3.1. Volatile Organic Compounds
Hydrocarbons, both aliphatic and aromatic, are the main VOCs (Volatile Organic Compounds) that are generated from the petroleum industry [
7]. Regulated emissions and leaks both contribute to the VOCs concentration [
12]. Aromatic VOCs are mainly comprised of Benzene, Toluene, Ethylbenzene, and Xylene (BTEX) [
13], and benzene is of particular concern since it is carcinogenic. VOCs are associated with a number of potential health effects. Short term exposure to VOCs may cause dizziness, fatigue, nausea and depression. Exposure to certain VOCs may even result in mutations and cancers [
14], and others may cause damage to the central nervous system, kidneys and liver [
15].
The detection of VOCs in air typically follows one of two strategies, active and passive sampling. In active sampling, a pump draws air from the sample location through an adsorbent substrate at a specific flow rate, or into a collection container, or to an in-situ continuous sensor [
16], depending on the type of VOC and concentration to be measured. The benefits of using active samplers are the collection of greater sample volume, and independence from environmental factors, unlike passive sampling. On the other hand, passive sampling does not call for the usage of pumps to draw the air sample, but uses the principle of diffusion through a static air layer into a collection medium. For example, a badge containing a sorbent is contacted with air, for a specific amount of time. Since passive sampling does not involve the usage of pumps, it is less bulky and easier to use. It is cheaper than active sampling but dependent on uncontrolled environmental factors.
It is important to know the atmospheric lifetime of VOCs to gain an understanding of the distances they might travel in air. A higher atmospheric lifetime indicates that the VOC can travel a greater distance in the atmosphere, possibly leading to impacts much farther away from the emission source.
Table 3 tabulates the atmospheric lifetime of several common VOCs [
17].
From
Table 3 it is possible to infer that methane has the longest atmospheric lifetime among the listed VOCs and hence can travel a long distance from the source of release, when compared to styrene, which has an atmospheric lifetime of just 4.9 h. Among the common aromatics (BTEX), benzene has the maximum lifetime of 9.4 days, while the lifetime of m-xylene is only 11.8 h.
In urban areas close to refineries and petrochemical industries, the air quality is lowered by the combined effect of industrial emissions and traffic emissions. This was observed in Lithuania, which has a very large crude oil refinery. Benzene concentrations in air in the urban region exceeded the rural region by 0.63 ppb [
13]. Air quality at rural regions near these industries was influenced only by the industrial emissions and traffic emissions played a minor role.
VOCs can also be aliphatic in nature. In a study performed in three sampling sites in Turkey, around large petroleum refining and petrochemical industries, it was observed that the aliphatic VOCs and their derivatives were predominant. Ethylene dichloride was the main pollutant, followed by ethyl alcohol and acetone [
12]. It was stated that there is a high likelihood of finding ethylene dichloride near these petroleum refining and petrochemical industries as a result of its usage in these industries. In this case, ethylene dichloride is used as a lead scavenger for preventing deposition of lead oxide, in leaded gasoline blends. It is also formed as an intermediate compound in the petrochemical industries during production of vinyl chloride.
In a study of hydrocarbon source signatures at La Porte in Houston, Texas it was observed that in addition to BTEX, other major VOCs were ethene, propene, 1-butene, C2-C4 alkanes, hexane, cyclohexane, methylcyclohexane, isopropylbenzene, and styrene [
18]. Some sites at which concentration measurements were taken were far away from major roads and highways such that traffic did not significantly affect the measurements. This is very important and we can infer that a non-urban region is the best indicator of the pollutant concentration from the industry, especially for VOCs. From a similar study around a major petroleum refinery in Greece, it was concluded that though aromatic hydrocarbons are a major source of VOCs, saturated hydrocarbons are also prevalent [
7], which originated from the evaporation of oil products and organic solvents. In Tarragona region, Spain, which has many major petrochemical and chemical industries, ambient air concentrations of a group of 65 VOCs were measured [
14]. It was assessed that average contributions to total VOCs made by BTEX, aromatic compounds, halogenated compounds, alkanes and alkenes ranged between 24% and 41%, 8% and 19%, 11% and 19%, 12% and 36%, and 15% and 33%, respectively.
Wind speed and direction are meteorological parameters that play a major role in the distribution of emissions and may vary from day to day. VOCs concentration decreases when wind speed is high, due to greater dilution. When the predominant wind direction is such that wind carries contaminants from the source to the sampling site, the concentration observed will be higher in this case, when compared to other wind directions.
Seasonal variations of VOCs have also been observed in various cases. In a study at Turkey, it was found that the VOC concentration was maximum in summer due to evaporation from fugitive sources, as a result of higher temperatures. The concentration was comparatively lower in the other seasons [
12]. Diurnal variations in VOCs can also occur, and concentrations can be high at night. This can be attributed to the relatively calmer atmosphere at night than during the day, and the lack of photochemical activity, which acts to degrade VOCs in the atmosphere. During the day, the number of highly reactive species is lower compared to the low reactive VOC species, due to photochemical reactions. On the other hand, the low reactive VOC species are more abundant due to accumulation [
19].
At a petrochemical industrial area in Yokohama Japan, 39 aliphatic and aromatic VOCs were measured in the ambient air, and aliphatic VOCs were the majority. n-Butane resulting from gasoline evaporative emissions and was found to be 3.5–6 ppb, while consumer products and petrochemical refining were the sources of isobutene in addition to gasoline evaporation. Propane was generated from the usage of LPG and natural gas processing [
19]. Aromatic VOCs, mainly BTEX was also present in large amounts. Among BTEX, toluene was found to be most abundant (20.1 µg/m
3) followed by ethyl benzene (12.6 µg/m
3), benzene (6.7 µg/m
3), and xylenes (m,p-Xylene: 9.1 µg/m
3, o-Xylene 2.9 µg/m
3). BTEX concentrations in some regions have been tabulated for comparison in
Table 4.
From the above Tables, it is possible to calculate the concentrations of BTEX in Asian and Non-Asian countries. For the Asian countries, the average concentrations are 2.14 ppbv Benzene, 6.11 ppbv Toluene, 1.33 ppbv Ethylbenzene, and 3.34 ppbv total Xylene. On the other hand, for non-Asian countries, the average concentrations are 1.82 ppbv Benzene, 3.01 ppbv Toluene, 0.57 ppbv Ethylbenzene, and 1.99 ppbv total Xylene. Based on this data it can be observed that the concentrations of all components of BTEX tend to be higher in Asia, when compared to other continents. This could be due to a variety of reasons including temperature differences. However, increasing energy demand and industrialization, and looser regulatory criteria may also be a contributing factor. When total VOCs are taken into consideration, the mean (arithmetic mean) of all VOCs for industrial areas is 100.19 ppbv. This is the highest among all regions. This is followed by urban areas, which have 24.33 ppbv of total VOCs. Total VOCs concentrations are least in rural areas, where the mean concentration of total VOCs is 18.61 ppbv. This could be attributed to the fact that industrial areas have the highest concentration of VOCs due to emission from industries, as well as traffic emissions. VOCs in urban areas mainly arise from vehicular emissions. On the contrary, rural areas are lower on emissions as they have very little traffic or industrial emissions.
Since a refinery site has a huge number of potential point sources of release, it is common to consider it as an area source. Once a pollutant concentration at a site is found, Inverse Dispersion Modeling (IDM) can calculate its emission rate. In a petroleum refinery in Northern China, this method has been used to find the emission rates [
26] using a three-step approach. First, VOC concentrations and meteorological conditions were measured. The next step involved the quantification of the relationship coefficient between the VOC emission rate of the source and the ambient VOC concentration, under the actual meteorological conditions measured. As a result, the monthly VOC emission rate was calculated [
26].
VOC emission is an increasing problem, with similar scenarios in many major regions of the world. In addition to the VOCs patterns in the regions discussed so far, the trend is similar in Beijing (China) where field VOCs measurements were made for a petroleum refinery. VOCs were monitored for a period spanning from June to September 2011. A 61 ppbv rise in the ambient VOCs was observed in this time span [
27].
3.2. Greenhouse Gases
The petroleum industry consumes a significant amount of fossil fuels for operational purposes, and this contributes to greenhouse gas emissions. The greenhouse gases associated with the petroleum refining and petrochemical industries are mainly methane, carbon dioxide and nitrous oxide [
28]. In 2006, it was noted that equipment leakage and venting of CH
4 and CO
2 from petroleum and natural gas industries contributed to 317 million metric tons of CO
2 equivalent (MMTCO
2e). Out of this, 261 MMTCO
2e of CH
4 and 28.50 MMTCO
2e of CO
2 were emitted from the natural gas industry, whereas 27.74 MMTCO
2e of CH
4 and 0.29 MMTCO
2e of CO
2 were emitted from the petroleum industry.
Table 5 provides a clear comparison to understand the level of greenhouse gases emissions from each stage of the oil and gas industry.
Many processes involved are very energy intensive, like alkylation, isomerization and so on. In recent years, increasing stringent fuel specifications pose a greater challenge for the oil and gas industry. To meet these requirements, the industry is using more energy in order to remove contaminants such as sulfur. Heat and power are persistent needs with the former being required at higher levels, often indirectly in the form of steam. Many refineries meet their heat and electricity needs by producing it themselves from gas, Fluid Catalytic Cracking coke and fuel oil. In this case, the fuel requirement, and hence the greenhouse gases emissions, are higher than they are for the refineries which procure these utilities from elsewhere.
Sometimes, though a few processes may be less energy intensive, the volume of input material processed might be high [
29]. The greenhouse gas emissions are also linked to the complexity of the refineries, where in general the more-complex refineries emit more such gases. Even during the production stage, greenhouse gas emissions are significant. In 2013, the carbon dioxide emissions during oil production and processing were estimated to be 130 kg CO
2 per Ton of Oil Equivalent (TOE) [
30,
31]. For each ton of hydrocarbon production, there would be release of 1 kg of methane, and 0.4 kg of nitrogen oxides [
30].
The count of wells and the production rate from each well have an important effect on the green house gases emissions. During the well completion phase it is crucial for the methane released to be flared or recovered. Intensity of emissions mainly depends on the ultimate recovery from the well [
32].
Table 6 gives a detailed picture of the type of greenhouse gases emitted from various sources in the processing stage, and the sources and type of greenhouse gases emitted [
33].
Throughout the operation of the petroleum industry, some gases are vented intermittently during equipment failures and maintenance activities [
34]. Hence, blowdown and purge can also result in emissions of methane, where blowdown refers to the venting of gases contained in an equipment, pressure vessel, or pipeline, and purge refers to the process of replacing air in the equipment with other gases. This is done after maintenance, to restore an oxygen-free environment after the maintenance activity is over, for safety purposes. When this happens, some purge gas may be released as the air is displaced from the equipment [
35]. The factors that affect the amount of methane released from maintenance and blowdown are the frequency of blowdown (times/year/equipment), volume of methane released per blowdown event (function of pressure, volume within the equipment and gas composition) and disposition of the blowdown gas.
There are some solutions for minimizing the energy consumption in processes and thereby the magnitude of greenhouse gases emitted. Two main approaches are heat integration (heat management) and waste heat recovery, where the existence of large temperature differences between different streams provides an opportunity to recover heat and reduce overall energy use [
29]. Pinch analysis is a very effective technique to achieve a high level of energy efficiency [
36]. It is a rigorous approach that can lead to significant improvements in process and utilities, which in turn has many benefits like reducing operating costs, debottlenecking processes, and improving efficiency. It enables the identification of streams that have the maximum potential for economical heat recovery.
Pinch analysis has achieved energy savings of 10–25% (expressed as a percentage of total purchased fuel) in oil refining areas, and 15–25% in the petrochemical industry [
36]. It can also contribute to savings in water and hydrogen consumption. Another area of reduction in energy consumption is an increase in furnace efficiency, which refers to the ratio of the heat transferred to the material and heat supplied to the furnace. The heat losses in the furnace can include heat storage in the structure of the furnace, losses from the outer walls of the furnace, heat transported out through load conveyors, radiation losses, heat carried by the cold air infiltration into furnace, and heat carried by excess air in burners [
37]. To increase the efficiency of the furnace and minimize heat loss, one could reduce the temperature of flue gas, lower the volume of flue gas by minimizing excess air, optimize the fuel air-ratio, and use combustion air that contains a higher amount of oxygen. Waste heat recovery is another way to improve furnace efficiency, where we can utilize the heat energy in the exit gases for the process. Recuperators and regenerators are a common option to consider for waste heat recovery.
3.3. Particulate Matter
Particulate matter (PM) is of increasing concern in recent years as studies have shown links with a variety of health impacts such as respiratory and cardiovascular morbidity, and even lung cancer [
38]. According to the International Commission on Radiological Protection (ICRP) Publication 30, particulate matter that is larger in size (over 10 µm) is intercepted by the nasal, tracheal mucus and is concentrated mainly in the naso-pharyngeal region [
39,
40]. Particulate matter of very small diameter is considered to be the most problematic as it penetrates deep into the lungs, thereby depositing in the bronchioles and alveoli in the pulmonary parenchyma region. Therefore, in many regions the focus of measurement and regulation is on particulate matter with a size of less than 2.5 µm (PM2.5). According to World Health Organization (WHO) guidelines, the annual arithmetic mean concentration of PM2.5 should not exceed 10 µg/m
3, while the 24-h mean concentration of PM2.5 should be below 25 µg/m
3 [
41]. Particulate matter in air is sampled using integrated sampling systems, typically for a period of 24 h. Filter substrates in the system capture PM, from which it is possible to measure the mass concentration using gravimetry. Captured PM can also then be analyzed for its chemical components where desired [
16].
Based on the formation mechanism, particulate matter can be classified into two types, namely primary and secondary particulate matter. Primary particulate matter is directly emitted from the source to the atmosphere, and a major source from the petroleum refining industry is the Fluid Catalytic Cracking Unit (FCC) [
8]. Secondary particulate matter is mainly formed in the atmosphere from precursor gases like SO
2, NO
x and VOCs, from photochemical reactions or liquid phase reactions in fog droplets or clouds. SO
2 and NO
x are oxidized to form sulfuric acid and nitric acid respectively, and SO
2 oxidation can follow two routes. In the first one, SO
2 is oxidized to sulfate through photochemical reactions during the day, and the presence of oxides of nitrogen and hydrocarbons accelerates the process of photochemical oxidation. Alternatively, SO
2 can be oxidized to sulfate in the droplets in clouds. When molecules of air pass through the clouds, the soluble SO
2 gas gets trapped by the water molecules, where it is then oxidized to sulfate. Secondary sulfates and secondary nitrates scatter light well, resulting in hazy conditions and reduced visibility. In rural areas, the ratio of PM2.5 to PM10 is higher since secondary aerosol sources contribute considerably to PM2.5 formation [
42,
43]. Secondary aerosol includes particulates formed as secondary organic aerosols (SOA) as a result of the atmospheric and photochemical oxidation of VOCs.
Hydrocarbon emissions from the petroleum refining and petrochemical industries increase the rate of the first mechanism (i.e., formation of secondary sulfates from photochemical oxidation). In a study at Edmonton Canada, the PM2.5 level was measured over a six-year span [
44]. Secondary organic aerosols were the main contributor followed by secondary nitrate and secondary sulfate. Secondary nitrate constituted 16.7% of the mass concentration of measured PM2.5. Correlations of nitrate were observed with NO
2, CO, VOCs such as BTEX and other aromatic hydrocarbons, and alkanes. Further evaluation revealed that the dominant wind directions for the local sources were from the petroleum refineries and petrochemical plants on one side and the local roads (vehicular exhaust) on the other side. Secondary sulfate was found to be the third major component, contributing 15.4% to the mass concentration of measured PM2.5. It was determined that the petroleum refining and petrochemical industry was the main source for the presence of secondary sulfates at the Edmonton sampling location. At Rio de Janeiro, a study was carried out to evaluate emissions of aerosol particulates [
45]. It was observed that oil combustion and traffic were accountable for 52–75% of the fine aerosols emitted. Industrial emissions also increase the amount of sulfur present within PM2.5 emissions [
46].
There is also the possibility of the presence of metals in the particulate matter due to the emissions from this industry. A study on the effect of oil refinery and vehicular emissions on airborne particulate matter in Gela, Italy, revealed that there were many metals and metalloids in the airborne particulate matter [
47]. When assessed further, it was found that the presence of arsenic, molybdenum, nickel, sulfur, selenium, vanadium and zinc, was attributed to the petrochemical industry emissions. Vehicular emissions contributed to the presence of copper, lead, platinum, palladium, antimony and some portion of zinc. These metals are potentially toxic, and their association with the atmosphere aerosols can have adverse health impacts [
47,
48]. A similar experiment conducted at Houston, Texas showed the presence of metals in airborne particulate matter, resulting from emissions during non-routine operations like startup, shutdown, maintenance activities and equipment failures [
8]. Scandium, vanadium, molybdenum and nickel are common due to combustion of crude oil and its fractions. Lighter lanthanoids are enriched in FCC catalysts and hence refining operations can also contribute to the presence of metals in PM10.
Apart from these pollutants, petrochemical industries also produce Persistent Organic Pollutants (POPs), which resist environmental degradation by many processes. Though the Stockholm Convention banned the usage of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs), and polychlorinated biphenyls (PCBs) in 2001, a study was performed at Tarragona County in Catalonia (Spain) in the year 2002 to determine the concentrations of POPs in the soil, from chemical and petrochemical industries [
49]. This was used to assess the environmental exposure to PCDD/Fs, PCBs, polychlorinated naphthalenes (PCNs) and polycyclic aromatic hydrocarbons (PAHs). The results obtained for the exposure to pollutants in the case of petrochemical industries, has been tabulated in
Table 7.
The measured levels were on the higher side when compared to other chemical industries and unpolluted sites. The soil samples in the region also contained heavy metals like manganese, lead, chromium.
3.4. Ozone
Ozone is a secondary pollutant that is formed from the reaction between VOCs, oxides of nitrogen, and sunlight. Hence, ozone production depends on the amount of VOCs, NO
x, and also the ratio of VOC/NO
x. The petrochemical industry can be a significant source of hydrocarbons (VOCs) and NO
x [
10,
18], and hence it contributes to a favorable atmosphere for the formation of ozone. In the case of ozone resulting from coal-fired power plant emissions (which contain NO
x but very little VOCs), hydrocarbons have to be provided from the surrounding region, and the power plant location relative to sources of reactive hydrocarbons is important in determining the impact on ozone formation [
10]. However, the petrochemical industry can release both ozone precursors and therefore can have a more direct impact on regional ozone formation. The effects can be worse in summer when temperatures and ozone production rates are higher [
50]. Other factors such as vehicular emissions also increase ozone formation by contributing to high levels of ozone precursors. Hence the urban zones/ traffic prone regions near petrochemical sites are at high risk of ozone formation. The presence of ground level ozone and airborne particles is a characteristic of photochemical smog, which is characterized by haze and reduced visibility. Topography and meteorology such as inversions also influence smog retention by trapping the pollutants. Smog can cause respiratory problems, coughing, irritation, reduced lung function, asthma, lung inflammation and damage to the lung lining [
51].
Surface ozone is abundant in urban areas due to the large number of emission sources, but can be more abundant in rural areas when nitrogen oxides and VOCs are transported in the downwind direction. The ozone formation in such cases, is more sensitive to the presence of NO
x as VOCs are already present in large amounts due to the abundance of Biogenic Volatile Organic Compounds (BVOCs) emitted by vegetation, and this trend was noted in a study in Korea [
52]. High levels of ozone were observed in Chuncheon, in comparison to Seoul, which was located upwind, though the emissions were less in Chuncheon than Seoul. Ozone moving from Seoul to Chuncheon was more sensitive to NO
x. Chuncheon had abundant VOCs during day, hence NO
x emitted from Seoul played a major role in the ozone formation.
Other than the concentration level of pollutants (VOCs and NO
x), another important parameter to be considered is their photochemical reactivity, determined by the Relative Incremental Reactivity (RIR) analysis [
5]. In a study on concentration of VOCs and its impact on ozone concentration at Wuhan, China, it was observed that ethane, ethene and toluene were the most abundant VOC species, with concentration levels 5.2 ± 0.2 ppbv, 3.3 ± 0.2 ppbv, and 2.0 ± 0.1 ppbv respectively [
5]. However, from the RIR analysis it was found that ethene and toluene were the main species contributing to the formation of ozone although ethane was present at higher concentrations.
Hydrocarbon release events with a frequency of several times in a year—which occur intermittently and unexpectedly—can cause ozone peaks. In the Houston and Galveston area, which has several petrochemical complexes, irregular emissions generated plumes that had rapid ozone formation [
53]. In another event at Galveston Bay, 3000 kg of ethene was released over a time period of half an hour, and the ozone concentration measured within this event plume was 50 ppb higher than the concentration that was observed outside the plume, and 100 ppb higher than the regional background. In Houston, a discrete emission event releasing more than the allowable range of hydrocarbons occurs rarely at any single industrial facility, but if all industrial facilities are considered as a whole, then a large emission event releasing greater than 1000 kg of hydrocarbons occurred every week [
54]. By modeling hundreds of such release events in Houston, it was found that approximately 1.5% of all the emission events that have occurred in that area produced over 10 ppb of additional ozone, 0.5% of emission events produced over 70 ppb of additional ozone, compared with the results with no emission events [
54]. Episodic events can have significant regional effects as they can cause ozone peaks in downwind regions.