BUILDING ENVIRONMENTAL PREDICTION MODEL FOR SWINE GESTATION BARNS
2019-02-12T18:22:05Z (GMT) by
There are over six million gestation sows in the United States and most of them are kept in gestation stalls. The inside environment of large livestock buildings requires advanced environmental control systems to maintain animal health and optimize animal production efficiency. The ventilation rate, inside temperature, and supplemental heating and cooling are the main control variables to manage the barn environment. About 144 barn-months of unpublished thermal data were obtained from six commercial gestation houses by the National Air Emission Monitoring Study (NAEMS). The data from this site was reviewed, corrected, and re-analyzed to improve its quality, completion, accuracy and reliability using the methods of comparison between onsite measurements and data collected from nearest weather stations, introducing corrected models to adjust the onsite data, substituting invalid and missing onsite data by weather station data and other improved methodologies. The data completeness for solar radiation, relative humidity, atmospheric pressure, outside temperature, and wind speed and direction were increased by 5.6 to 17.9%. The six NAEMS gestation barns were used to test and validate a building environmental prediction model (BEPM) based on known thermodynamic and heat transfer principles for simultaneously predicting inside temperatures and ventilation rates. The BEPM inputs included the weather, the building dimensions and materials, geographical location and building orientation, and sow herd characteristics. Predictions of ventilation rates and inside temperatures followed the expected yearly patterns as the measured NAEMS data. Four combinations of heat production rate and inside temperature submodel combinations CIGR-T, CIGR-T2, US-T, and US-T2 were compared and evaluated based on the root-mean-square-deviation and fitness tests to determine the best submodel combination. The average predicted and measured means of ventilation rate were 24.8 and 24.1 m3/s for NAEMS Site IA4B, 27.5 and 24.9 m3/s for Site NC4B, and 24.6 and 23.9 m3/s for Site OK4B, respectively. The average predicted and measured means of inside temperature were 20.3 and 19.7°C for IA4B, 23.3 and 22.9°C for NC4B, and 20.8 and 20.9°C for OK4B, respectively, based on their top performing submodel combinations. The overall optimal combination of four different submodels was determined to be the CIGR-T2 submodel, which consisted of the CIGR International Commission of Agricultural and Biosystems Engineering heat production rate equations for sows and a second order polynomial regression of inside versus outside temperatures in the temperature control region between the minimum and maximum temperature setpoints. The CIGR-T2 submodel simultaneously predicted the daily mean ventilation rate and daily mean inside temperature with good performance. The average RMSDs of the three sites for ventilation rate and inside temperature were 7.05 m3/s and 2.78°C, respectively. Sensitivity tests simulated based on the optimal BEPM (CIGR-T2) showed that annual total energy costs including electricity for powering fans and supplemental heat were influenced significantly by the minimum inside temperature setpoint, the thickness of ceiling insulation, and the minimum ventilation rate. This BEPM can be used for energy usage predictions, cooling and heating systems analysis and design, and as an important module of process-based gas emission models. It can be expanded to other livestock species (swine farrowing and finishing, egg laying operations, freestall dairy barns, etc.) by changing the heat production rate prediction submodel.