Storage Stability of Volatile Organic Compounds from Petrochemical Plant of China in Different Sample Bags

According to the emission characteristics of volatile organic compounds (VOCs) in the petrochemical plants of China, the storage stability of VOCs for two different bags, polyester aluminum (PEA) and polyvinyl fluoride (PVF), was investigated in this study by comparing the adsorption of gas samples. A series of experiments were carried out to study the impact of different factors of sampling in the petrochemical industry. The results showed that the C2∼C3 substances can be adsorbed by the Tedlar bag, and after being refilled with pure nitrogen, the VOCs adsorbed previously by the bag material can be released. The aromatic hydrocarbon VOCs with larger molecular weight had a relatively lower recovery rate than the smaller molecular weights. And the average recovery of PEA airbags was significantly better than that of PVF airbags for storing stationary VOCs in the refinery of China. More kinds of substances can be detected in the airbags that had been added with helium protective gas, and it had a higher recovery rate for both kinds of simple bags after 24 hours of storage time, which indicated that the airbags without protective gas had adsorbed these substances.


Introduction
Recently, with the development of the urbanization, the emission of the VOCs has increased greatly [1]. e components of the VOCs are complicated and have a variety of harmfulness to the environment and the health of the people. erefore, they had drawn public attention [2][3][4]. At present, researchers have focused on the study of VOCs from indoor gas, and there are few studies on VOCs emitted from refinery and petrochemistry [5,6]. e petrochemical industry in China is the main emission source of VOCs, so the control of the VOCs in the petrochemical plant has been an urgent problem [7]. e key of the management and control is how to gather and analyze the components and the concentration of the VOCs. And most of the current researchers paid more attention on the improvement of analysis and treatment methods [8][9][10][11]. However, the collection and storage of samples, as the premise of VOCs analysis, also play an indispensable role in the accuracy of the entire analysis. e importance of VOCs storage stability is also equivalent to the analysis method of VOCs [12,13].
Whole-air sampling and sorbent-based tubes sampling methods are the most widely used techniques for collecting gaseous VOCs under on-site or laboratory conditions [14,15]. Among them, whole-air sampling techniques include airbag sampling and tank sampling. Each of these three methods has advantages and disadvantages. Sorbent-based tubes sampling has the advantages of small size, light weight, being easy to carry, and having multiple choices of adsorbent, which can be equipped with different targets; however, the sampling process is susceptible to human contamination; it is easy to react with the target compound to cause acquisition loss and the analysis process is complicated. e container sampling method has the advantages of quick and easy, which can collect multiple mixture samples at the same time and repeat the analysis in one sampling. However, sampling is time-consuming and labor-intensive and has high requirements on cleaning and vacuuming. It causes adsorption on high-concentration samples and is not easy to clean. e airbag sampling method is convenient and cheap for sampling and analysis, and multiple samples can be withdrawn from the bag for repetitive analysis [16], even though the compounds in the sampling bag are prone to physical leakage and chemical effects that cause the concentration to decay. In view of the complex composition of stationary source volatile organic compounds (VOCs) and the high concentration of exhaust gas in each component of the refinery, the airbag method was used for sampling and analysis in this study [17,18].
According to the emission characteristics of VOCs in the petrochemical plants of China, the two materials airbags (PEA and PVF), most commonly used for airbag sampling, were selected for analysis. A series of research and analysis were carried out on the storage stability of VOCs, and the adsorption effect of airbags on gas samples during the experiment was studied. A series of research and analysis were carried out on the storage stability of VOCs, and the leakage loss and the adsorption of gas samples by the airbags during the experiment were studied. e purpose is to explore the differences between the two materials for the collection and storage of VOCs and to provide certain theoretical and technical support for the collection, accurate analysis, and governance research of stationary source volatile organic compounds in petrochemical enterprises.

Experimental Program.
As VOCs collected in sampling bags were typically analyzed within a short period of time (usually less than 24-48 hours after sampling) [5,15,19], there were also a few experiments for analysis within three days. e stability of VOCs from stationary sources in the refinery and petrochemistry was evaluated in the sampling bags for 6 days in this study. e concentration of VOCs was analyzed at different times after sampling, at 6 h, 12 h, 18 h, and 24 h on the first day, and on the other 2nd, 3rd, 4th, 5th, and 6th days, using a gas chromatograph (Agilent 7980). e storage stability in the two kinds of bags was represented by the ratio R, which is shown in the following: where R is the recovery rate, C t stands for the concentration of "t" days stored in the bag (ppb), and C i is the initial concentration of the VOCs (ppb). When the ratio of C i /C r is within the range of 1σ precision error, that is, between 0.9 and 1.1, which means the value of VOCs is close to the initial concentration [16], and the stability can be considered as steady. Due to the large amounts of samples, the average of two parallel experiments was conducted for analysis.

Sampling
Method. e sampling site selected the sampling hole in the prehydrogenation fractionation section of a refinery in Qingdao. And the sample airbags and sampling system are shown in Figure 1.
e sampling system was installed, and the air tightness was checked as shown in Figure 1. After the system was installed, the filter head was removed, the front end of the sampling tube was blocked, and the vacuum pressure gauge was installed on the front end of the regulating valve with a tee and connected to the Teflon tubing through the joint where the vacuum box was passed. Starting pumping gas, when the vacuum pressure gauge reached 13 kPa, the regulating valve was closed. If the vacuum pressure gauge decreased by 0.15 kPa within 1 minute, the system was supposed to have good air tightness and there was no air leakage. If the air leakage was found, a sectional inspection should be carried out to find the leakage points and causes and then repaired in time.
In order to investigate the stability of VOCs stored in the sampling bag for 6 days, sampling was performed at a fixed flow rate of 500 mL/min by using airbags 1∼4 at the sampling location (airbag 1 and airbag 2 were PEA materials and airbags 3 and 4 were PVF (Tedlar) materials). At the same time, the airbags 5∼8 (the materials of the airbag 5 and 6 were PEA, and the airbags 7 and 8 were PVF (Tedlar)) were filled with 100 mL of helium gas sampled at the same position to provide for subsequent tests. Each airbag was sent to the laboratory for analysis. Taking packaging and transportation into consideration, the volume of PEA airbag and PVF airbag was 2 L, and the volume of gas sampled was about 1.6 L per sample.

Instruments and Methods.
e analytical instrument in this study was Agilent 7980 A. And the needle injection was used for the sample injection method and injection volume was 1 mL. e chromatographic conditions were as follows: the split ratio was 100 : 1, the postinlet temperature was 250°C, the pressure was 13.6 psi, and septum purge flow rate was 5 mL/min. e chromatographic column was Agilent 19091S-433, and the column parameters was 50 m × 200 μm × 0.5 μm. e initial column temperature was 35°C and increased to 300°C with 0.3 ml/min flow rate. e carrier gas was high purity helium, and the flow rate was 25 mL/min. When examining the storage stability of VOCs for the two airbags, the entire operation cycle of the first analysis was settled as 90 minutes. After getting the material type and the substance with longest residence time, the operation cycle was changed as shown in Table 1.

e Stability of VOCs Stored in Two Types of Airbags
e Adaptability of the Two Type of Airbags. e chromatograms of the sample stored in the two airbags for 6 hours on the sampling day are shown in the Figure 2 and Table 2. ere were 16 substances that can be detected in the PEA airbags by chromatography. Meanwhile, there were 23 substances detected in the PVF (Tedlar) airbags. e above substances include alkanes, benzenes, halogenated hydrocarbons, and olefins.
It can be seen clearly that more species had been detected in the PVF (Tedlar) sampling bag than the PEA sampling bag. And C1∼C3 was not detected in the PVF (Tedlar) sampling bag. It was possible that the C1∼C3 substances were adsorbed by the material of PVF (Tedlar) bag or leaked from the bag. e C5∼C8 can be detected in both airbags. After the fifth day's detection, two sampling bags were refilled with pure nitrogen and placed for 24 hours. Methane and propane were not detected in the PVF (Tedlar) airbag on the sixth day, but isobutane was detected ( Figure 3). It can be inferred that methane and propane leaked through the penetration of PVF airbags, and isobutane was adsorbed on the PVF material. On the other hand, 2-methyl-2-butene, 3methylpentane, 2-methylhexane, 2,3-two methylpentane, toluene, and other substances were detected in the PEA airbag ( Figure 3). is indicated that the sample bag released volatile organic compounds adsorbed previously on the PEA bag material.

e Relative Recovery Rate of VOCs in PVF and PEA
Airbags. 14 substances that can be detected in both airbags had been chosen for studying the stability of the airbag. And the 14 substances are listed in Table 2. e storage stability of the VOCs was studied for 6 days; the histograms showed the ratio of the VOCs concentration (C t (ppb)) measured on the "t" day (ppb) to the VOCs concentration (C i (ppb)) on the first day of the initial measurement. e error bars on each histogram represent the precision error. e black dotted line in each diagram indicates the ratio of C t /C i (ppb/ppb) of 1 : 1.
It can be seen clearly in Figure 4 that halogen volatile organic compounds, trichloromethane, had more than 90%   recovery rate in the first two days, and it reduced by 12% in 24 h-48 h. It had a significant decay on the third day and sustained until the sixth day. On the sixth day, the recovery rate in PVF had decreased to 23% in PVF bags and 29% in PEA bags. e benzene was usually used as a standard comparison substance in the chromatographic detection, and it had a higher stability in the airbags. e recovery rate of benzene in both types of airbags in the first 3 days was greater than 0.9, as shown in Figure 5(a). ere was a certain degree of attenuation from the third day to the sixth day. e recovery rate in the PEA bags on the sixth day was 76% and 71% in the PVF bags. In Figure 5(b), the methylheptane was not detected in the PEA airbag. e R value in the first 2 days in the PVF airbag was greater than 0.9; it decreased to 0.86 on the third day and decreased to 0.66 on the sixth day. is result showed that the aromatic hydrocarbon VOCs with larger molecular weight had a relatively lower recovery rate than the smaller molecular.
For the olefins, as shown in Figure 6, the recovery rates of 2-methyl-2-butene and 1,3-cyclohexadiene in the first two days of the two airbags were both greater than 90%. And it began to decrease on the third day in varying degrees. e decay rate of 1,3-cyclohexene recovery was faster than 2-methyl-2-butene, and the recovery rates of both olefins in PEA bags were greater than that in PVF bags. On the sixth day, the recovery rate of 2-methyl-2-butene in PEA was 46% (PVF: 30%), and the recovery rate of 1,3-cyclohexadiene in PEA was 28.2% (PVF: 24.5%). e results in Figure 7 showed that, for the alkanes, the concentrations of isopentane and 2,2-dimethylpropane on the first day (t � 6 h) were higher than other alkanes in the same sampling bag. Between the first day (t � 6 h) and last day (t � 6 d), 9 alkanes which can be detected in both kinds of airbags as shown in Figure 7 were selected for the comparison. It can be seen clearly that PEA bags had a higher recovery rate than PVF bags for the alkanes. e recovery rate of 2,2-dimethylpropane in PEA gas bag was greater than 90% within 24 h, and in the PVF bags, it stabilized until 18 h and then began to decay. Because 2,2-dimethylpropane was a highly volatile compound, the decay increased as time went by in the airbags. On the 6th day, the recovery rates in the two airbags were 49.6% (PEA) and 28.1% (PVF), respectively. e recovery rate of cyclopentane was always the highest in 6 days in all the alkanes gases, followed by cyclohexane. e recovery rates of cyclopentane and cyclohexane in the first two days were higher than 90% in the two airbags, and there was a small attenuation on the third day, which decreased by 8.85% (PEA) and 7.3% (PVF) and 8.8% (PEA) and 8% (PVF), respectively. e recovery rate of cyclopentane and cyclohexane on the sixth day was still maintained at 76.95% (PEA) and 72.10% (PVF) and 72% (PEA) and 70% (PVF), respectively, which was much higher  Journal of Analytical Methods in Chemistry than the other alkanes. e recovery rates of the other alkanes had the same tendency as the cyclopentane and cyclohexane. ey have recovery rates higher than 90% in the first two days, and there was a decline in different degrees from the third day to the sixth day. e recovery rates on day 6 are listed in Table 3. And it also can be seen that the recovery rate decreased with the increase of the molecular weight with the exception for the two most stable alkanes (cyclopentane and cyclohexane) and the most unstable alkanes (2,2-dimethylpropane).
e recovery rates of all VOCs during 6 days in all periods were taken into account together. It can be seen that the recovery of organic matters in all the studies had a decreasing trend with the change of storage time. e reason may be speculated that some VOCs themselves are highly active and easily decomposed and changed (e.g., 2,2dimethylpropane). A large number of water molecules were in the gas sample, on one hand and VOCs can be indirectly adsorbed through chemical interactions (hydrogen bonding, etc.) or physical interactions (dissolution, etc.) with the water molecules. In addition, the sampling bag materials have active sites that can adsorb or catalyze VOCs.
From the overall trend, during the 6-day storage period, PEA airbags had a better storage stability for VOCs than PVF. And the average recovery rate of PEA bags (79.6 ± 4%) was better than PVF bags (72.9 ± 4%). In addition, if only the longest period of the experiment (t � 6 days) was compared, the average recovery rate of PEA (53.2 ± 4%) was 8.2% higher than PVF (44.6 ± 4%).
e R values of the target      VOCs in the PEA were higher than those of the PVF in all time periods analyzed in this experiment. e results showed that the storage performance of VOCs in PEA bags is better than that of PVF in the storage period of 6 days, which could promote the accuracy of VOCs analysis.

e Effect of Shielding Gas (Helium) on the Stability.
According to Xia et al.'s research on the air sampling canister, it was found that after the inner wall of the air sampling canister was inertized, the benzene, n-hexane, and dichloromethane in the tank could be stabilized in the canister for 84 days and the concentration was basically not attenuated [20]. In this study, 100 mL of inert gas (helium gas) was introduced into the airbag before sampling to simulate the inertization of the inner wall of the Summa tank. And the airbag was stored in a dark place for 24 hours and then vacuumed and resampled to study the change trend of VOCs recovery with time.
In this study, the material of the airbags 5 and 6 was PEA, airbags 7 and 8 was PVF air, airbags 9 and 10 was PEA filled with helium gas for 24 h, and airbags 11 and 12 was PVF airbag filled with helium gas, respectively. e sample volume is 1.6 L and was taken back to the laboratory for chromatographic analysis.
e analysis was carried out from the substance type, initial concentration, and stability. e results are shown in Figure 3.
It can be seen clearly in Figure 8 that the kinds of substances in the PEA airbag were increased by filling with 1.8 6h 12h 18h 24h 48h 72h 96h 120 h144 h   Journal of Analytical Methods in Chemistry Table 3: e VOCs recovery rates in 6 days.
VOCs  the protective gas, named n-pentane, toluene, 4-methylheptane, and an unrecognized substance with a retention time of 31.532 min, while the substances in the PVF airbag were increased by 10 kinds, with 1 isobutane, 1 C 7 H 14 isomer, 7 C 8 H 18 isomers, and 1 C 9 H 20 isomer. It is shown that the airbag filled with inert gas can detect more kinds of substances after 24 hours of storage time, which indicated that the airbag without the protective gas had adsorbed these substances, and the added inert gas can replace the macromolecular substances adsorbed on the airbag material.
e basic information and peak area of the substances detected in the two airbags are listed in Table 4.

Stability Analysis.
In this part, 10 kinds of VOCs that can be detected in both kinds of airbags were chosen for the stability analysis of the airbags. e stability of VOCs stored in airbags under two conditions was compared. e recovery rates of the two airbags filled with helium gas and the two unfilled airbags sampled after 24 hours are shown in  Figure 3, in which bag5/bag6 were PEA airbags that were stored without helium condition, bag7/bag8 were PEA airbags that had been stored with helium condition, bag9/ bag10 were PVF airbags that were stored without helium condition, and bag11/bag12 were PVF airbags with helium condition. It can be seen from Figure 3 that the decay pattern of the VOCs with the addition of helium gas was similar to the VOCs that was without helium gas. Except the benzene, the stability of other VOCs in airbags was improved after 24 hours with the helium gas addition. e average recovery rate of the PEA airbag without adding helium gas after 6-day storage time was 78.43%, relatively; with adding the helium gas, the average recovery rate can be 80.93%. On the 6th day, the PEA airbag with helium added was 57.04%, which was nearly 5 percentage points higher than the unfilled recovery rate of 52.55%. e same decay trends can be seen from the PVF gas bags. e average recovery rate for 6 days was 75.03% and 72.04% with and without helium gas added, respectively. And on the 6th day, the average recovery rate of the PVF airbag added with helium was 49.24%, which was higher than the unfilled airbags that was 44.58%. Hence, in order to ensure the varieties and concentration of the stored samples and make the storage time longer, the helium gas should be added for 24 hours before vacuuming for sampling.

Conclusion
(1) Compared with PEA sampling bags, more VOCs can be detected in PVF (Tedlar) sampling bags. However, no C1∼C3 organic matter was detected in the PVF (Tedlar) sampling bags. e reason for the analysis may be that C1∼C2 small molecular substances leak and lose through permeation, and C3 may be adsorbed by the material of the PVF (Tedlar) bags. (2) Except for the two most stable alkanes, cyclopentanes and cyclohexane, and the most unstable 2,2-dimethylpropane, the recovery rate of the remaining alkanes decreased with the increase of molecular weight. e average recovery rate of all VOCs in PEA bags was better than that in PVF bags. On the 6th day, the recovery rate for the PEA bags was 8.2% higher than the PVF bags. e PEA airbags can produce more reliable data in the sampling analysis of VOCs. (3) After adding the helium protective gas, more kinds of substances can be detected. And the average recovery rate for PEA and PVF gas bags after vacuuming for 24 hours was 80.93% and 75.03%, respectively. at was higher than recovery rate without the addition of helium 78.43% and 72.04%, respectively.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.