barn was naturally ventilated through automatically regulated curtains on the side walls. Additional ventilation during warm weather was achieved by opening the doors. Ceiling fans in the barn were not operated during the measurement period.
2.2.
105
Production and feed information
The barn had Holstein dairy cows with production information shown in Table 1. There was no fixed feeding schedule with feed delivered 1e2 times per day for the milking cows and every other day for the heifers. The dry cows were fed every 2e3 days. The feed was a mixed ration with composition shown in Table 1. The animal mass was obtained from production sheets provided by management.
2.3.
Gas concentrations and climate data
The concentrations of CH4, NH3, N2O, and CO2 were measured using a photo-acoustic multi-gas analyser 1309 (Lumasense Technologies SA, Ballerup, Denmark). The analyser was calibrated by the manufacturer prior to installation in the barn. The detection thresholds of the gases were: 0.2 ppm NH3, 0.03 ppm N2O, 0.4 ppm CH4 and 1.5 ppm CO2, with an accuracy of ±2e3%. Gas concentrations were measured at 3 indoor locations and 2 outdoor locations. The indoor sampling locations (L3eL5 in Fig. 1) at a height of 2 m above the floor were chosen to cover the different groups of cows. The hourly means of all measured values at all the indoor locations were used as the indoor concentrations. The outdoor locations were chosen along either side of the barn at a height of about 2 m above the ground. The background concentration for each hour was considered to be the lower of the two values from both outdoor locations as it was most likely a better representation of the air entering the barn due to changing wind directions. Air was sampled from the various locations using 10e30 m long polytetrafluoroethylene (PTFE) tubes with an inner diameter of 3.2 mm. Air was
Fig. 1 e Layout of the dairy cow barn where measurements were conducted (diagrams not to scale).
106
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Table 1 e Production information and composition in rations fed to dairy cows during the period when emission measurements were conducted. Production information and ingredient in the feed
High producing cows
Low producing cows
Transition cowsa
Dry cowsb
Breeding heifersc
% of Dry matter Production information Animal number Animal mass, kg Milk yield, kg head1 d1
77 (66)d 650 32
Feedstuff in the ration (% of dry matter) Corn silage Alfalfa-grass silage Grass hay Straw High moisture corn Custom supplemente Minerals
39.4 32.2 2.0 2.1 10.3 14.1 0
45.1 39.0 2.2 2.3 0 11.4 0
23.6 18.2 0 20.5 4.9 32.8 0
40.4 32.3 16.2 0 0 10.1 1.0
41.9 36.4 0 13.7 0 6.8 1.2
Forage % Concentrate %
75.7 24.3
88.6 11.4
62.3 37.7
88.9 11.1
92.0 8.0
a b c d e
55 700 19
14 600 25, (28)d
30 700 e
45 400 e
Up to three weeks after calving. 3 Months to calving. 11 Months to first insemination. Data during the fall measurement period. The average weight of the cattle was obtained from management reports. Purchased from a commercial feed supplier (contained 34% crude protein, 6% fat, 4.5% crude fibre, minerals and vitamins).
continuously drawn from all five locations by a vacuum pump (KNF Neuberger Inc., Trenton, NJ, USA) through an 8valve manifold (Campbell Scientific, Inc., Logan, UT, USA) at a collection rate of about 10 l min1. The valve manifold was controlled by a CR-1000 datalogger (Campbell Scientific, Inc., Logan, UT, USA), which was programmed to sequentially open one valve for the acoustic gas analyser to subsample. Air from locations L3 and L4 (Fig. 1), which covered the main area of the barn with the highest cow density (170 of the total 215 cows) was sampled for 15 min each. Air was sampled for 10 min from location L5 in the side barn which had 45 cows and for 10 min from each of the two outdoor locations. Gas concentrations for the first 4 min of sampling in each location were discarded to account for a complete flushing of the acoustic analyser when the valve changed from one location to the next. The PTFE tube inlets were fitted with 1 mm pore size membrane filters (Millipore Corporation, Billerica, MA, USA) to trap dust particles. The average air temperature and relative humidity were measured every 5 min using Tiny-Tag data loggers with 10K NTC thermistor and capacitive sensors (Gemini Data Loggers, Chichester, UK). The sensors were calibrated by the manufacturer before deployment. Three data loggers were placed close to the gas sampling locations inside the barn, and two data loggers were placed outside the barn (Fig. 1). A cup anemometer (F460, Climatronics Corp., Newton, PA, USA) and a wind vane (R.M. Young, Model 05102, Traverse, MI, USA) that measured wind speed and direction, respectively, where mounted on a tower 100 m away from the barn and at a height of about 4 m from the ground. The wind data was recorded at 1 min intervals on a 21X datalogger (Campbell Scientific, Inc., Logan, UT, USA). Hourly mean values of the climate data were used in the analyses.
2.4.
Ventilation and emission rates
The ventilation rate in animal barns was estimated by calculating the mass balance of a gas with a known source such as CO2. The CO2 source strength was estimated through empirical equations based on production level and animal characteristics (CIGR, 2002). This approaches assumes CO2 is produced from the animals with negligible production from manure or other sources, no consumption of CO2 within the barn, ideal mixing of the air within the barn and a presentative sampling of CO2 concentration. Given these assumptions, the difference between indoor and outdoor concentrations was related to its production rate and hence barn ventilation rate (Blanes & Pedersen, 2005; Ni, Vinckier, Hendriks, & Coenegrachts, 1999; Pedersen et al., 2008): VR ¼ VRHPU Ftot CF
(1)
where VRHPU is the ventilation rate per heat producing unit (m3 h1 HPU1), Ftot is the heat produced by the cows and heifers (W), and CF is a correction factor for the heat produced at any temperature Ti ( C). CF ¼ 1 þ 0:00004ð20 Ti Þ3 VRHPU ¼
0:21ðRAÞ ðCO2 indoors CO2 outdoors Þ106
(2)
(3)
In Eq. (3), 0.21 is the CO2 production (m3 h1 HPU1) and 1 HPU is 1000 W of heat produced by the animals at 20 C, RA is the hourly relative animal activity measured with an activity monitoring system (see below), CO2indoors is the indoor CO2 concentration (ppmv), and CO2outdoors is the outdoor CO2 concentration (ppmv). The total heat (Ftot ¼ Fdairy
107
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
þ Fheifers) produced by the dairy cows and heifers was calculated using CIGR (2002) equations: (4)
or decreased animal movement in the barn. Information regarding the schedule of management routines that influenced animal movement was equally obtained from the farmer.
(5)
2.6.
cows
Fdairy cow ¼ 5:6m0:27 þ 22Y Fheifer ¼ 7:64m0:69 þ G
23 57:27 þ 0:302m 1 E 1 0:171G
Modelling methane emissions
where m (kg) is the average mass of a dairy cow or heifer, Y (kg d1) is the average daily milk production, G (kg d1) is the daily gain in weight of a heifer (0.55 kg d1 measured in this barn) and E is the average metabolisable energy content of the feed (11 MJ kg1 dry matter), which was assumed to be constant. Eqs. (4) and (5) were calculated for each animal in a category and multiplied by the total number of animals in the category. The emission rate of a gas was the product of the enhanced concentration (difference between indoor and outdoor concentrations expressed in g m3) and the ventilation rate.
Enteric CH4 emission rate was estimated for the dairy cows according to the IPCC (2006) Tier 2 method using the actual productivity parameters and feed characteristics for the animals in the barn (e.g. live weight, milk production data, dry matter, nutrients and gross energy intake) and the predicted methane conversion factor (Ym) according to Ellis et al. (2007).
2.5.
Gas concentrations measured inside and outside the barn are presented in Table 2. Indoor gas concentrations were higher in the spring than in the fall, partly due to the higher animal population in the spring (11 more cows in the spring than in the autumn e Table 1). Furthermore, an expected lower ventilation rate in the spring (which was colder, 7.4 C) than the fall (which was warmer, 10.3 C) might also explain the higher gas concentrations in the spring. The background concentrations were within the range of values reported in another study that used similar instrumentation and sampling technique in a dairy cow barn: 431 ppm for CO2, 5.7 ppm for CH4, 0.71 ppm for NH3 and 0.35 ppm for N2O (Wu, Zhang, & Kai, 2012). It is expected that emissions from nearby sources (manure tank, farm land and machinery) are likely to influence outdoor concentrations. Delayed tracking of NH3 concentrations by the acoustic analyser (Rom & Zhang, 2010) and adsorption/desorption in the sampling lines (Mukhtar et al., 2003) may have also influenced concentrations measured, especially from the background locations. The variations in the average hourly concentrations indicated low concentrations in the daytime and high concentrations at night (Fig. 2). This shows that a better indoor air quality is achievable in naturally ventilated animal barns within the daytime as compared to the night-time. The lower concentrations in the daytime relative to night-time are related to higher daytime ventilation rates. The concentrations of all the gases never exceeded the Ontario provincial time-weighted average exposure limits, which are 5000 ppm
Animal activity
The activity or movement of the cows was measured with an activity monitoring system which is incorporated in the ALPRO™ dairy herd management system (Tumba, Sweden). Animal activity derived by the activity monitoring system is primarily used to determine when animals are in heat so as to identify optimal insemination time and for monitoring animal health deviations. The system consists of a compact collarmounted activity meter around the neck of each cow. The collar has a sensor that detects motion when a magnetic ball moves in a cavity surrounded by copper coils. The movement of the ball is converted into an electrical potential and the impulse is registered, stored and remotely transmitted to an antenna installed in the barn for onward processing by the ALPRO™ software in a computer. With the exception of the heifers, all the dairy cows (176 in the spring and 165 in the autumn) were mounted with collars that measured animal activity. The activity is recorded as either 1 or 0 within each 14 s that the animal moves, resulting in a maximum of 255 counts for each hour. A sum of the counts within each hour represents the hourly activity. The total activity for each hour in the barn was calculated as the mean activity for all the cows. The relative activity on a diurnal basis was calculated as the ratio of the activity for a specific hour and the mean activity for each day. In addition to direct measurements of animal movement using activity meters, qualitative physical observations were made as to the time of day with increased
3.
Results and discussion
3.1.
Gas concentrations and climate data
Table 2 e Gas concentrations and environmental data during measurements in the dairy cow barn (mean ± standard deviation). Gas/climate parameter
Spring measurements Indoor
CO2, ppm CH4, ppm NH3, ppm N2O, ppm Air temperature, C Relative humidity, %
1000 ± 335 68 ± 39 3.8 ± 2.1 0.47 ± 0.08 7.4 ± 2.2 76 ± 14
Fall measurements
Outdoor 453 7.9 1.1 0.37 1.0 79
± 43 ± 5.9 ± 0.4 ± 0.03 ± 6.4 ± 22
Indoor 685 ± 35 ± 2.2 ± 0.39 ± 10.3 ± 83 ±
119 13 0.6 0.04 3.9 10
Outdoor 422 8.7 1.3 0.36 7.0 89
± 22 ± 2.2 ± 0.2 ± 0.03 ± 5.3 ± 15
108
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Fig. 2 e Diurnal variations in gas concentrations with values representing hourly averages for the spring (FebruaryeApril).
for CO2, 1000 ppm for CH4, and 25 ppm for NH3 and N2O (OML, 2010).
3.2.
Animal activity and ventilation rates
The activity of the cows showed a distinct diurnal pattern with two peaks (Fig. 3). As expected, the standard deviations show a larger variation in animal activity within the daytime as compared to the nigh-time. Despite the similarity in activity patterns for both seasons, there was a slight difference in the times when the minimum and maximum activities were measured. The minimum activity was measured at about 5 and 6 a.m. in the spring and in the autumn, receptively. The morning activities peaked at different times as well: about 8 and 10 a.m. in the spring and the autumn respectively. However, the evening activities peaked at the same time at about 7 p.m. The time lag in daytime activities for the different seasons could be due to changes in the time when routine duties were initiated in the barn (lighting, examining the cows, cleaning and feeding). It was also possible that changes in daylight saving time might have affected routine duties in the barn, which then affected the movement of the cows. The evening activity was related to animal movements resulting from feeding, milking and routine animal examinations by the workers at the end of the day. Animal activity patterns that are similar to those in the current study have been reported in other barns with the use
of infrared detectors (CIGR, 2002; Ngwabie et al., 2011). Figure 4 shows how the animal activity measured with the DeLaval activity monitoring system compares with measurements obtained using an infrared passive detector in another study (Ngwabie et al., 2011). The barn used by Ngwabie et al. (2011) and the barn used here had similar management systems (milking robots, manure management). Both measurement methods show close similarity in cow activity variation patterns, but for a difference in magnitude. This could be explained by the fact that the infrared sensory technique considers all movements (animals, workers) in its field of view (Pedersen & Pedersen, 1995), and the field of view may not include all the cows in the barn. In addition, the detection of a temperature change against a background (barn floor) when an animal moves might also affect the measurement, especially when an animal moves against a background made up of other animals. On the other hand, the activity monitoring system considers all cows, with no other external factors, indicating the likelihood of a higher spatial resolution measurement. Another advantage is that there is no additional operational cost and time if the system is already mounted and used by the farmer for herd management optimisation. The average ventilation rates during the measurement periods are presented in Table 3. Higher ventilation rates in the autumn than in the spring could be related to the higher indoor temperatures in the autumn (3 C higher) and because the barn doors and side wall curtains were mostly open in the autumn given the higher outdoor air temperature (Table 2).
109
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Fig. 3 e Average diurnal variations in dairy cow activity measured using the DeLaval activity monitoring system during the spring (FebruaryeApril) and the fall (September and October). The vertical lines represent the standard deviations for similar hours throughout the measurement period.
Errors in estimating the ventilation rate could arise from the uncertainty in the CO2 production factor used in Eq. (3) and also from the representativeness of measurement locations due to imperfect mixing of air in the barn (Van Buggenhout et al., 2009; Wu, Zhai, Zhang, & Nielsen, 2012). Discrepancies have been reported in ventilation rates between direct measurements at exhaust fans and the CO2 mass balance of 2e17% (De Sousa & Pedersen, 2004) and 6.5% (Hinz & Linke, 1998a, 1998b). In another study that compared ventilation rates using four methods over 41 d in a mechanically ventilated building for pigs, the CO2 balance method was 8% lower than the ventilation rate measured using a fan (Blanes & Pedersen, 2005). Diurnal variations in the hourly CO2 ventilation rates in the spring and the autumn showed high values in the daytime
and low values at night (Fig. 5). The observed ventilation rate pattern was related not only to the animal activity, but also to other factors such as wind speed (R2 ¼ 0.50 and 0.31 in the spring and fall respectively) and wind direction (R2 ¼ 0.54 and 0.33 in the spring and autumn respectively). Although the wind speed (Table 3) was similar in both seasons, the ventilation rate had a better correlation with values in the spring than in the fall. The construction of a large anaerobic digester tank close to the barn in the autumn (Fig. 1) might have influenced the relationship between the wind and the ventilation rate in the later part of the measurement. The large spike in the autumn ventilation rate at about 10 a.m. could be related to the opening of the side wall curtains during this period due to the high indoor air temperature.
3.3.
Emission rates
3.3.1.
Measured emissions
Diurnal variations were observed in the emissions rates with a peak in the morning and in the evening (Fig. 6). Detailed hourly information of the emission variation was expected in this measurement since the hourly ventilation rates were used in calculating the emissions. Diurnal variations in emissions are common in animal barns (Ngwabie et al., 2011; Zhu et al., 2012) and are related to several factors: ventilation rate, management routines and temperature changes. Spikes in NH3 emissions, which is emitted predominantly from the manure, was related to periods of high urinating/defecating frequency which was observed in the mornings and in the
Table 3 e Barn ventilation rates and outdoor wind speed data during the measurements.
Fig. 4 e A comparison of the animal activity in the present measurements with reported values measured using an infrared passive detector inside a dairy barn.
Spring Fall
Ventilation rate, m3 LU1 h1
Wind speed, m s1
531 ± 452 824 ± 492
3.6 ± 2.0 3.2 ± 1.8
Mean ± standard deviation, 1 LU ¼ 500 kg animal mass.
110
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
evenings. Lower NH3 emissions were measured during the night when the ventilation rate and temperature were lower and the cows were resting. The mean emission rates are presented in Table 4. Despite a 55% increase in the mean CO2 balance ventilation rate from the spring to the fall season, the average emission factors dropped by just 12% for CH4, 33% for NH3 and 29% for N2O over the same period. There was a drop in gas concentrations from the autumn to the fall associated with the increased ventilation rate (Table 2). A change in the CH4 emission factor proportional to ventilation rate was not expected, as it is mostly controlled by animal numbers and diet. The larger changes in NH3 and to an extent, N2O emission factors could be related to some other factors that influence gas concentrations in the barn such as animal number (Table 1), temperature (Table 2) and manure retention time inside the barn. Changes in the supplementary bedding material from straw in the spring to solid digestate in the fall might also have affected the emission rates of NH3 and N2O due to potential changes in the rate of immobilisation for the different bedding materials (Tasistro, Cabrera, Ritz, & Kissel, 2008). Fig. 5 e Diurnal variations in ventilation rates calculated as the average for each hour in the spring (FebruaryeApril) and in the fall (September and October).
3.3.2.
Modelled CH4 emissions
Modelled enteric CH4 emission rates and nutritional composition for the different categories of dairy cows using the IPCC (2006) Tier 2 method is presented in Table 5. Regarding additional local barn data in Table 1, all the rations contained a higher proportion of forage than concentrate, however,
Fig. 6 e Diurnal variations in emission rates from the barn with dairy cows, with values representing hourly averages in the spring (FebruaryeApril) and in the fall (September and October).
111
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Table 4 e Emission rates from the dairy cow barn calculated using the CO2 mass balance ventilation rate. Spring 1
Fall
1
CH4, g LU h 13.9 ± 5.3 (4.5e34.5) 12.2 ± 5.5 (3.0e32.1) NH3, g LU1 h1 0.64 ± 0.32 (0.09e2.06) 0.43 ± 0.22 (0.09e1.47) N2O, mg LU1 h1 41.3 ± 21.7 (0.45e128.93) 29.4 ± 21.7 (0e123.82) 1 LU ¼ 500 kg animal mass. Total LU in the spring and fall were 272 and 258 respectively. Mean ± standard deviation (range).
rations fed to the high producing cows and transition cows had a relatively lower proportion of forages (62e76%) compared with rations fed to low producing cows, dry cows and heifers (89e92% of forages in the ration). This was reflected in the chemical composition of the rations, as both neutral detergent fibre (NDF) and acid detergent fibre (ADF) contents were on average 24% higher in rations for low producing cows, dry cows and heifers compared with rations for high producing and transition cows (Table 5). Average dry matter intake was 67% higher for lactating dairy cows compared with that for dry cows and heifers (19.2 vs. 11.5 kg head1 d1, Table 5). Higher dry matter intake and lower NDF and ADF contents in the ration tend to decrease the fraction of gross energy intake converted to CH4 in dairy cattle (Ellis et al., 2007), therefore predicted CH4 conversion factor (Ym) was on average 17% higher for dry cows and heifers compared with that for lactating dairy cows (0.063 ± 0.012 for dry cows and heifers versus 0.054 ± 0.011 for lactating cows).
3.3.3. Enteric CH4 emission factors derived using the measured emissions Measured CH4 emissions from the dairy barn during the spring and fall measurement campaigns were equivalent to 411 and 360 g head1 d1, respectively resulting in a mean
emission rate of 385 g CH4 head1 d1 over the two seasons (Table 6). If it is assumed that all the CH4 measured in the current study was entirely from enteric fermentation, and that a non-lactating dairy animal (dry cows and heifers > 1 year, 75 out of 216 in this barn) produces about 60% of enteric CH4 relative to that from a lactating cow (Holter & Young, 1992), then an estimated enteric CH4 emission of 447 g head1 d1 per lactating cow can be derived (Table 6). This value is higher than the range of enteric CH4 emissions for lactating dairy cows reported in Canada (329e424 g head1 d1) from a review of 11 studies (Boadi, Benchaar, Chiquette, & Masse, 2004). In comparison, the estimated enteric CH4 emission rate for a lactating dairy cow calculated according to the IPCC (2006) Tier 2 method, but using the actual dry matter, nutrients and gross energy intake and predicted CH4 conversion factor (Ym) according to Ellis et al. (2007) was 363 ± 73 g head1 d1 (Table 6). The higher CH4 emission rate observed from direct measurements in the current study, compared with both previously reported and the current modelled emission rates, may be attributed to uncertainty associated with ventilation estimates (through CO2 balance) and the potential contributions of CH4 emitted from the liquid manure present in the temporary holding pit located in one end of the barn leading to higher CH4 concentrations (Fig. 1). Generally, indoor pits located beneath the floor can temporarily hold liquid dairy manure before being pumped outdoor where the main storage facility is located (StatisticsCanada, 2003). Using laboratory scale storage units, Sommer, Petersen, Sørensen, Poulsen, and Møller (2007) demonstrated that the presence of a small quantity of inoculum (~8%) could support the immediate production of appreciable amounts of CH4 from manure held over 10 d. Therefore CH4 could be generated in liquid manure held temporarily for about 2e3 weeks in under-floor storage pits where complete cleaning after each emptying is usually not carried out (Sommer et al.,
Table 5 e Nutrients composition, dry matter intake (DMI) and gross energy intake (GEI) and estimated rates of CH4 emissions from enteric fermentation by different dairy cattle groups in the barn where emission measurements were undertaken. High producing cows Nutrient composition (g kg1) Crude protein (CP) Ether Extract (EE) Neutral detergent fibre (NDF) Acid detergent fibre (ADF) Ash Non-fibre carbohydrate (NFC)a Total digestible nutrients (TDN), % Dry matter intake (DMI), kg head1 d1 Gross energy intake (GEI) (MJ1 head1 d1)b Methane conversion factor (Ym) (fraction of GEI)c Enteric CH4 production (g CH4 head1 d1)d a b c d
1
Low producing cows
Transition cows
Dry cows
Breeding heifers
182.3 34.3 322.3 198.4 80.5 380.6 76.8
175.3 31.9 375.5 253.6 79.3 338.0 72.8
141.8 37.2 392.2 254.4 74.8 354.0 72.7
116.3 32.2 457.9 295.7 61.4 331.2 67.3
143.1 30.2 428.9 297.1 73.6 324.2 68.8
21.3 385.0 0.052 ± 0.010
19.4 355.1 0.054 ± 0.011
17.0 312.4 0.057 ± 0.011
12.4 228.3 0.063 ± 0.012
11.0 200.9 0.062 ± 0.012
362 ± 73
348 ± 69
320 ± 64
260 ± 52
223 ± 45
Calculated: NFC (g kg ) ¼ 1000 (CP þ EE þ NDF þ Ash) (NRC, 2001). Calculated: GEI (MJ kg1) ¼ DMI [(23.6 CP þ 39.8 ether extract þ 17.3 NFC þ 18.9 NDF)/1000] (Ramin & Huhtanen, 2013). Based on the prediction equation of Ellis et al. (2007) using DMI, NDF and ADF intake. Converted from MJ head1 d1 (assuming an energy density of 55.65 MJ kg1 CH4, IPCC, 2006).
112
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Table 6 e Enteric CH4 emission factors derived for dairy cows using the measured CH4 emissions from the barn. Measured emissions corrected with CH4 from manure
Estimated enteric CH4 emissionsa
83.1 216
83.1 216
67.3 ± 13.5 216
360 NE 360
385 NE 385
385 25.1b 360
312 ± 63 e 312 ± 63
420 252
447 268
418 251
363 ± 73 241 ± 49
Measured emissions Spring
Fall
Average
90.7 221
75.5 210
Emissions per animal (g CH4 head1 d1) All sources 411 Emissions from manure NE Emissions from all cattle 411 Emission portioningc Lactating cows Non-lactating dairy cattle
Total CH4 emissions (kg CH4 d1) Number of animals in the barn
475 285
For units conversions: 141 lactating cows ¼ 187 LU; 75 non-lactating cows ¼ 78 LU. a Estimated using IPCC (2006) Tier 2 method with current barn input data and a 20% uncertainty. NE: no emissions from the manure was assumed. b Average CH4 emission of 25.1 g animal1 day1 from indoor manure pit was assumed according to Kinsman et al. (1995). c Partitioning of total measured emission assumed that non-lactating dairy cow (dry cows and heifers > 1 year, 75 out of 216 in this barn) produced about 60% of enteric CH4 relative to that from a lactating cow (Holter & Young, 1992).
2007; Wood, VanderZaag, Wagner-Riddle, Smith, & Gordon, 2014). However, very little published data is available for CH4 emissions from dairy manure held temporarily in indoor pits in Canadian dairy barns. We are aware of only one study that measured CH4 emissions from the manure holding pit underneath the floor in a dairy barn in Canada which provided an average emission rate equivalent to 25.1 g head1 d1 (about 6% of measured CH4 emissions from the barn) from manure held for about 21 days in an eastern Ontario dairy barn with 118 lactating Holstein cows (Kinsman, Sauer, Jackson, & Wolynetz, 1995). In another similar study in Germany, an average emission rate of 56 g head1 d1 from liquid manure held temporary for 23 d before transferred to the main storage in a dairy barn housing 34 lactating cows and 9 dry cows was measured (Marik & Levin, 1996). Methane formation in liquid manure can be influenced both by residence time and temperature (Sommer et al., 2007). Higher average CH4 emissions reported by Marik and Levin (1996) compared with those reported by Kinsman et al. (1995) for indoor manure held over a similar residence time (approximately 3 weeks) may be due to the differences in temperature where the studies were conducted (average annual temperature 10.5 C at the site studied by Marik and Levin (1996) versus 6.6 C at the site studied by Kinsman et al. (1995). Our study was conducted in a similar temperature zone (with an average annual temperature of 6.7 C) to that of Kinsman et al. (1995). Furthermore, manure was temporary held for about 2e3 weeks in the indoor pit with an outdoor extension, before being pumped to the long-term storage tank in the present study. If we assume that potential contributions of CH4 from manure held underneath the floor in the current study was similar to the average emissions reported by Kinsman et al. (1995) which was 25.1 g head1 d1), then this will result in an average emission rate of 418 g head1 d1 for a lactating cow, which falls within the range of enteric CH4 emission rates reported for lactating dairy cows in Canada (329e424 g head1 d1, Boadi et al., 2004).
Based on the ratio of the modelled enteric CH4 emission (312 g head1 d1) and the mean measured barn CH4 emission (385 g head1 d1) in Table 6, enteric fermentation accounted for about 81% of the barn emission. This showed that the manure could account for 19% of the barn emissions. However if the reported CH4 manure emission of ~25 g head1 d1 (Kinsman et al., 1995) was considered, enteric fermentation will rather account for over 90% of the barn emissions. Our direct and modelled measurements showed that CH4 emissions from the manure in the current barn can be as high as 73 g head1 d1 (2.5 g LU1 h1). Several researchers have measured CH4 and NH3 emissions from naturally ventilation barns in the cold climatic regions of North America and Europe. Table 7 presents average values of some reported emissions where enteric fermentation was not separated from manure emissions and how they compare to the present study. Generally, the measured CH4 emission factor was within the range of reported values while the measured NH3 emission was somewhat lower than reported values. Differences in emission factors are likely related to differences in regional climate, feed composition, productivity level, building and the management systems. There is little available information of N2O emissions from dairy barns as the emission factor is expected to be small e.g. 0.1 g N2O LU1 d1 have been measured in a barn with lactating cows (Adviento-Borbe et al., 2010). The higher N2O emissions in the current study, equivalent to 0.85 g LU1 d1 could be due to favourable conditions for its production resulting from the mixing of bedding material with the manure.
4.
Conclusions
This study used animal activity obtained from an activity monitoring system in automated barns for dairy cows to model the ventilation rate on an hourly basis using the CO2 balance method. The emission factors of CH4, NH3 and N2O were subsequently generated and their diurnal variation
113
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Table 7 e Summary CH4 and NH3 emission factors from naturally ventilated dairy barns. CH4 12.2e13.9 g LU1 h1 10.60a g LU1 h1 18.1b g LU1 h1 10.8e11.4 g LU1 h1 14.5e17.9 g LU1 h1 e
NH3
Country, ventilation type
Source
0.43e0.64 g LU1 h1 e e 0.81e1.0 g LU1 h1 2.9e3.5 g LU1 h1
Ontario e Canada, CO2 mass balance e Alberta e Canada, dispersion model Sweden, CO2 mass balance Germany, CO2 mass balance
0.3e2.8 g LU1 h1
Switzerland, Tracer ratio
Present direct measurement Present modelled results (McGinn & Beauchemin, 2012) (Ngwabie et al., 2011) (Samer, Berg, et al., 2011; Samer, Fiedler, et al., 2011) (Schrade et al., 2012)
1 LU ¼ 500 kg animal mass. a Calculated from the total modelled enteric emission of 67,336 g CH4 day1 (Table 5) and a mean live unit of 265 LU in Table 4. b Recalculated from 15.1 g cow1 h1 and 600 kg of cow mass.
patterns were studied. Enteric CH4 emission factor was also generated using the IPCC (2006) Tier 2 method with local barn input data. The following conclusions were drawn from this study: The activity monitoring system provides a high resolution measurement of animal activity (movement) since all animals are considered. It has the potential to improve the temporal resolutions of the ventilation rate from 24 to 1 h when calculated using the CO2 balance method and the resulting emission variation pattern. The mean ventilation rates were 531 and 824 m3 LU1 h1 in the spring and fall respectively with pronounced diurnal variations. Diurnal variations were observed in the emissions rates with a peak in the morning and in the evening that were related to management routines and climatic conditions in the barn. The peaks were more prominent for CH4 and NH3 than for N2O. The average measured emission factors were 13.9 g CH4 LU1 h1, 0.64 g NH3 LU1 h1 and 41.3 mg N2O LU1 h1 in the spring and 12.2 g CH4 LU1 h1, 0.43 g NH3 LU1 h1 and 29.4 mg N2O LU1 h1 in the fall. Emissions factors dropped by 12% for CH4, 33% for NH3 and 29% for N2O from the spring to the fall. The average measured barn CH4 emission factor from all sources and the entire dairy cow category was equivalent to 385 g head1 d1 (13.08 g LU1 h1) while the modelled enteric emission factor was 312 g head1 d1 (10.58 g CH4 LU1 h1). An attempt at source apportioning indicated that indoor manure could account for up to 73 g head1 d1 (2.5 g CH4 LU1 h1). A modelled lactating cow emitted about 363 g head1 d1 (11.08 g CH4 LU1 h1) while a modelled non-lactating cows emitted about 241 g head1 d1 (9.67 g CH4 LU1 h1). The ratio of the modelled enteric CH4 emission to the measured barn CH4 emission indicated that enteric fermentation accounted for about 81% of the total barn emission. Additional activity measurements using other methods alongside the activity monitoring system is recommended for full validation of its use for ventilation rate calculations. Further research using direct measurement methods needs to be carried to determine the contribution of indoor manure to barn CH4 emissions.
Acknowledgement Sponsorship for this projected was provided by the Agricultural Greenhouse Gas Program funded by Agriculture and Agri-Food Canada- project number 1585-16-3-1-14. Jordan Forsyth, the proprietor and management of Clovermead Dairy Farm were indispensable in the execution of this project.
references
Adviento-Borbe, M. A. A., Wheeler, E. F., Brown, N. E., Topper, P. A., Graves, R. E., Ishler, V. A., et al. (2010). Ammonia and greenhouse gas flux from manure in freestall barn with dairy cows on precision fed rations. Transactions of the ASABE, 53, 1251e1266. Blanes, V., & Pedersen, S. (2005). Ventilation flow in pig houses measured and calculated by carbon dioxide, moisture and heat balance equations. Biosystems Engineering, 92, 483e493. Bluteau, C. V., Masse, D. I., & Leduc, R. (2009). Ammonia emission rates from dairy livestock buildings in Eastern Canada. Biosystems Engineering, 103, 480e488. Boadi, D., Benchaar, C., Chiquette, J., & Masse, D. (2004). Mitigation strategies to reduce enteric methane emissions from dairy cows: update review. Canadian Journal of Animal Science, 84, 319e335. CIGR. (2002). Climatization of animal houses. Forth report of working group. Danish Institute of Agricultural Sciences, Research Centre Bygholm. Cornou, C., & Lundbye-Christensen, S. (2012). Modeling of sows diurnal activity pattern and detection of parturition using acceleration measurements. Computers and Electronics in Agriculture, 80, 97e104. Costa, A., Borgonovo, F., Leroy, T., Berckmans, D., & Guarino, M. (2009). Dust concentration variation in relation to animal activity in a pig barn. Biosystems Engineering, 104, 118e124. De Sousa, P., & Pedersen, S. (2004). Ammonia emission from fattening pig houses in relation to animal activity and carbon dioxide production. Agricultural Engineering International: The CIGR Ejournal, 6. Manuscript BC 04 003. Ellis, J. L., Kebreab, E., Odongo, N. E., McBride, B. W., Okine, E. K., & France, J. (2007). Prediction of methane production from dairy and beef cattle. Journal of Dairy Science, 90, 3456e3466. Environment-Canada. (2012). National inventory report 1990e2010: Greenhouse gas sources and sinks in Canada e Executive summary. Government of Canada.
114
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Harper, L. A., Flesch, T. K., Powell, J. M., Coblentz, W. K., Jokela, W. E., & Martin, N. P. (2009). Ammonia emissions from dairy production in Wisconsin. Journal of Dairy Science, 92, 2326e2337. Hinz, T., & Linke, S. (1998a). A comprehensive experimental study of aerial pollutants in and emissions from livestock buildings. Part 1. Methods. Journal of Agricultural Engineering Research, 70, 111e118. Hinz, T., & Linke, S. (1998b). A comprehensive experimental study of aerial pollutants in and emissions from livestock buildings. Part 2: results. Journal of Agricultural Engineering Research, 70, 119e129. Holter, J. B., & Young, A. J. (1992). Methane prediction in dry and lactating Holstein cows. Journal of Dairy Science, 75, 2165e2175. IPCC. (2006). Guidelines for national greenhouse gas inventories. In Emissions from livestock and manure management (Vol. 4). AFOLU Intergovernmental Panel on Climate Change. Jayasundara, S., & Wagner-Riddle, C. (2014). Greenhouse gas emissions intensity of Ontario milk production in 2011 compared with 1991. Canadian Journal of Animal Science, 94, 155e173. Kinsman, R., Sauer, F. D., Jackson, H. A., & Wolynetz, M. S. (1995). Methane and carbon dioxide emissions from dairy cows in full lactation monitored over a six-month period. Journal of Dairy Science, 78, 2760e2766. Li, C. S., Salas, W., Zhang, R. H., Krauter, C., Rotz, A., & Mitloehner, F. (2012). Manure-DNDC: a biogeochemical process model for quantifying greenhouse gas and ammonia emissions from livestock manure systems. Nutrient Cycling in Agroecosystems, 93, 163e200. Marik, T., & Levin, I. (1996). A new tracer experiment to estimate the methane emissions from a dairy cow shed using sulfur hexafluoride (SF6). Global Biogeochemical Cycles, 10, 413e418. McGinn, S. M., & Beauchemin, K. A. (2012). Dairy farm methane emissions using a dispersion model. Journal of Environmental Quality, 41, 73e79. McGinn, S. M., Flesch, T. K., Harper, L. A., & Beauchemin, K. A. (2006). An approach for measuring methane emissions from whole farms. Journal of Environmental Quality, 35, 14e20. Mukhtar, S., Rose, A., Capareda, S., Boriack, C., Lacey, R., Shaw, B., et al. (2003). Assessment of ammonia adsorption onto teflon and LDPE tubing used in pollutant stream conveyance. Agricultural Engineering International: The CIGR Ejournal., 5. Manuscript BC 03 012. Ngwabie, N. M., Jeppsson, K. H., Gustafsson, G., & Nimmermark, S. (2011). Effects of animal activity and air temperature on methane and ammonia emissions from a naturally ventilated building for dairy cows. Atmospheric Environment, 45, 6760e6768. Ni, J. Q., & Heber, A. J. (2008). Sampling and measurement of ammonia at animal facilities. In Advances in agronomy (Vol. 98, pp. 201e269). San Diego, USA: Elsevier Academic Press Inc. Ni, J. Q., Heber, A. J., Darr, M. J., Lim, T. T., Diehl, C. A., & Bogan, B. W. (2009). Air quality monitoring and on-site computer system for livestock and poultry environment studies. Transactions of the ASABE, 52, 937e947. Ni, J., Vinckier, C., Hendriks, J., & Coenegrachts, J. (1999). Production of carbon dioxide in a fattening pig house under field conditions. II. Release from the manure. Atmospheric Environment, 33(22), 3697e3703. NRC. (2001). National Research Council. Nutrient requirements of dairy cattle (7th revised ed.). Washington, DC: The National Academies Press. Ogink, N. W. M., Mosquera, J., Calvet, S., & Zhang, G. (2013). Methods for measuring gas emissions from naturally ventilated livestock buildings: developments over the last decade and perspectives for improvement. Biosystems Engineering, 116(3), 297e308.
OML (Ontario Ministry of Labour). (2010). Occupational exposure limits for Ontario workplaces. Pedersen, S., Blanes-Vidal, V., Joergensen, H., Chwalibog, A., Haeussermann, A., Heetkamp, M. J. W., et al. (2008). Carbon dioxide production in animal houses: a literature review. Agricultural Engineering International, 10. Manuscript BC 08 008. Pedersen, S., & Pedersen, C. B. (1995). Animal activity measured by infrared detectors. Journal of Agricultural Engineering Research, 61, 239e246. Ramin, M., & Huhtanen, P. (2013). Development of equations for predicting methane emissions from ruminants. Journal of Dairy Science, 96, 2476e2493. Rom, H. B., & Zhang, G. Q. (2010). Time delay for aerial ammonia concentration measurements in livestock buildings. Sensors, 10, 4634e4642. Samer, M., Berg, W., Muller, H. J., Fiedler, M., Glaser, M., Ammon, C., et al. (2011). Radioactive (85)kr and CO2 balance for ventilation rate measurements and gaseous emissions quantification through naturally ventilated barns. Transactions of the ASABE, 54, 1137e1148. Samer, M., Fiedler, M., Muller, H. J., Glaser, M., Ammon, C., Berg, W., et al. (2011). Winter measurements of air exchange rates using tracer gas technique and quantification of gaseous emissions from a naturally ventilated dairy barn. Applied Engineering in Agriculture, 27, 1015e1025. Schrade, S., Zeyer, K., Gygax, L., Emmenegger, L., Hartung, E., & Keck, M. (2012). Ammonia emissions and emission factors of naturally ventilated dairy housing with solid floors and an outdoor exercise area in Switzerland. Atmospheric Environment, 47, 183e194. Sheppard, S. C., Bittman, S., Swift, M. L., Beaulieu, M., & Sheppard, M. I. (2011). Ecoregion and farm size differences in dairy feed and manure nitrogen management: a survey. Canadian Journal of Animal Science, 91, 459e473. Sommer, S. G., Petersen, S. O., Sørensen, P., Poulsen, H. D., & Møller, H. B. (2007). Methane and carbon dioxide emissions and nitrogen turnover during liquid manure storage. Nutrient Cycling in Agroecosystems, 78, 27e36. Statistics-Canada. (2003). Manure storage in Canada (Vol. 1). No. 1. Catalogue no. 21-021-MIE. Tasistro, A. S., Cabrera, M. L., Ritz, C. W., & Kissel, D. E. (2008). Manipulating bedding materials and PLT (TM) to reduce NH3 emissions from broiler manure. Bioresource Technology, 99(6), 1952e1960. Van Buggenhout, S., Van Brecht, A., Ozcan, S. E., Vranken, E., Van Malcot, W., & Berckmans, D. (2009). Influence of sampling positions on accuracy of tracer gas measurements in ventilated spaces. Biosystems Engineering, 104, 216e223. Wheeler, E. F., Weiss, R. W. J., & Weidenboerner, E. (2000). Evaluation of instrumentation for measuring aerial ammonia in poultry houses. Journal of Applied Poultry Research, 9, 443e452. Wood, J. D., VanderZaag, A. C., Wagner-Riddle, C., Smith, E. L., & Gordon, R. J. (2014). Application of a high-temporal resolution method to estimate ammonia emissions from farmland. Nutrient Cycling in Agroecosystems, 99, 95e105. Wu, W. T., Zhai, J., Zhang, G. Q., & Nielsen, P. V. (2012). Evaluation of methods for determining air exchange rate in a naturally ventilated dairy cattle building with large openings using computational fluid dynamics (CFD). Atmospheric Environment, 63, 179e188. Wu, W. T., Zhang, G. Q., & Kai, P. (2012). Ammonia and methane emissions from two naturally ventilated dairy cattle buildings and the influence of climatic factors on ammonia emissions. Atmospheric Environment, 61, 232e243. Zhu, Z., Dong, H., & Zhou, Z. (2012). Ammonia and greenhouse gas emissions from a dairy cattle barn with a daily manure collection system. Transactions of the ASABE, 55, 1959e1964.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/issn/15375110
Research Paper
Measurements of emission factors from a naturally ventilated commercial barn for dairy cows in a cold climate Ngwa M. Ngwabie a,*, Andrew Vanderzaag b, Susantha Jayasundara c, Claudia Wagner-Riddle c a
Department of Environmental Science, Faculty of Science, University of Buea, Box 63, Buea, South West Region, Cameroon b Agriculture and Agri-Food Canada, Ottawa, Ontario, Canada c School of Environmental Science, Ontario Agricultural College, University of Guelph, Guelph, Ontario N1G 2W1, Canada
article info
Emission rates of CH4, N2O and NH3 were measured in a commercial free-stall barn that
Article history:
housed 141 lactating dairy cows, and 75 dry cows and replacement heifers. Animal activity,
Received 3 February 2014
measured using the ALPRO™ dairy herd management system was used together with the
Received in revised form
CO2 balance method to calculate the ventilation rate. Methane emission was also modelled
16 August 2014
using the IPCC Tier 2 method. Animal activity variations similar to reported patterns
Accepted 26 August 2014
indicated that the activity monitoring system provided high resolution measurements
Published online 16 September 2014
since all cows were considered. Diurnal variations were observed in the emissions with
Keywords: Animal activity Emission factor Dairy cows Greenhouse gases Enteric fermentation
1.
mean values of 12.2e13.9 g CH4 LU1 h1, 0.43e0.64 g NH3 LU1 h1 and 29.4 e41.3 mg N2O LU1 h1. Modelled enteric CH4 emission was 312 g CH4 head1 d1 (10.58 g CH4 LU1 h1). It was estimated that indoor manure emitted 73 g CH4 head1 d1 (2.5 g CH4 LU1 h1), with enteric fermentation representing 81% of the total barn CH4 emission. Lactating cows emitted about 363 g CH4 head1 d1 (11.42 g CH4 LU1 h1) while non-lactating cows emitted 241 g CH4 head1 d1 (9.67 g CH4 LU1 h1). Crown Copyright © 2014 Published by Elsevier Ltd on behalf of IAgrE. All rights reserved.
Introduction
The agricultural sector is a significant contributor to greenhouse gas (GHG) emissions, accounting for 8% of the total 2010 GHG emissions in Canada, indicating a 19% increase from the 1990 level (Environment-Canada, 2012). In particular, agriculture accounts for 24% and 72% of the Canadian CH4 and N2O
emissions, respectively, with the main sources coming from cattle and pig raising as well as the use of synthetic nitrogen fertilisers. In total, Canadian livestock contribute about 60% of the total agricultural GHG emissions. Increasing trends in estimated national emission factors have been observed in some livestock categories, e.g. CH4 emission from enteric fermentation in dairy cows increased
* Corresponding author. Tel.: þ237 71643209. E-mail address: [email protected] (N.M. Ngwabie). http://dx.doi.org/10.1016/j.biosystemseng.2014.08.016 1537-5110/Crown Copyright © 2014 Published by Elsevier Ltd on behalf of IAgrE. All rights reserved.
104
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Nomenclature ADF AU CF
acid detergent fibre animal unit (500 kg animal mass) correction factor for the heat produced at any temperature Ti ( C) CO2indoors CO2 indoor concentration (ppmv) CO2outdoors CO2 outdoor concentration (ppmv) CP crude protein DMI dry matter intake (kg head1 d1) E average metabolisable feed energy content (MJ kg1 dry matter) EE ether extract G daily gain in weight of a heifer (kg d1) GEI gross energy intake (MJ1 head1 d1) GHG greenhouse gas HPU heat produced unit IPCC Intergovernmental Panel on Climate Change LU livestock unit (500 kg animal mass) M average animal mass (kg) NDF neutral detergent fibre NE no emissions assumed NFC non-fibre carbohydrate (g kg1) PTFE polytetrafluoroethylene RA relative animal activity measured temperature ( C) Ti TDN total digestible nutrients (%) ventilation rate per heat producing unit VRHPU (m3 h1 HPU1) Y average daily milk production (kg d1) methane conversion factor Ym total heat produced by dairy cows and heifers Ftot (W) Fdairy cows heat produced by dairy cows (W) Fheifers heat produced by heifers (W) from 109.4 kg head1 yr1 in 1990 to 127.1 kg head1 yr1 in 2010, with an associated increase in milk production (Environment-Canada, 2012). However, significant differences still exist between emission factors at the barn, provincial and national levels due to the simplified approach needed for national scale inventories and the diversified nature of dairy management and climatic conditions (Environment-Canada, 2012; IPCC, 2006). Canadian dairy cows are mostly kept indoors in naturally ventilated barns but in some cases they are sent outdoor for a few hours during the day in warm weather (Sheppard, Bittman, Swift, Beaulieu, & Sheppard, 2011). On average, Canadian dairy cattle manure storage is evenly distributed among solid and liquid forms (~40% each), with ~20% being deposited on pastures; however in certain provinces, the proportion of dairy manure handled as liquid can be as high as 89% or as low as 20% (Environment-Canada, 2012; Statistics-Canada, 2003). Although several measurements have been conducted to quantify and study the variation patterns of NH3 and GHG emissions from dairy cow barns in the cold regions of North America and Europe (Harper et al., 2009; Ngwabie, Jeppsson, Gustafsson, & Nimmermark, 2011; Zhu, Dong, & Zhou, 2012), only a few measurements have been conducted in Canada (Bluteau, Masse, & Leduc, 2009;
McGinn & Beauchemin, 2012; McGinn, Flesch, Harper, & Beauchemin, 2006). More measurements are therefore needed to fully understand and quantify emissions from commercial dairy cow barns in Canada. However, direct measurements can be expensive and the application of models (IPCC, 2006; Li et al., 2012) to estimate emission factors may provide a more reliable and cheaper alternative especially when input data from specific barns is used. Estimates from such models for Canadian dairy cows at the provincial level have carried out (Jayasundara & Wagner-Riddle, 2014) and they need to be validated with direct measurements. The determination of emission factors from livestock buildings requires measurements of gas concentrations and ventilation rates. While several techniques and instruments to measure gas concentrations in livestock buildings have been used and extensively reviewed (Ni & Heber, 2008; Ni et al., 2009; Ogink, Mosquera, Calvet, & Zhang, 2013; Wheeler, Weiss, & Weidenboerner, 2000), measurements of ventilation rates, especially in naturally ventilated buildings (popular in the dairy industry) remains highly difficult. The CO2 balance method is recommended for ventilation rate determination in naturally ventilated buildings but it has a temporal resolution of 24 h, which can be improved if the animal activity is known (CIGR, 2002). Several methods have been applied to obtain values for the animal activity: modelling (CIGR, 2002; Cornou & Lundbye-Christensen, 2012), infrared detectors (Pedersen & Pedersen, 1995) and cameras (Costa, Borgonovo, Leroy, Berckmans, & Guarino, 2009). There is a need for easily used methods that can also improve the spatial resolution of the measured activity since most detectors and sensors have a limited field of view. A system that is increasingly used in highly automated barns is the ALPRO™ herd management system which is incorporated into the DeLaval system (Tumba, Sweden). Amongst its features is an activity monitoring system that is used to determine when animals are in oestrus so as to identify optimal insemination times and for monitoring animal health deviations. It is a promising tool for emission research which may provide high spatial resolution data for animal activity to be used in ventilation rate determination by the CO2 balance method. This is because it measures the activity of each animal that is fitted with a collar. Given the considerations mentioned above, measurements were conducted in a naturally ventilated barn for dairy cows during the spring and the fall transitional seasons with the aim of (a) studying diurnal variations (variations within a day) in the emissions the CH4, N2O and NH3; (b) determining the emission factors of these gases through direct measurements; and (c) comparing CH4 emission factors obtained through direct measurements and through modelling using the IPCC (2006) Tier 2 method with local barn data.
2.
Materials and methods
2.1.
Dairy cow barn
Measurements were carried out during the transitional seasons in the spring (FebruaryeApril) and in the fall (September and October) of 2012 in a dairy cow barn located
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
near Drayton (Latitude 43 450 18.2200 N and Longitude 80 400 16.6000 W) in Southern Ontario, Canada (Fig. 1). The barn had an automatic milking system with open access to the milking machine. The barn was divided into five sections based on the animal category. The main barn had sections for high milk producing cows, low milk producing cows and for replacement heifers. A corridor connected the main barn to a side barn. The corridor housed transition cows while the side barn contained dry cows. All sections in the barn had cubicles on an elevated platform where the cows rested. The bedding material had a rubber mat base which was covered with a layer of straw during the spring measurements. During the autumn measurements, the rubber mat was covered with a layer of solid digestate as bedding material. The solid digestate was obtained by mechanically compressing digestate obtained from a biodigester which became operational during the autumn. The main barn had a solid concrete floor which was mechanically scraped six times a day (12 a.m., 4 a.m., 8 a.m., 12 p.m., 4 p.m., 7 p.m.), with each cleaning session lasting 1 h. The slurry was dumped into a covered pit at one end of the barn. The pit had an outdoor extension from which the manure held for 2e3 weeks was pumped into a nearby storage tank in the spring or into a biodigester in the fall. The dry and transition cows were on slated floors. The slats were over a manure channel with an outdoor extension for temporary manure storage before being pumped into the storage tank or the biodigester. The barn was naturally ventilated through automatically regulated curtains on the side walls. Additional ventilation during warm weather was achieved by opening the doors. Ceiling fans in the barn were not operated during the measurement period.
2.2.
105
Production and feed information
The barn had Holstein dairy cows with production information shown in Table 1. There was no fixed feeding schedule with feed delivered 1e2 times per day for the milking cows and every other day for the heifers. The dry cows were fed every 2e3 days. The feed was a mixed ration with composition shown in Table 1. The animal mass was obtained from production sheets provided by management.
2.3.
Gas concentrations and climate data
The concentrations of CH4, NH3, N2O, and CO2 were measured using a photo-acoustic multi-gas analyser 1309 (Lumasense Technologies SA, Ballerup, Denmark). The analyser was calibrated by the manufacturer prior to installation in the barn. The detection thresholds of the gases were: 0.2 ppm NH3, 0.03 ppm N2O, 0.4 ppm CH4 and 1.5 ppm CO2, with an accuracy of ±2e3%. Gas concentrations were measured at 3 indoor locations and 2 outdoor locations. The indoor sampling locations (L3eL5 in Fig. 1) at a height of 2 m above the floor were chosen to cover the different groups of cows. The hourly means of all measured values at all the indoor locations were used as the indoor concentrations. The outdoor locations were chosen along either side of the barn at a height of about 2 m above the ground. The background concentration for each hour was considered to be the lower of the two values from both outdoor locations as it was most likely a better representation of the air entering the barn due to changing wind directions. Air was sampled from the various locations using 10e30 m long polytetrafluoroethylene (PTFE) tubes with an inner diameter of 3.2 mm. Air was
Fig. 1 e Layout of the dairy cow barn where measurements were conducted (diagrams not to scale).
106
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Table 1 e Production information and composition in rations fed to dairy cows during the period when emission measurements were conducted. Production information and ingredient in the feed
High producing cows
Low producing cows
Transition cowsa
Dry cowsb
Breeding heifersc
% of Dry matter Production information Animal number Animal mass, kg Milk yield, kg head1 d1
77 (66)d 650 32
Feedstuff in the ration (% of dry matter) Corn silage Alfalfa-grass silage Grass hay Straw High moisture corn Custom supplemente Minerals
39.4 32.2 2.0 2.1 10.3 14.1 0
45.1 39.0 2.2 2.3 0 11.4 0
23.6 18.2 0 20.5 4.9 32.8 0
40.4 32.3 16.2 0 0 10.1 1.0
41.9 36.4 0 13.7 0 6.8 1.2
Forage % Concentrate %
75.7 24.3
88.6 11.4
62.3 37.7
88.9 11.1
92.0 8.0
a b c d e
55 700 19
14 600 25, (28)d
30 700 e
45 400 e
Up to three weeks after calving. 3 Months to calving. 11 Months to first insemination. Data during the fall measurement period. The average weight of the cattle was obtained from management reports. Purchased from a commercial feed supplier (contained 34% crude protein, 6% fat, 4.5% crude fibre, minerals and vitamins).
continuously drawn from all five locations by a vacuum pump (KNF Neuberger Inc., Trenton, NJ, USA) through an 8valve manifold (Campbell Scientific, Inc., Logan, UT, USA) at a collection rate of about 10 l min1. The valve manifold was controlled by a CR-1000 datalogger (Campbell Scientific, Inc., Logan, UT, USA), which was programmed to sequentially open one valve for the acoustic gas analyser to subsample. Air from locations L3 and L4 (Fig. 1), which covered the main area of the barn with the highest cow density (170 of the total 215 cows) was sampled for 15 min each. Air was sampled for 10 min from location L5 in the side barn which had 45 cows and for 10 min from each of the two outdoor locations. Gas concentrations for the first 4 min of sampling in each location were discarded to account for a complete flushing of the acoustic analyser when the valve changed from one location to the next. The PTFE tube inlets were fitted with 1 mm pore size membrane filters (Millipore Corporation, Billerica, MA, USA) to trap dust particles. The average air temperature and relative humidity were measured every 5 min using Tiny-Tag data loggers with 10K NTC thermistor and capacitive sensors (Gemini Data Loggers, Chichester, UK). The sensors were calibrated by the manufacturer before deployment. Three data loggers were placed close to the gas sampling locations inside the barn, and two data loggers were placed outside the barn (Fig. 1). A cup anemometer (F460, Climatronics Corp., Newton, PA, USA) and a wind vane (R.M. Young, Model 05102, Traverse, MI, USA) that measured wind speed and direction, respectively, where mounted on a tower 100 m away from the barn and at a height of about 4 m from the ground. The wind data was recorded at 1 min intervals on a 21X datalogger (Campbell Scientific, Inc., Logan, UT, USA). Hourly mean values of the climate data were used in the analyses.
2.4.
Ventilation and emission rates
The ventilation rate in animal barns was estimated by calculating the mass balance of a gas with a known source such as CO2. The CO2 source strength was estimated through empirical equations based on production level and animal characteristics (CIGR, 2002). This approaches assumes CO2 is produced from the animals with negligible production from manure or other sources, no consumption of CO2 within the barn, ideal mixing of the air within the barn and a presentative sampling of CO2 concentration. Given these assumptions, the difference between indoor and outdoor concentrations was related to its production rate and hence barn ventilation rate (Blanes & Pedersen, 2005; Ni, Vinckier, Hendriks, & Coenegrachts, 1999; Pedersen et al., 2008): VR ¼ VRHPU Ftot CF
(1)
where VRHPU is the ventilation rate per heat producing unit (m3 h1 HPU1), Ftot is the heat produced by the cows and heifers (W), and CF is a correction factor for the heat produced at any temperature Ti ( C). CF ¼ 1 þ 0:00004ð20 Ti Þ3 VRHPU ¼
0:21ðRAÞ ðCO2 indoors CO2 outdoors Þ106
(2)
(3)
In Eq. (3), 0.21 is the CO2 production (m3 h1 HPU1) and 1 HPU is 1000 W of heat produced by the animals at 20 C, RA is the hourly relative animal activity measured with an activity monitoring system (see below), CO2indoors is the indoor CO2 concentration (ppmv), and CO2outdoors is the outdoor CO2 concentration (ppmv). The total heat (Ftot ¼ Fdairy
107
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
þ Fheifers) produced by the dairy cows and heifers was calculated using CIGR (2002) equations: (4)
or decreased animal movement in the barn. Information regarding the schedule of management routines that influenced animal movement was equally obtained from the farmer.
(5)
2.6.
cows
Fdairy cow ¼ 5:6m0:27 þ 22Y Fheifer ¼ 7:64m0:69 þ G
23 57:27 þ 0:302m 1 E 1 0:171G
Modelling methane emissions
where m (kg) is the average mass of a dairy cow or heifer, Y (kg d1) is the average daily milk production, G (kg d1) is the daily gain in weight of a heifer (0.55 kg d1 measured in this barn) and E is the average metabolisable energy content of the feed (11 MJ kg1 dry matter), which was assumed to be constant. Eqs. (4) and (5) were calculated for each animal in a category and multiplied by the total number of animals in the category. The emission rate of a gas was the product of the enhanced concentration (difference between indoor and outdoor concentrations expressed in g m3) and the ventilation rate.
Enteric CH4 emission rate was estimated for the dairy cows according to the IPCC (2006) Tier 2 method using the actual productivity parameters and feed characteristics for the animals in the barn (e.g. live weight, milk production data, dry matter, nutrients and gross energy intake) and the predicted methane conversion factor (Ym) according to Ellis et al. (2007).
2.5.
Gas concentrations measured inside and outside the barn are presented in Table 2. Indoor gas concentrations were higher in the spring than in the fall, partly due to the higher animal population in the spring (11 more cows in the spring than in the autumn e Table 1). Furthermore, an expected lower ventilation rate in the spring (which was colder, 7.4 C) than the fall (which was warmer, 10.3 C) might also explain the higher gas concentrations in the spring. The background concentrations were within the range of values reported in another study that used similar instrumentation and sampling technique in a dairy cow barn: 431 ppm for CO2, 5.7 ppm for CH4, 0.71 ppm for NH3 and 0.35 ppm for N2O (Wu, Zhang, & Kai, 2012). It is expected that emissions from nearby sources (manure tank, farm land and machinery) are likely to influence outdoor concentrations. Delayed tracking of NH3 concentrations by the acoustic analyser (Rom & Zhang, 2010) and adsorption/desorption in the sampling lines (Mukhtar et al., 2003) may have also influenced concentrations measured, especially from the background locations. The variations in the average hourly concentrations indicated low concentrations in the daytime and high concentrations at night (Fig. 2). This shows that a better indoor air quality is achievable in naturally ventilated animal barns within the daytime as compared to the night-time. The lower concentrations in the daytime relative to night-time are related to higher daytime ventilation rates. The concentrations of all the gases never exceeded the Ontario provincial time-weighted average exposure limits, which are 5000 ppm
Animal activity
The activity or movement of the cows was measured with an activity monitoring system which is incorporated in the ALPRO™ dairy herd management system (Tumba, Sweden). Animal activity derived by the activity monitoring system is primarily used to determine when animals are in heat so as to identify optimal insemination time and for monitoring animal health deviations. The system consists of a compact collarmounted activity meter around the neck of each cow. The collar has a sensor that detects motion when a magnetic ball moves in a cavity surrounded by copper coils. The movement of the ball is converted into an electrical potential and the impulse is registered, stored and remotely transmitted to an antenna installed in the barn for onward processing by the ALPRO™ software in a computer. With the exception of the heifers, all the dairy cows (176 in the spring and 165 in the autumn) were mounted with collars that measured animal activity. The activity is recorded as either 1 or 0 within each 14 s that the animal moves, resulting in a maximum of 255 counts for each hour. A sum of the counts within each hour represents the hourly activity. The total activity for each hour in the barn was calculated as the mean activity for all the cows. The relative activity on a diurnal basis was calculated as the ratio of the activity for a specific hour and the mean activity for each day. In addition to direct measurements of animal movement using activity meters, qualitative physical observations were made as to the time of day with increased
3.
Results and discussion
3.1.
Gas concentrations and climate data
Table 2 e Gas concentrations and environmental data during measurements in the dairy cow barn (mean ± standard deviation). Gas/climate parameter
Spring measurements Indoor
CO2, ppm CH4, ppm NH3, ppm N2O, ppm Air temperature, C Relative humidity, %
1000 ± 335 68 ± 39 3.8 ± 2.1 0.47 ± 0.08 7.4 ± 2.2 76 ± 14
Fall measurements
Outdoor 453 7.9 1.1 0.37 1.0 79
± 43 ± 5.9 ± 0.4 ± 0.03 ± 6.4 ± 22
Indoor 685 ± 35 ± 2.2 ± 0.39 ± 10.3 ± 83 ±
119 13 0.6 0.04 3.9 10
Outdoor 422 8.7 1.3 0.36 7.0 89
± 22 ± 2.2 ± 0.2 ± 0.03 ± 5.3 ± 15
108
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Fig. 2 e Diurnal variations in gas concentrations with values representing hourly averages for the spring (FebruaryeApril).
for CO2, 1000 ppm for CH4, and 25 ppm for NH3 and N2O (OML, 2010).
3.2.
Animal activity and ventilation rates
The activity of the cows showed a distinct diurnal pattern with two peaks (Fig. 3). As expected, the standard deviations show a larger variation in animal activity within the daytime as compared to the nigh-time. Despite the similarity in activity patterns for both seasons, there was a slight difference in the times when the minimum and maximum activities were measured. The minimum activity was measured at about 5 and 6 a.m. in the spring and in the autumn, receptively. The morning activities peaked at different times as well: about 8 and 10 a.m. in the spring and the autumn respectively. However, the evening activities peaked at the same time at about 7 p.m. The time lag in daytime activities for the different seasons could be due to changes in the time when routine duties were initiated in the barn (lighting, examining the cows, cleaning and feeding). It was also possible that changes in daylight saving time might have affected routine duties in the barn, which then affected the movement of the cows. The evening activity was related to animal movements resulting from feeding, milking and routine animal examinations by the workers at the end of the day. Animal activity patterns that are similar to those in the current study have been reported in other barns with the use
of infrared detectors (CIGR, 2002; Ngwabie et al., 2011). Figure 4 shows how the animal activity measured with the DeLaval activity monitoring system compares with measurements obtained using an infrared passive detector in another study (Ngwabie et al., 2011). The barn used by Ngwabie et al. (2011) and the barn used here had similar management systems (milking robots, manure management). Both measurement methods show close similarity in cow activity variation patterns, but for a difference in magnitude. This could be explained by the fact that the infrared sensory technique considers all movements (animals, workers) in its field of view (Pedersen & Pedersen, 1995), and the field of view may not include all the cows in the barn. In addition, the detection of a temperature change against a background (barn floor) when an animal moves might also affect the measurement, especially when an animal moves against a background made up of other animals. On the other hand, the activity monitoring system considers all cows, with no other external factors, indicating the likelihood of a higher spatial resolution measurement. Another advantage is that there is no additional operational cost and time if the system is already mounted and used by the farmer for herd management optimisation. The average ventilation rates during the measurement periods are presented in Table 3. Higher ventilation rates in the autumn than in the spring could be related to the higher indoor temperatures in the autumn (3 C higher) and because the barn doors and side wall curtains were mostly open in the autumn given the higher outdoor air temperature (Table 2).
109
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Fig. 3 e Average diurnal variations in dairy cow activity measured using the DeLaval activity monitoring system during the spring (FebruaryeApril) and the fall (September and October). The vertical lines represent the standard deviations for similar hours throughout the measurement period.
Errors in estimating the ventilation rate could arise from the uncertainty in the CO2 production factor used in Eq. (3) and also from the representativeness of measurement locations due to imperfect mixing of air in the barn (Van Buggenhout et al., 2009; Wu, Zhai, Zhang, & Nielsen, 2012). Discrepancies have been reported in ventilation rates between direct measurements at exhaust fans and the CO2 mass balance of 2e17% (De Sousa & Pedersen, 2004) and 6.5% (Hinz & Linke, 1998a, 1998b). In another study that compared ventilation rates using four methods over 41 d in a mechanically ventilated building for pigs, the CO2 balance method was 8% lower than the ventilation rate measured using a fan (Blanes & Pedersen, 2005). Diurnal variations in the hourly CO2 ventilation rates in the spring and the autumn showed high values in the daytime
and low values at night (Fig. 5). The observed ventilation rate pattern was related not only to the animal activity, but also to other factors such as wind speed (R2 ¼ 0.50 and 0.31 in the spring and fall respectively) and wind direction (R2 ¼ 0.54 and 0.33 in the spring and autumn respectively). Although the wind speed (Table 3) was similar in both seasons, the ventilation rate had a better correlation with values in the spring than in the fall. The construction of a large anaerobic digester tank close to the barn in the autumn (Fig. 1) might have influenced the relationship between the wind and the ventilation rate in the later part of the measurement. The large spike in the autumn ventilation rate at about 10 a.m. could be related to the opening of the side wall curtains during this period due to the high indoor air temperature.
3.3.
Emission rates
3.3.1.
Measured emissions
Diurnal variations were observed in the emissions rates with a peak in the morning and in the evening (Fig. 6). Detailed hourly information of the emission variation was expected in this measurement since the hourly ventilation rates were used in calculating the emissions. Diurnal variations in emissions are common in animal barns (Ngwabie et al., 2011; Zhu et al., 2012) and are related to several factors: ventilation rate, management routines and temperature changes. Spikes in NH3 emissions, which is emitted predominantly from the manure, was related to periods of high urinating/defecating frequency which was observed in the mornings and in the
Table 3 e Barn ventilation rates and outdoor wind speed data during the measurements.
Fig. 4 e A comparison of the animal activity in the present measurements with reported values measured using an infrared passive detector inside a dairy barn.
Spring Fall
Ventilation rate, m3 LU1 h1
Wind speed, m s1
531 ± 452 824 ± 492
3.6 ± 2.0 3.2 ± 1.8
Mean ± standard deviation, 1 LU ¼ 500 kg animal mass.
110
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
evenings. Lower NH3 emissions were measured during the night when the ventilation rate and temperature were lower and the cows were resting. The mean emission rates are presented in Table 4. Despite a 55% increase in the mean CO2 balance ventilation rate from the spring to the fall season, the average emission factors dropped by just 12% for CH4, 33% for NH3 and 29% for N2O over the same period. There was a drop in gas concentrations from the autumn to the fall associated with the increased ventilation rate (Table 2). A change in the CH4 emission factor proportional to ventilation rate was not expected, as it is mostly controlled by animal numbers and diet. The larger changes in NH3 and to an extent, N2O emission factors could be related to some other factors that influence gas concentrations in the barn such as animal number (Table 1), temperature (Table 2) and manure retention time inside the barn. Changes in the supplementary bedding material from straw in the spring to solid digestate in the fall might also have affected the emission rates of NH3 and N2O due to potential changes in the rate of immobilisation for the different bedding materials (Tasistro, Cabrera, Ritz, & Kissel, 2008). Fig. 5 e Diurnal variations in ventilation rates calculated as the average for each hour in the spring (FebruaryeApril) and in the fall (September and October).
3.3.2.
Modelled CH4 emissions
Modelled enteric CH4 emission rates and nutritional composition for the different categories of dairy cows using the IPCC (2006) Tier 2 method is presented in Table 5. Regarding additional local barn data in Table 1, all the rations contained a higher proportion of forage than concentrate, however,
Fig. 6 e Diurnal variations in emission rates from the barn with dairy cows, with values representing hourly averages in the spring (FebruaryeApril) and in the fall (September and October).
111
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Table 4 e Emission rates from the dairy cow barn calculated using the CO2 mass balance ventilation rate. Spring 1
Fall
1
CH4, g LU h 13.9 ± 5.3 (4.5e34.5) 12.2 ± 5.5 (3.0e32.1) NH3, g LU1 h1 0.64 ± 0.32 (0.09e2.06) 0.43 ± 0.22 (0.09e1.47) N2O, mg LU1 h1 41.3 ± 21.7 (0.45e128.93) 29.4 ± 21.7 (0e123.82) 1 LU ¼ 500 kg animal mass. Total LU in the spring and fall were 272 and 258 respectively. Mean ± standard deviation (range).
rations fed to the high producing cows and transition cows had a relatively lower proportion of forages (62e76%) compared with rations fed to low producing cows, dry cows and heifers (89e92% of forages in the ration). This was reflected in the chemical composition of the rations, as both neutral detergent fibre (NDF) and acid detergent fibre (ADF) contents were on average 24% higher in rations for low producing cows, dry cows and heifers compared with rations for high producing and transition cows (Table 5). Average dry matter intake was 67% higher for lactating dairy cows compared with that for dry cows and heifers (19.2 vs. 11.5 kg head1 d1, Table 5). Higher dry matter intake and lower NDF and ADF contents in the ration tend to decrease the fraction of gross energy intake converted to CH4 in dairy cattle (Ellis et al., 2007), therefore predicted CH4 conversion factor (Ym) was on average 17% higher for dry cows and heifers compared with that for lactating dairy cows (0.063 ± 0.012 for dry cows and heifers versus 0.054 ± 0.011 for lactating cows).
3.3.3. Enteric CH4 emission factors derived using the measured emissions Measured CH4 emissions from the dairy barn during the spring and fall measurement campaigns were equivalent to 411 and 360 g head1 d1, respectively resulting in a mean
emission rate of 385 g CH4 head1 d1 over the two seasons (Table 6). If it is assumed that all the CH4 measured in the current study was entirely from enteric fermentation, and that a non-lactating dairy animal (dry cows and heifers > 1 year, 75 out of 216 in this barn) produces about 60% of enteric CH4 relative to that from a lactating cow (Holter & Young, 1992), then an estimated enteric CH4 emission of 447 g head1 d1 per lactating cow can be derived (Table 6). This value is higher than the range of enteric CH4 emissions for lactating dairy cows reported in Canada (329e424 g head1 d1) from a review of 11 studies (Boadi, Benchaar, Chiquette, & Masse, 2004). In comparison, the estimated enteric CH4 emission rate for a lactating dairy cow calculated according to the IPCC (2006) Tier 2 method, but using the actual dry matter, nutrients and gross energy intake and predicted CH4 conversion factor (Ym) according to Ellis et al. (2007) was 363 ± 73 g head1 d1 (Table 6). The higher CH4 emission rate observed from direct measurements in the current study, compared with both previously reported and the current modelled emission rates, may be attributed to uncertainty associated with ventilation estimates (through CO2 balance) and the potential contributions of CH4 emitted from the liquid manure present in the temporary holding pit located in one end of the barn leading to higher CH4 concentrations (Fig. 1). Generally, indoor pits located beneath the floor can temporarily hold liquid dairy manure before being pumped outdoor where the main storage facility is located (StatisticsCanada, 2003). Using laboratory scale storage units, Sommer, Petersen, Sørensen, Poulsen, and Møller (2007) demonstrated that the presence of a small quantity of inoculum (~8%) could support the immediate production of appreciable amounts of CH4 from manure held over 10 d. Therefore CH4 could be generated in liquid manure held temporarily for about 2e3 weeks in under-floor storage pits where complete cleaning after each emptying is usually not carried out (Sommer et al.,
Table 5 e Nutrients composition, dry matter intake (DMI) and gross energy intake (GEI) and estimated rates of CH4 emissions from enteric fermentation by different dairy cattle groups in the barn where emission measurements were undertaken. High producing cows Nutrient composition (g kg1) Crude protein (CP) Ether Extract (EE) Neutral detergent fibre (NDF) Acid detergent fibre (ADF) Ash Non-fibre carbohydrate (NFC)a Total digestible nutrients (TDN), % Dry matter intake (DMI), kg head1 d1 Gross energy intake (GEI) (MJ1 head1 d1)b Methane conversion factor (Ym) (fraction of GEI)c Enteric CH4 production (g CH4 head1 d1)d a b c d
1
Low producing cows
Transition cows
Dry cows
Breeding heifers
182.3 34.3 322.3 198.4 80.5 380.6 76.8
175.3 31.9 375.5 253.6 79.3 338.0 72.8
141.8 37.2 392.2 254.4 74.8 354.0 72.7
116.3 32.2 457.9 295.7 61.4 331.2 67.3
143.1 30.2 428.9 297.1 73.6 324.2 68.8
21.3 385.0 0.052 ± 0.010
19.4 355.1 0.054 ± 0.011
17.0 312.4 0.057 ± 0.011
12.4 228.3 0.063 ± 0.012
11.0 200.9 0.062 ± 0.012
362 ± 73
348 ± 69
320 ± 64
260 ± 52
223 ± 45
Calculated: NFC (g kg ) ¼ 1000 (CP þ EE þ NDF þ Ash) (NRC, 2001). Calculated: GEI (MJ kg1) ¼ DMI [(23.6 CP þ 39.8 ether extract þ 17.3 NFC þ 18.9 NDF)/1000] (Ramin & Huhtanen, 2013). Based on the prediction equation of Ellis et al. (2007) using DMI, NDF and ADF intake. Converted from MJ head1 d1 (assuming an energy density of 55.65 MJ kg1 CH4, IPCC, 2006).
112
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Table 6 e Enteric CH4 emission factors derived for dairy cows using the measured CH4 emissions from the barn. Measured emissions corrected with CH4 from manure
Estimated enteric CH4 emissionsa
83.1 216
83.1 216
67.3 ± 13.5 216
360 NE 360
385 NE 385
385 25.1b 360
312 ± 63 e 312 ± 63
420 252
447 268
418 251
363 ± 73 241 ± 49
Measured emissions Spring
Fall
Average
90.7 221
75.5 210
Emissions per animal (g CH4 head1 d1) All sources 411 Emissions from manure NE Emissions from all cattle 411 Emission portioningc Lactating cows Non-lactating dairy cattle
Total CH4 emissions (kg CH4 d1) Number of animals in the barn
475 285
For units conversions: 141 lactating cows ¼ 187 LU; 75 non-lactating cows ¼ 78 LU. a Estimated using IPCC (2006) Tier 2 method with current barn input data and a 20% uncertainty. NE: no emissions from the manure was assumed. b Average CH4 emission of 25.1 g animal1 day1 from indoor manure pit was assumed according to Kinsman et al. (1995). c Partitioning of total measured emission assumed that non-lactating dairy cow (dry cows and heifers > 1 year, 75 out of 216 in this barn) produced about 60% of enteric CH4 relative to that from a lactating cow (Holter & Young, 1992).
2007; Wood, VanderZaag, Wagner-Riddle, Smith, & Gordon, 2014). However, very little published data is available for CH4 emissions from dairy manure held temporarily in indoor pits in Canadian dairy barns. We are aware of only one study that measured CH4 emissions from the manure holding pit underneath the floor in a dairy barn in Canada which provided an average emission rate equivalent to 25.1 g head1 d1 (about 6% of measured CH4 emissions from the barn) from manure held for about 21 days in an eastern Ontario dairy barn with 118 lactating Holstein cows (Kinsman, Sauer, Jackson, & Wolynetz, 1995). In another similar study in Germany, an average emission rate of 56 g head1 d1 from liquid manure held temporary for 23 d before transferred to the main storage in a dairy barn housing 34 lactating cows and 9 dry cows was measured (Marik & Levin, 1996). Methane formation in liquid manure can be influenced both by residence time and temperature (Sommer et al., 2007). Higher average CH4 emissions reported by Marik and Levin (1996) compared with those reported by Kinsman et al. (1995) for indoor manure held over a similar residence time (approximately 3 weeks) may be due to the differences in temperature where the studies were conducted (average annual temperature 10.5 C at the site studied by Marik and Levin (1996) versus 6.6 C at the site studied by Kinsman et al. (1995). Our study was conducted in a similar temperature zone (with an average annual temperature of 6.7 C) to that of Kinsman et al. (1995). Furthermore, manure was temporary held for about 2e3 weeks in the indoor pit with an outdoor extension, before being pumped to the long-term storage tank in the present study. If we assume that potential contributions of CH4 from manure held underneath the floor in the current study was similar to the average emissions reported by Kinsman et al. (1995) which was 25.1 g head1 d1), then this will result in an average emission rate of 418 g head1 d1 for a lactating cow, which falls within the range of enteric CH4 emission rates reported for lactating dairy cows in Canada (329e424 g head1 d1, Boadi et al., 2004).
Based on the ratio of the modelled enteric CH4 emission (312 g head1 d1) and the mean measured barn CH4 emission (385 g head1 d1) in Table 6, enteric fermentation accounted for about 81% of the barn emission. This showed that the manure could account for 19% of the barn emissions. However if the reported CH4 manure emission of ~25 g head1 d1 (Kinsman et al., 1995) was considered, enteric fermentation will rather account for over 90% of the barn emissions. Our direct and modelled measurements showed that CH4 emissions from the manure in the current barn can be as high as 73 g head1 d1 (2.5 g LU1 h1). Several researchers have measured CH4 and NH3 emissions from naturally ventilation barns in the cold climatic regions of North America and Europe. Table 7 presents average values of some reported emissions where enteric fermentation was not separated from manure emissions and how they compare to the present study. Generally, the measured CH4 emission factor was within the range of reported values while the measured NH3 emission was somewhat lower than reported values. Differences in emission factors are likely related to differences in regional climate, feed composition, productivity level, building and the management systems. There is little available information of N2O emissions from dairy barns as the emission factor is expected to be small e.g. 0.1 g N2O LU1 d1 have been measured in a barn with lactating cows (Adviento-Borbe et al., 2010). The higher N2O emissions in the current study, equivalent to 0.85 g LU1 d1 could be due to favourable conditions for its production resulting from the mixing of bedding material with the manure.
4.
Conclusions
This study used animal activity obtained from an activity monitoring system in automated barns for dairy cows to model the ventilation rate on an hourly basis using the CO2 balance method. The emission factors of CH4, NH3 and N2O were subsequently generated and their diurnal variation
113
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Table 7 e Summary CH4 and NH3 emission factors from naturally ventilated dairy barns. CH4 12.2e13.9 g LU1 h1 10.60a g LU1 h1 18.1b g LU1 h1 10.8e11.4 g LU1 h1 14.5e17.9 g LU1 h1 e
NH3
Country, ventilation type
Source
0.43e0.64 g LU1 h1 e e 0.81e1.0 g LU1 h1 2.9e3.5 g LU1 h1
Ontario e Canada, CO2 mass balance e Alberta e Canada, dispersion model Sweden, CO2 mass balance Germany, CO2 mass balance
0.3e2.8 g LU1 h1
Switzerland, Tracer ratio
Present direct measurement Present modelled results (McGinn & Beauchemin, 2012) (Ngwabie et al., 2011) (Samer, Berg, et al., 2011; Samer, Fiedler, et al., 2011) (Schrade et al., 2012)
1 LU ¼ 500 kg animal mass. a Calculated from the total modelled enteric emission of 67,336 g CH4 day1 (Table 5) and a mean live unit of 265 LU in Table 4. b Recalculated from 15.1 g cow1 h1 and 600 kg of cow mass.
patterns were studied. Enteric CH4 emission factor was also generated using the IPCC (2006) Tier 2 method with local barn input data. The following conclusions were drawn from this study: The activity monitoring system provides a high resolution measurement of animal activity (movement) since all animals are considered. It has the potential to improve the temporal resolutions of the ventilation rate from 24 to 1 h when calculated using the CO2 balance method and the resulting emission variation pattern. The mean ventilation rates were 531 and 824 m3 LU1 h1 in the spring and fall respectively with pronounced diurnal variations. Diurnal variations were observed in the emissions rates with a peak in the morning and in the evening that were related to management routines and climatic conditions in the barn. The peaks were more prominent for CH4 and NH3 than for N2O. The average measured emission factors were 13.9 g CH4 LU1 h1, 0.64 g NH3 LU1 h1 and 41.3 mg N2O LU1 h1 in the spring and 12.2 g CH4 LU1 h1, 0.43 g NH3 LU1 h1 and 29.4 mg N2O LU1 h1 in the fall. Emissions factors dropped by 12% for CH4, 33% for NH3 and 29% for N2O from the spring to the fall. The average measured barn CH4 emission factor from all sources and the entire dairy cow category was equivalent to 385 g head1 d1 (13.08 g LU1 h1) while the modelled enteric emission factor was 312 g head1 d1 (10.58 g CH4 LU1 h1). An attempt at source apportioning indicated that indoor manure could account for up to 73 g head1 d1 (2.5 g CH4 LU1 h1). A modelled lactating cow emitted about 363 g head1 d1 (11.08 g CH4 LU1 h1) while a modelled non-lactating cows emitted about 241 g head1 d1 (9.67 g CH4 LU1 h1). The ratio of the modelled enteric CH4 emission to the measured barn CH4 emission indicated that enteric fermentation accounted for about 81% of the total barn emission. Additional activity measurements using other methods alongside the activity monitoring system is recommended for full validation of its use for ventilation rate calculations. Further research using direct measurement methods needs to be carried to determine the contribution of indoor manure to barn CH4 emissions.
Acknowledgement Sponsorship for this projected was provided by the Agricultural Greenhouse Gas Program funded by Agriculture and Agri-Food Canada- project number 1585-16-3-1-14. Jordan Forsyth, the proprietor and management of Clovermead Dairy Farm were indispensable in the execution of this project.
references
Adviento-Borbe, M. A. A., Wheeler, E. F., Brown, N. E., Topper, P. A., Graves, R. E., Ishler, V. A., et al. (2010). Ammonia and greenhouse gas flux from manure in freestall barn with dairy cows on precision fed rations. Transactions of the ASABE, 53, 1251e1266. Blanes, V., & Pedersen, S. (2005). Ventilation flow in pig houses measured and calculated by carbon dioxide, moisture and heat balance equations. Biosystems Engineering, 92, 483e493. Bluteau, C. V., Masse, D. I., & Leduc, R. (2009). Ammonia emission rates from dairy livestock buildings in Eastern Canada. Biosystems Engineering, 103, 480e488. Boadi, D., Benchaar, C., Chiquette, J., & Masse, D. (2004). Mitigation strategies to reduce enteric methane emissions from dairy cows: update review. Canadian Journal of Animal Science, 84, 319e335. CIGR. (2002). Climatization of animal houses. Forth report of working group. Danish Institute of Agricultural Sciences, Research Centre Bygholm. Cornou, C., & Lundbye-Christensen, S. (2012). Modeling of sows diurnal activity pattern and detection of parturition using acceleration measurements. Computers and Electronics in Agriculture, 80, 97e104. Costa, A., Borgonovo, F., Leroy, T., Berckmans, D., & Guarino, M. (2009). Dust concentration variation in relation to animal activity in a pig barn. Biosystems Engineering, 104, 118e124. De Sousa, P., & Pedersen, S. (2004). Ammonia emission from fattening pig houses in relation to animal activity and carbon dioxide production. Agricultural Engineering International: The CIGR Ejournal, 6. Manuscript BC 04 003. Ellis, J. L., Kebreab, E., Odongo, N. E., McBride, B. W., Okine, E. K., & France, J. (2007). Prediction of methane production from dairy and beef cattle. Journal of Dairy Science, 90, 3456e3466. Environment-Canada. (2012). National inventory report 1990e2010: Greenhouse gas sources and sinks in Canada e Executive summary. Government of Canada.
114
b i o s y s t e m s e n g i n e e r i n g 1 2 7 ( 2 0 1 4 ) 1 0 3 e1 1 4
Harper, L. A., Flesch, T. K., Powell, J. M., Coblentz, W. K., Jokela, W. E., & Martin, N. P. (2009). Ammonia emissions from dairy production in Wisconsin. Journal of Dairy Science, 92, 2326e2337. Hinz, T., & Linke, S. (1998a). A comprehensive experimental study of aerial pollutants in and emissions from livestock buildings. Part 1. Methods. Journal of Agricultural Engineering Research, 70, 111e118. Hinz, T., & Linke, S. (1998b). A comprehensive experimental study of aerial pollutants in and emissions from livestock buildings. Part 2: results. Journal of Agricultural Engineering Research, 70, 119e129. Holter, J. B., & Young, A. J. (1992). Methane prediction in dry and lactating Holstein cows. Journal of Dairy Science, 75, 2165e2175. IPCC. (2006). Guidelines for national greenhouse gas inventories. In Emissions from livestock and manure management (Vol. 4). AFOLU Intergovernmental Panel on Climate Change. Jayasundara, S., & Wagner-Riddle, C. (2014). Greenhouse gas emissions intensity of Ontario milk production in 2011 compared with 1991. Canadian Journal of Animal Science, 94, 155e173. Kinsman, R., Sauer, F. D., Jackson, H. A., & Wolynetz, M. S. (1995). Methane and carbon dioxide emissions from dairy cows in full lactation monitored over a six-month period. Journal of Dairy Science, 78, 2760e2766. Li, C. S., Salas, W., Zhang, R. H., Krauter, C., Rotz, A., & Mitloehner, F. (2012). Manure-DNDC: a biogeochemical process model for quantifying greenhouse gas and ammonia emissions from livestock manure systems. Nutrient Cycling in Agroecosystems, 93, 163e200. Marik, T., & Levin, I. (1996). A new tracer experiment to estimate the methane emissions from a dairy cow shed using sulfur hexafluoride (SF6). Global Biogeochemical Cycles, 10, 413e418. McGinn, S. M., & Beauchemin, K. A. (2012). Dairy farm methane emissions using a dispersion model. Journal of Environmental Quality, 41, 73e79. McGinn, S. M., Flesch, T. K., Harper, L. A., & Beauchemin, K. A. (2006). An approach for measuring methane emissions from whole farms. Journal of Environmental Quality, 35, 14e20. Mukhtar, S., Rose, A., Capareda, S., Boriack, C., Lacey, R., Shaw, B., et al. (2003). Assessment of ammonia adsorption onto teflon and LDPE tubing used in pollutant stream conveyance. Agricultural Engineering International: The CIGR Ejournal., 5. Manuscript BC 03 012. Ngwabie, N. M., Jeppsson, K. H., Gustafsson, G., & Nimmermark, S. (2011). Effects of animal activity and air temperature on methane and ammonia emissions from a naturally ventilated building for dairy cows. Atmospheric Environment, 45, 6760e6768. Ni, J. Q., & Heber, A. J. (2008). Sampling and measurement of ammonia at animal facilities. In Advances in agronomy (Vol. 98, pp. 201e269). San Diego, USA: Elsevier Academic Press Inc. Ni, J. Q., Heber, A. J., Darr, M. J., Lim, T. T., Diehl, C. A., & Bogan, B. W. (2009). Air quality monitoring and on-site computer system for livestock and poultry environment studies. Transactions of the ASABE, 52, 937e947. Ni, J., Vinckier, C., Hendriks, J., & Coenegrachts, J. (1999). Production of carbon dioxide in a fattening pig house under field conditions. II. Release from the manure. Atmospheric Environment, 33(22), 3697e3703. NRC. (2001). National Research Council. Nutrient requirements of dairy cattle (7th revised ed.). Washington, DC: The National Academies Press. Ogink, N. W. M., Mosquera, J., Calvet, S., & Zhang, G. (2013). Methods for measuring gas emissions from naturally ventilated livestock buildings: developments over the last decade and perspectives for improvement. Biosystems Engineering, 116(3), 297e308.
OML (Ontario Ministry of Labour). (2010). Occupational exposure limits for Ontario workplaces. Pedersen, S., Blanes-Vidal, V., Joergensen, H., Chwalibog, A., Haeussermann, A., Heetkamp, M. J. W., et al. (2008). Carbon dioxide production in animal houses: a literature review. Agricultural Engineering International, 10. Manuscript BC 08 008. Pedersen, S., & Pedersen, C. B. (1995). Animal activity measured by infrared detectors. Journal of Agricultural Engineering Research, 61, 239e246. Ramin, M., & Huhtanen, P. (2013). Development of equations for predicting methane emissions from ruminants. Journal of Dairy Science, 96, 2476e2493. Rom, H. B., & Zhang, G. Q. (2010). Time delay for aerial ammonia concentration measurements in livestock buildings. Sensors, 10, 4634e4642. Samer, M., Berg, W., Muller, H. J., Fiedler, M., Glaser, M., Ammon, C., et al. (2011). Radioactive (85)kr and CO2 balance for ventilation rate measurements and gaseous emissions quantification through naturally ventilated barns. Transactions of the ASABE, 54, 1137e1148. Samer, M., Fiedler, M., Muller, H. J., Glaser, M., Ammon, C., Berg, W., et al. (2011). Winter measurements of air exchange rates using tracer gas technique and quantification of gaseous emissions from a naturally ventilated dairy barn. Applied Engineering in Agriculture, 27, 1015e1025. Schrade, S., Zeyer, K., Gygax, L., Emmenegger, L., Hartung, E., & Keck, M. (2012). Ammonia emissions and emission factors of naturally ventilated dairy housing with solid floors and an outdoor exercise area in Switzerland. Atmospheric Environment, 47, 183e194. Sheppard, S. C., Bittman, S., Swift, M. L., Beaulieu, M., & Sheppard, M. I. (2011). Ecoregion and farm size differences in dairy feed and manure nitrogen management: a survey. Canadian Journal of Animal Science, 91, 459e473. Sommer, S. G., Petersen, S. O., Sørensen, P., Poulsen, H. D., & Møller, H. B. (2007). Methane and carbon dioxide emissions and nitrogen turnover during liquid manure storage. Nutrient Cycling in Agroecosystems, 78, 27e36. Statistics-Canada. (2003). Manure storage in Canada (Vol. 1). No. 1. Catalogue no. 21-021-MIE. Tasistro, A. S., Cabrera, M. L., Ritz, C. W., & Kissel, D. E. (2008). Manipulating bedding materials and PLT (TM) to reduce NH3 emissions from broiler manure. Bioresource Technology, 99(6), 1952e1960. Van Buggenhout, S., Van Brecht, A., Ozcan, S. E., Vranken, E., Van Malcot, W., & Berckmans, D. (2009). Influence of sampling positions on accuracy of tracer gas measurements in ventilated spaces. Biosystems Engineering, 104, 216e223. Wheeler, E. F., Weiss, R. W. J., & Weidenboerner, E. (2000). Evaluation of instrumentation for measuring aerial ammonia in poultry houses. Journal of Applied Poultry Research, 9, 443e452. Wood, J. D., VanderZaag, A. C., Wagner-Riddle, C., Smith, E. L., & Gordon, R. J. (2014). Application of a high-temporal resolution method to estimate ammonia emissions from farmland. Nutrient Cycling in Agroecosystems, 99, 95e105. Wu, W. T., Zhai, J., Zhang, G. Q., & Nielsen, P. V. (2012). Evaluation of methods for determining air exchange rate in a naturally ventilated dairy cattle building with large openings using computational fluid dynamics (CFD). Atmospheric Environment, 63, 179e188. Wu, W. T., Zhang, G. Q., & Kai, P. (2012). Ammonia and methane emissions from two naturally ventilated dairy cattle buildings and the influence of climatic factors on ammonia emissions. Atmospheric Environment, 61, 232e243. Zhu, Z., Dong, H., & Zhou, Z. (2012). Ammonia and greenhouse gas emissions from a dairy cattle barn with a daily manure collection system. Transactions of the ASABE, 55, 1959e1964.