Sources of Particular Pollutants in Ambient Air at a Petrochemical Enterprise

Xue Jianliang; Zhao Dongfeng; Cheng Jianguang; Chen Lu; Liu Wei

(1. College of Chemical and Enνironmental Engineering, Shandong Uniνersity of Science and Technology, Qingdao, Shandong 266590; 2.College of Chemical Engineering, China Uniνersity of Petroleum(East China), Qingdao, Shandong 266580)

Sources of Particular Pollutants in Ambient Air at a Petrochemical Enterprise

Xue Jianliang1,2; Zhao Dongfeng2; Cheng Jianguang1; Chen Lu2; Liu Wei2

(1. College of Chemical and Enνironmental Engineering, Shandong Uniνersity of Science and Technology, Qingdao, Shandong 266590; 2.College of Chemical Engineering, China Uniνersity of Petroleum(East China), Qingdao, Shandong 266580)

The study on source apportionment of particular pollutants in ambient air at a petrochemical enterprise is the basis of the control over air pollution. Through analyzing particular pollutants in the samples collected from one petrochemical enterprise in northwestern China, the sources of particular pollutants were discussed. The test results showed that concentrations of particular pollutants in different sites were remarkably different. Results showed that the sampling sites with higher concentrations of particular pollutants, including toluene, xylenes, NH3and H2S, were located at the boundary of the petrochemical enterprise. Instead, the concentrations of NMHC in the ambient air sampling sites were higher than those at the boundary of the petrochemical enterprise. The sampling sites with higher concentrations of particular pollutants were located in the area that was close to the petrochemical enterprise. The results obtained from the Pearson correlation coefficients analyses, the factor analyses, andχ2-tests of the particular pollutants had revealed that NH3, H2S, toluene and xylenes at all sampling sites came from the same source, while NMHC might come from some other sources besides the petrochemical enterprise.

particular pollutants; sources apportionment; Pearson correlation coefficient

1 Introduction

Following the accelerated process of industrialization in China, the air pollutants discharged from the industrial enterprises have greatly affected the people’s health[1-2]. Among these air pollutants, the benzene homologues (viz.: benzene, toluene, ethylbenzene and xylenes), the nonmethane hydrocarbons (NMHC), ammonia, and hydrogen sulfide are particular environmental pollutants emanating from petroleum and petrochemical industries[3-6]. These pollutants have brought forth some negative impact, such as poor air visibility and other health problems[4,7]. For example, the benzene homologues are toxic compounds and some of them are known as carcinogens[8]. However, NMHC are important chemicals taking part in the photochemical processes in the troposphere[9]. Furthermore, during the photochemical process, these particular environmental pollutants can react with each other to form secondary pollutants[10].

Recently, many researches have been focusing on monitoring the concentration or characteristics of air pollutants in some regions. While a few related study worked on the contribution of some specific emissions of particular pollutants from petrochemical enterprises. In this paper, some correlation methods were carried out to analyze the correlation between samples collected from the petrochemical enterprise and those collected around a typical petrochemical enterprise. The results could provide some qualitative analyses about the contribution of petrochemical enterprise to the particular pollutants.

2 Experimental

2.1 The selected particular pollutants of petrochemical enterprise and their standardized analytical methods

The particular pollutants of the petrochemical enterprise (e.g. NMHC, benzene homologues, ammonia, and hydrogen sulfide) were analyzed. Concentrations of these pol-lutants were monitored in order to verify the contribution of the petrochemical enterprise to the pollution of ambient air. Each monitoring item and their standardized analytical methods are shown in Table 1.

Table 1 Items monitored and their standard analytical methods

2.2 The specified sampling sites

The target petrochemical enterprise is situated in northwestern China. According to the local meteorological information, this research chose 11 monitoring sites around the petrochemical enterprise. All the 11 monitoring sites are listed in Table 2, with each site showing its relative distance to the centre and its direction to the petrochemical enterprise. The petrochemical enterprise is located in a vast plain. And, there is no other industrial enterprises except the said petrochemical enterprise in the surrounding.

This research used the samples collected at the border of the petrochemical enterprise to study the concentrations of pollutant sources. According to the pertaining rule, this research had chosen four sites as the pollutant sources of the petrochemical enterprise.

2.3 Quality control and quality assurance

The concentrations of pollutants at the eleven monitoring sites could represent the pollutant concentrations of the area. In the process of monitoring, the sampling was strictly performed in accordance with the relevant standards, and the pollutant quantification was performed using the five-point calibration approach. The blank test was also adopted. And the analytical methods did not find detectable amounts of the target analytes. The measured concentrations of the pollutants stipulated in the National Institute of Standards and Technology (NIST) 1941 were between 90% to 116% of the specified values. Therefore, the average recoveries varied from 85% to 104%.

Table 2 Location of sites for monitoring atmospheric particulate matter and their relative direction to the petrochemical enterprise

3 Results and Discussion

3.1 Distribution of the particular pollutants

All the samples of particular pollutants at 11 sampling sites around the petrochemical enterprise were collected. The mass concentrations of pollutants are shown in Table 3.

Table 3 Mass concentrations of specific pollutants collected at sampling sites of petrochemical enterprisemg/m3

Table 3 shows that the concentrations of particular pollutants at different sampling sites were significantly different. Results showed that the sampling sites with higher concentrations of the particular pollutants, such as toluene, xylenes, NH3and H2S, were located at the boundary of the petrochemical enterprise. For example, the highest concentration of toluene, xylenes, NH3and H2S was 4.89, 4.21, 0.13, and 0.013 mg/m3, respectively. Particularly, the concentration of toluene, xylenes and H2S at the boundary of petrochemical enterprise could be 10 times more than their respective concentration at the ambient air sampling sites. Instead, the concentration of NMHC at ambient air sampling sites was higher than that at the boundary of petrochemical enterprise.

To study the spatial distribution of these specific pollutants at the ambient air sampling sites, Table 3 also indicates that the sampling sites with higher concentration of these specific pollutants were located in areas that were near the petrochemical enterprise. Taking the sampling sites #1 and #3 as the examples, it could be seen that the mass concentration of NMHC, toluene, xylenes, NH3and H2S at the sampling site #1 was 3.86, 0.162, 0.17, 0.13, and 0.008 mg/m3, respectively. Whereas the concentration of these related pollutants at the sampling site #3 was only 3.72, 0.188, 0.179, 0.12, and 0.005 mg/m3, respectively.

3.2 Further analyses on sources of particular pollutants

The application of a statistical study would be helpful to identification of the sources of particular pollutants in order to shed light on the correlation between the concentrations of the pollutants. This paper applied the Principal Component Analysis to find out the most important components inside, and then rotated the factors with the Varimax rotation method in an attempt to find a factor solution that was equal to that obtained in the initial extraction and would have simpler and more explicit interpretation. Then, these rotated factors could reasonably interpret the possible sources of the ambient air pollution. Finally, theχ2-test was used to verify the previous results.

(1) The Pearson Correlation Coefficients and Factor Analysis of specific pollutants

To further investigate the sources of the particular pollutants, the Pearson Correlation Coefficient was adopted to carry out the correlation analyses[10]. In statistics, the Pearson Correlation Coefficient can reflect the relationship between 2 variables, e.g.: 0＜|rp|≤0.2 indicates a weak correlation or no correlation; 0.2＜|rp|≤0.4 indicates a low correlation; 0.4＜|rp|≤0.6 indicates a medium correlation; 0.6＜|rp|≤0.8 indicates a significant correlation; and 0.8＜|rp|＜1 indicates a high correlation. The formula of Pearson Correlation Coefficient is as follows:

in whichXandYare two different variables, andkis the number of samples.

The Pearson Correlation Coefficients of particular pollutant concentrations are presented in Table 4, and the factor loading graph of particular pollutant is depicted in Table 5.

Table 4 Pearson Correlation Coefficients of specific pollutants collected at ambient sampling sites and at boundary of petrochemi

It can be seen from Table 4 that the Pearson concentration correlation of 4 pollutants was 0.901, which meant a significant correlation. The result showed that NH3, H2S, toluene and xylenes in all sampling sites came from the same source. Upon considering the industrial enterprises besides the target petrochemical enterprise, the petrochemical enterprise was considered as the only source for emission of NH3, H2S, toluene and xylenes.

However, the Pearson Correlation Coefficient fell down to 0.593 when NMHC was included. Thus it could be inferred that there were many other NMHC sources besides the petrochemical enterprise. As for the concentration correlation of the particular pollutants at the ambient sampling sites and at the boundary of petrochemical enterprise, the rest four pollutants except NMHC as a subset could be better inferred as the source.

It can be seen from Table 5 that the loading values on each factor have been polarized and became more responsive. The loading of Component 1 on NH3, and H2S were 0.833 and 0.870, respectively, which were much higherthan that on NMHC. Since previous results showed that NH3, and H2S had a high correlation, thus these 2 points could work together to con firm that they might come from the same source (the target petrochemical enterprise). According to the manufacturing process of the petrochemical enterprise, and by analyzing the emission sources of the petrochemical enterprise, it could be inferred that they were probably originated from the wastewater treatment plant in the petrochemical enterprise.

Table 5 The factor loading graph of specific pollutan

Similarly, the component 2 had high loading on benzene and methylbenzene, which were 0.714 and 0.790, respectively. Since previous results also showed that there was a high correlation between toluene and xylenes, it could also prove that they might come from the petrochemical enterprise. And through further analyses on the manufacturing process of petrochemical enterprise, it was verified that benzene and methylbenzene might come from the catalytic reforming unit or the gasoline pyrolysis equipment.

Component 3 had high loading on NMHC, which was equal to 0.677. Based on the previous analyses, it could be inferred that the petrochemical enterprise was not its sole source and there might be some other sources that had contributed to the air pollution.

In summary, NH3, H2S, toluene and xylenes were probably emanated from petrochemical enterprise. NH3and H2S mainly came from wastewater treatment plant, while toluene and xylenes could be originated from the catalytic reforming unit, the naphtha pyrolysis unit, and the fractionation unit along with the fugitive emission sources at the petrochemical enterprise.

(2)χ2-test

The values of χ2-test is shown in Table 6, wherepis the difference correlation between two specific pollutants. The valuep≥0.05 indicates that the difference between two specific pollutants is not significant. Instead, the valuep＜0.05 indicates that the difference between two specific pollutants is significant.

Table 6 Values ofχ2-test among 5 specific pollutants at the ambient air sampling sites

It has been shown in Table 6 that the values ofχ2-test between NH3, H2S, toluene and xylenes were all ≥0.05, that is to say, the sources of NH3, H2S, toluene and xylenes are the same. The values were significantly different between NMHC and other components, thus NMHC could come from other sources. This is the reason why the concentration of NMHC at ambient air sampling sites was higher than that at the boundary of petrochemical enterprise.

4 Conclusions

By analyzing the concentration characteristics of particular pollutants and their correlation, as well as by applying the Principal Component Analysis and the Factor Analysis to the study on the source of particular air pollutants, the following conclusions could be drawn up:

(1) Concentrations of particular pollutants in different sites were significantly different. Results showed that the sampling sites with higher concentrations of particular pollutants, such as toluene, xylenes, NH3and H2S, were located at the boundary of the petrochemical enterprise. Instead, the concentrations of NMHC at ambient air sampling sites were higher than those at the boundary of petrochemical enterprise.

(2) The sampling sites with higher concentrations of particular pollutants were located in the area near the petrochemical enterprise.

(3) Application of the statistical study would be helpful to the analysis on the source of the specific pollutants to shed light on the correlation of the concentrations of the pollutants. By means of the Pearson Correlation Coefficients, the Factor Analysis, and theχ2-test of specific pollutants, the results showed that NH3, H2S, toluene and xylenes, which were collected at all sampling sites, camefrom the same sources and NMHC could come from other sources besides the petrochemical enterprise.

Acknowledgments:This study was financially supported by the Fundamental Research Funds for the Central Universities (No. 13CX06055A), and Key Technology Development Projects of Qingdao Economic and Technological Development Zone (No. 2013-1-58).

[1] Chameides W L, Fehsenfeld F, Rodgers M O, et al. Ozone precursor relationships in the ambient air[J]. Journal of Geophysical Research: Atmospheres, 1992, 97(D5): 6037-6055

[2] Kniemeyer O, Fischer T, Wilkes H, et al. Anaerobic degradation of ethylbenzene by a new type of marine sulfate-reducing bacterium[J]. Appl Environ Microbiol, 2003, 69(2): 760-768

[3] Li He, Liu Yu H, Luo Na, et al. Biodegradation of benzene and its derivatives by a psychrotolerant and moderately haloalkaliphilicPlanococcussp. strain ZD22[J]. Research in Microbiology, 2006, 157(7): 629-636

[4] Prenafeta-Boldu F X, Vervoort J, Grotenhuis J T C, et al. Substrate interactions during the biodegradation of benzene, toluene, ethylbenzene, and xylene (BTEX) hydrocarbons by the fungusCladophialophorasp. strain T1[J]. Appl Environ Microbiol, 2002, 68(6): 2660-2665

[5] M?llmann-Coers M, Klemp D, Mannschreck K, et al. Statistical study of the diurnal variation of modeled and measured NMHC contributions[J]. Atmospheric Environment, 2002, 36(S1): 109-122

[6] Zervas E, Montagne X, Lahaye J. The influence of gasoline formulation on specific pollutant emissions[J]. Journal of the Air & Waste Management Association, 2011, 49(11): 1304-1314

[7] Abdulkareem A S. Urban air pollution evaluation by computer simulation: A case study of petroleum refining company in Nigeria[J]. Leonardo Journal of Sciences, 2006, 4(6): 17-28

[8] Rooney-Varga J N, Anderson R T, Fraga J L, et al. Microbial communities associated with anaerobic benzene degradation in a petroleum-contaminated aquifer[J]. Appl Environ Microbiol, 1999, 65(7): 3056-3063

[9] Rappenglü Ck B, Fabian P, Kalabokas P, et al. Quasi-continuous measurements of non-methane hydrocarbons (NMHC) in the Greater Athens area during medcaphot-trace[J]. Atmospheric Environment, 1998, 32(12): 2103-2121

[10] Miyuki Noguchi, Mun-joo Bae, Kyoko Yamashita, et al. Correlation of the ratio of non-methane hydrocarbon and nitrogen oxide concentrations in the atmosphere and outdoor environment[J]. Building and Environment, 2009, 44(7): 1489-1492

Recieved date: 2013-08-25; Accepted date: 2013-11-03.

Xue Jianliang, E-mail: ll-1382@163.com.