Physical, chemical and biological properties of airborne particles emitted from animal confinement buildings

摘要

Airborne particles emitted from animal confinement buildings have been an environmental concern for some time. Relevant regulations are under discussion. However, one of the major obstacles is a lack of comprehensive understanding of their physical, chemical and biological properties that may be closely related to environmental and health effects. In addition, the use of receptor modeling requires known chemical source profiles. To address the current research need, airborne particles collected from six types of animal confinement buildings were subject to characterization. First, TSP, PM10 and PM2.5 concentrations were determined gravimetrically and their variations were investigated. PM concentrations were significantly affected by animal building type, season (ambient temperature) and feeding systems; while, animal density had no significant effect. Specifically, higher PM concentrations occurred in poultry buildings than swine buildings; PM concentrations decreased with ambient temperature; and wet feeding systems were associated with lower TSP and PM10 concentrations than dry feeding systems. A generalized linear model was established for estimating PM10 concentrations in swine buildings with animal building type, daily average ambient temperature, specific fan area, animal density, and TSP concentrations as predictors. The coefficient of determination (R2) of the proposed model was 0.907.Second, some previous studies reported that Federal Reference and Equivalent Method (FRM/FEM) PM samplers oversample PM10 and PM2.5 from agricultural sources. They proposed an indirect method that calculates PM10 and PM2.5 concentrations from TSP concentrations and the particle size distribution (PSD) of TSP. The conclusion and the proposed calculation method were established based on several assumptions. The present study shows that when different assumptions are employed, different conclusions and calculation results could be obtained. The PM10 and PM2.5 concentrations derived from different particle size analyzers could be significantly different. Among the four analyzers under investigation, Aerosizer DSP produced the most comparable PM10 concentrations to the gravimetric method. Third, the chemical composition of the PM10 and PM2.5 samples was examined, including inorganic elements and soluble ions. The present study revealed that PM chemical composition varied significantly with animal building type. PM samples from certain different types of animal confinement buildings, e.g., manure-belt layer hen and tom turkey, had significantly different chemical compositions, indicating a possibility of applying receptor models to determining PM contributions by different animal building types. Seasons had no significant effect on PM10 and a significant but slight effect on PM2.5 chemical compositions- the absence of strong seasonal variations is a good news for PM receptor modeling. Future efforts should be made to apply receptor models to animal production related air quality problems, and to compare receptor models with dispersion models to assess their respective advantages and limitations. PM2.5 samples from different types of animal confinement buildings had more similar chemical compositions than PM10 samples. One of the limitations associated with the present study is that the total mass fraction of investigated chemical species was low, typically less than 16%. Future investigations should attempt to characterize the rest of PM mass.Fourth, a total of 57 odorants were identified and quantified in TSP, PM10 and feed samples. Acetic acid and ethanol were the most abundant odorants; while, phenylacetic acid and (E,E)-2,4-decadienal were the top two odor contributors, considering their low odor thresholds in air. The odorant composition of PM samples varied significantly with animal building type. Compared to the TSP samples, the PM10 samples from different animal buildings were more similar in odorant composition and contributed on average 50% of the odor strength of TSP samples. The effect of seasons was also significant, but less substantial than that of animal building types. A gradual change in odorant composition from hot to mild to cold seasons was observed. TSP and PM10 samples were found to have significantly different odorant compositions and significantly higher odorant contents than feed samples, suggesting that the majority of particle-borne odorants may originate from sources other than feed. Different than the conclusions from some previous studies, the present study suggests that the majority of odorants exist in the gas phase rather than on particles. Fifth, airborne endotoxins and (1???3)-??-D-glucans in TSP samples were analyzed. Most measured airborne endotoxin concentrations exceeded the exposure limit proposed by Donham el al. (1995; 2000) and the threshold issued by the Dutch Health Council in the Netherlands, which may raise a health concern for farm workers as well as animals grown in confined environments. The present study revealed that animal building types had a significant effect on airborne endotoxins and (1???3)-??-D-glucan concentrations, but no significant effect on the contents of endotoxin and (1???3)-??-D-glucan in particles. By contrast, seasons had no significant effect on airborne endotoxin and (1???3)-??-D-glucan concentrations but a significant effect on their contents in particles, which increased with the daily average ambient temperature. Elevated indoor temperatures during the summer were considered to facilitate the growth and propagation of bacteria and fungi, thus leading to higher microbial contents in particles. A significant and positive correlation was identified between TSP and airborne endotoxin/ (1???3)-??-D-glucan concentrations, implying a possibility of applying existing dust control techniques for the mitigation of these two bioactive agents.