Simulation of air quality and operational cost to ventilate swine farrowing facilities in Midwest U.S. during winter.

We have developed a time-dependent simulation model to estimate in-room concentrations of multiple contaminants [ammonia (NH3), carbon dioxide (CO2), carbon monoxide (CO) and dust] as a function of increased ventilation with filtered recirculation for swine farrowing facilities. Energy and mass balance equations were used to simulate the indoor air quality (IAQ) and operational cost for a variety of ventilation conditions over a 3-month winter period for a facility located in the Midwest U.S., using simplified and real-time production parameters, comparing results to field data. A revised model was improved by minimizing the sum of squared errors (SSE) between modeled and measured NH3 and CO2. After optimizing NH3 and CO2, other IAQ results from the simulation were compared to field measurements using linear regression. For NH3, the coefficient of determination (R2) for simulation results and field measurements improved from 0.02 with the original model to 0.37 with the new model. For CO2, the R2 for simulation results and field measurements was 0.49 with the new model. When the makeup air was matched to hallway air CO2 concentrations (1,500 ppm), simulation results showed the smallest SSE. With the new model, the R2 for other contaminants were 0.34 for inhalable dust, 0.36 for respirable dust, and 0.26 for CO. Operation of the air cleaner decreased inhalable dust by 35% and respirable dust concentrations by 33%, while having no effect on NH3, CO2, in agreement with field data, and increasing operational cost by $860 (58%) for the three-month period.