Degree Type


Date of Award


Degree Name

Doctor of Philosophy


Agricultural and Biosystems Engineering

First Advisor

Hongwei Xin


The U.S. swine industry has undergone significant changes over the past 50 years. The industry has moved from many, smaller scale producers with multiple production stages on one farm to fewer, larger scale producers with production stages separated out to multiple farms. As pig production has consolidated, large sow facilities have become more common. These large sow farms have environmental concerns at the animal, barn, and ecosystem level. To address the questions on current breeding/gestation and farrowing environments the studies detailed in this dissertation were developed. This dissertation covers 33 months of continual monitoring at a commercial swine breeding-gestation-farrowing facility. The first paper characterizes the gaseous concentrations and emissions of two breeding/gestation barns, two farrowing rooms, and external manure storage for a 29-month period. These data fill a gap in the U.S. NH3 and GHG emissions inventory. The second paper quantifies the heat and moisture production rates of breeding/gestating sows and lactating sows with litters for a 16-month period. These data will help updating the standards for engineering design and operation of modern swine housing. The third paper compares heat mat vs. heat lamp as localized heating source for prewean piglets for three farrowing rooms over a 12-month period (16 farrowing cycles). The fourth paper compares two gas analyzers, a Fourier transform infrared spectrometer (FTIR) and a photoacoustic infrared spectrometer (PAS), for the field measurement of NH3 and GHG concentrations over a 5-month period in a swine facility.

A 4,300-sow farm was selected for the extensive field monitoring which employed a Mobile Air Emission Monitoring Unit equipped with state-of-the-art gas analyzers and a data acquisition system. The monitored portion of the facility consisted of a deep-pit breeding/early gestation barn (B/EG, 1800 head, 204 ±3.2 kg hd-1 (mean ±SE)), the deep-pit late gestation barn (LG, 1800 head, 219 ±3.0 kg hd-1), and two shallow-pit (pull-plug) farrowing rooms (40 sow/litter per room, 223 ±0.4 kg hd-1). A dynamic flux chamber was used to monitor gaseous emissions from the external manure storage for the farrowing rooms. The total heat production (THP) was determined using indirect animal calorimetry, latent heat production (LHP) was determined from mass balance, and sensible heat production (SHP) was calculated as the difference between THP and LHP. Three 40-crate farrowing rooms were equipped with 125W heat lamps in half of the crates and 290W 0.6m x 1.5m (2ft x 5ft) double heat mats shared between two crates in the other half of the crates. A temperature dependent, variable output controller regulates the power supply to the mats. The lamps were controlled on/off by the room ventilation system controller and turned off when the room temperature exceeded the set point by 5.5°C. Electricity usage of each half-room was measured separately with electric meters, and piglet performance was recorded by farm personnel and our research group. Additionally, infrared thermographs were taken for a 24-hr period several times during the lactation period to capture the heat source utilization by the piglets. The FTIR and PAS were installed side-by-side in the MAEMU and operated for 5 months under a range of durations per sample location (120s, 240s, 360s) and FTIR sample integration times (30s, 60s). The response time of the analyzers to known gas concentrations was also tested in a laboratory setting.

Results from the first paper show daily indoor NH3, CO2, N2O, and CH4 concentrations (ppm, mean ±SD) were 12.0 (±7.6), 1594 (±797), 0.31 (±0.11), and 28.5 (±9.8), respectively, in the breeding/gestation barns; and 9.7 (±4.1), 1536 (±701), 0.30 (±0.10), and 78.3 (±37), respectively, in the farrowing rooms. Daily emissions per animal unit (AU, 500 kg live weight) were 35.1 g NH3, 7.46 kg CO2, 0.17 g N2O, and 263.4 g CH4 for sows in the B/EG barn; and 28.2 g NH3, 6.50 kg CO2, 0.12 g N2O, and 201.3 g CH4 for sows in the LG barn. The average daily emissions per AU (sow and piglets) of the farrowing rooms during the lactation period (birth to weaning) were: 59.7 g NH3, 16.4 kg CO2, 0.73 g N2O, and 107 g CH4. For the monitored period, the external manure storage had the following average daily emission per m2 surface area: 1.26 g NH3, 137 g CO2, and 94.8g CH4, which was equivalent to daily emissions per AU in the farrowing rooms of 12.2 g NH3, 1055 g CO2, and 867 g CH4. The whole-farm average daily emissions per AU were 38.5 g NH3, 8.73 kg CO2 (including 7.3 kg from animal respiration), 0.25 g N2O, and 301 g CH4.

Results from the second paper show that THP at 20°C averages 1.8 W/kg for sows in the breeding/early gestation stage, 1.5 W/kg for sows in the late gestation stage, and 3.9 W/kg for sows and litters in week 0 of the lactation stage. The corresponding house-level LHP for the three stages averages 0.7 W/kg (early gestation), 0.6 W/kg (late gestation), and 2.1 W/kg (lactation, week 0). Finally the corresponding house-level SHP for the three stages averages 1.1 W/kg (early gestation), 0.9 W/kg (late gestation), and 1.8 W/kg (lactation, week 0). Compared with the ASABE standards, values from the current study for gestation sows in their early and late pregnancy stages showed increases of 28% and 8% in THP, 53% and 22% in LHP, and 16% and 2% in SHP, respectively. Values for lactating sows and litters during the first week after parturition showed increases of 23% in THP, 48% in LHP, and 11% in SHP relative to the ASABE standards. The reductions of THP from day to night for the three stages were 32% (early gestation), 27% (late gestation), and 7% (lactation). These data will help updating the standards for engineering design and operation of modern swine housing.

Results from the third paper found the average body weight gain (mean ±SE) of piglets in the mat and lamp regimens was, respectively, 224 (±5.7) g/d and 220 (±5.9) g/d. Prewean mortality (mean ±SE) for the mat and lamp regimens were, respectively, 7.8% (±0.4%) and 7.4% (±0.5%). Power use (mean ±SE) for the mat and lamp regimens was respectively, 0.66 (±0.06) kWh and 1.05 (±0.04) kWh per kg weaned pig. Overall, the heat sources were occupied for 58% and 56% of the time for mats and lamps, respectively. When the heat source was utilized, at least two piglets were present 76% and 87% of the time for mats and lamps, respectively. Overall, the mats and lamps performed similarly except for power use.

Results from the fourth paper found the FTIR and PAS had good agreement for NH3, CO2, and CH4 field measurements. The linear regression slopes for FTIR vs. PAS ranged from 1.002 to 1.052 for NH3, 0.980 to 1.002 for CO2, and 0.996 to 1.017 for CH4. The N2O concentrations were < 0.8 ppm on the PAS and < 0.6ppm on the FTIR and the two analyzers had poor agreement at the individual sample levels. The relative difference between FTIR and PAS concentrations was generally larger at lower concentrations, decreased sample location times, and large indoor-ambient concentration differences. The PAS had the fastest response times to T98 (time taken to display 98% of known concentration) for all gases, followed by the FTIR at 30s sample integration time. The FTIR at 60s sample integration time had the longest response times. This study revealed that the FTIR is comparable to the PAS for NH3, CO2, and CH4 measurements, although care must be taken when there exist large changes from location to location to allow sufficient time for the FTIR to respond. Further investigation of the instruments at higher N2O concentrations is needed to quantify their respective performance.

Overall, findings of these studies are beneficial to the swine industry by providing an environmental assessment of a Midwestern U.S. breeding-gestation-farrowing system, as well as to the advancement of the scientific knowledge. The gaseous emissions will helpful to the development and application of mitigation technologies. The new data on heat and moisture production rates will help updating the current ventilation design standards and allow for more precise environmental control of the production facilities. The heat source comparison demonstrates the similar results (piglet performance, piglet utilization) of both heat mat and heat lamp while indicating that further farrowing crate design modifications may be beneficial to piglet performance. The analyzer comparison demonstrates the suitability of the FTIR for animal air quality work while outlining situations that may be problematic due to the FTIR response time.

Copyright Owner

John Stinn



File Format


File Size

155 pages