عنوان

Improvement and Application of a New Ammonia Emission Inventory for Poultry and Swine Production in NC

Number

TL52338

.Language of Text, Soundtrack etc

انگلیسی

Title Proper

Improvement and Application of a New Ammonia Emission Inventory for Poultry and Swine Production in NC

General Material Designation

[Thesis]

First Statement of Responsibility

Dietrich, Yijia Zhao

Subsequent Statement of Responsibility

Reich, Brian

Name of Publisher, Distributor, etc.

North Carolina State University

Date of Publication, Distribution, etc.

2019

Text of Note

172 p.

Dissertation or thesis details and type of degree

Ph.D.

Body granting the degree

North Carolina State University

Text preceding or following the note

2019

Text of Note

As one of the criteria pollutants regulated under the National Ambient Air Quality Standards (NAAQS), fine particulate matter (i.e., PM2.5) has significantly negative impacts on human health and environmental quality. In Southeast United States (U.S.), secondary inorganic PM2.5 (iPM2.5) has been found to be the major component of the total PM2.5 mass. Ammonia (NH3) is a primary precursor gas for the formation of secondary iPM2.5 that is largely emitted from animal agriculture and fertilizer application. In particular, livestock and poultry production has been identified to be the largest source of NH3 emission in North Carolina (NC). Studying NH3 emissions from animal agriculture is critical for improving knowledge and understanding of potential impacts of those NH3 emissions on the secondary iPM2.5 formation. While air emissions from animal feeding operation (AFO) have been investigated for decades, research remains needs to improve NH3 emission inventory (EI) to systematically assess fate and transport of AFO NH3 emissions and associated impacts on a regional scale. Therefore, the research objectives of this study were 1. To establish a new AFO NH3 EI with improved spatial resolution in NC, 2. To predict ambient NH3 concentrations through an atmospheric chemical transport model, named CMAQ, coupled with the new EI 3. To identify the hotspots of ambient NH3 concentrations in response to AFO emissions and atmospheric conditions, 4. To demonstrate the spatial improvements in NH3 hotspot prediction with the use of the new EI as compared to the existing EPA's NH3 National Emission Inventory (NEI), and 5. To check the performance of CMAQ modeling with the new EI by comparing model results with measurements. For objective 1, the spatial distributions of swine and poultry production farms were identified with use of Google Earth and mapped by ArcMap. Farm-level NH3 emissions were determined through applying most up-to-date emission factors (EFs) and activity data (e.g., animal population) for animal production confinement, waste storage/treatment facilities (e.g., lagoon), and waste land application. Different from the existing 2014 EPA NEI, the EFs applied in this study allowed the knowledge reflection on local management practices and environment. It was discovered that there were totally 6,769 poultry and swine farms located in NC, in which 2,378 broiler farms, 871 layer farms, 1,228 turkey farms, and 2,292 swine farms. The NH3 emissions from poultry farms were found to have a seasonal pattern, in which summer had the highest emission. Among confinement, storage, and land application, animal confinement was found to the biggest contributor to total on-farm NH3 emission. For swine, seasonal contrast of NH3 emission from confinement was found to be significant with high winter emission due to high EFs applied associated with high anima weight applied in winter. Summer NH3 emission from swine lagoon was significantly greater than winter. Manure storage and land application contributed significant amounts to swine farm total emission. The new NH3 EI compared to the EPA NEI inventory showed a much improved spatial resolution of estimated NH3 emission in NC. The new EI had higher emissions for broiler and layer confinements than the EPA NEI. However, the total annual county level emissions from all source sectors in the new EI and EPA NEI were found to be similar. For objectives 2-5, the new NH3 EI was used as the key input to an atmospheric transport model to link emissions and ambient concentrations of NH3. The Community Multiscale Air Quality Modeling System (CMAQ) was applied to simulate the chemical fate and transport conditions in the Southeast region of the U.S. The new NH3 EI was processed using the Sparse Matrix Operator Kernel Emissions (SMOKE) modeling system as point source format for NC poultry and swine farms. The CMAQ simulations were done for 3 days in January, April, October, and 4 days in July with intention to capture the seasonal differences. The outputting NH3 ambient concentrations were compared to those from using 2014 EPA NEI database, field measurements at Clinton Site located in the center of intensive animal farming in southeast NC, and field data from two sites of National Atmospheric Deposition Program (NADP) Ammonia Monitoring Network (AMoN). The spatial distribution of predicted NH3 concentration was analyzed using GIS techniques. The modeling results show that NH3 concentrations were over predicted by CMAQ with both the new EI and EPA NEI as compared to the monitoring data at Clinton site. Spatial variability of predicted ambient NH3 concentrations and their hotspots were much enhanced by profiling local NH3 emissions as point sources in the new EI of this study. Due to the fact that NH3 emission from livestock and poultry waste were only applied in NC in the entire U.S. domain during CMAQ input preparation of this study, model predictions at locations nearby the NC geographical boundary regions show bias as results of artificial NH3 sink and the inaccurately simulated chemical fate and transport, so called "boundary issues". The model predictions were observed to have improved performance at the two sites of AMoN located in the central regions of NC and were considered not affected by the "boundary issue". In addition to the boundary issues, it was also observed that simulation time could be another factor affecting model performance. In general, the longer simulation (i.e., 4 days) in July showed improvement in the agreement of model prediction and measurements. Results also indicate that CMAQ model prediction performance varied by locations of sites and time periods. It is recommended to evaluate the impacts of location and meteorology parameters on the model prediction.

Subject Term

Agricultural engineering

Dietrich, Yijia Zhao

Reich, Brian

North Carolina State University

Electronic name

p

[Thesis]

276903

a

Y