Analysis of the Synergies of Air Pollutant and Greenhouse Gas Emission Reduction in Typical Chemical Enterprises

Abstract

In this study, we selected the production processes and main products of three typical chemical enterprises in Shanghai, namely SH Petrochemical (part of the oil-refining sector), SK Ethylene, and HS Chlor-Alkali, to quantitatively assess the synergistic effects across technology, policy, and emission mechanisms. The localized air pollutant levels and greenhouse gas emissions of the three enterprises were calculated. The synergistic effects between the end-of-pipe emission reductions for air pollutants and greenhouse gas emissions were analyzed using the pollutant reduction synergistic and cross-elasticity coefficients, including technology comparisons (e.g., acrylonitrile gas incineration (AOGI) technology vs. traditional flare). Based on these data, we used the SimaPro software and the CML-IA model to conduct a life cycle environmental impact assessment regarding the production and upstream processes of their unit products. By combining the life cycle method and the scenario simulation method, we predicted the trends in the environmental impacts of the three chemical enterprises after the implementation of low-carbon development policies in the chemical industry in 2030. We also quantified the synergistic effects of localized air pollutant and greenhouse gas (GHG) emission reductions within the low-carbon development scenario by using cross-elasticity coefficients based on life cycle environmental impacts. The research results show that, for every ton of air pollutant reduced through end-of-pipe treatment measures, the HS Chlor-Alkali enterprise would increase its maximum CO₂ emissions, amounting to about 80 tons. For SK Ethylene, the synergistic coefficient for VOC reduction and CO₂ emissions when using AOGI thermal incineration technology is superior to that for traditional flare thermal incineration. The activities of the three enterprises had an impact on several environmental indicators, particularly the fossil fuel resource depletion potential, accounting for 69.48%, 53.94%, and 34.23% of their total environmental impact loads, respectively. The scenario simulations indicate that, in a low-carbon development scenario, the overall environmental impact loads of SH Petrochemical (refining sector), SK Ethylene, and HS Chlor-Alkali would decrease by 3~5%. This result suggests that optimizing the upstream power structure, using “green hydrogen” instead of “grey hydrogen” in hydrogenation units within refining enterprises, and reducing the consumption of electricity and steam in the production processes of ethylene and chlor-alkali are effective measures in reducing carbon emissions in the chemical industry. The quantification of the synergies based on life cycle environmental impacts revealed that there are relatively strong synergies for air pollutant and GHG emission reductions in the oil-refining industry, while the chlor-alkali industry has the weakest synergies.

1. Introduction

Owing to the potential consequences of global climate change, reducing greenhouse gas (GHG) emissions has become a common goal for all countries in the world [1,2]. The chemical industry’s activities entail high energy consumption and high emissions. In China, this industry has become one of the main sources of carbon dioxide emissions [3]. At the same time, production processes in the chemical industry give rise to a variety of air pollutant emissions, posing a threat to the atmosphere and human health [4]. Studies have shown that air pollutant and GHG emissions are homologous [5], and the realization of synergistic reductions in the emissions of air pollutants and GHGs is vital for the sustainable development of the chemical industry.

Scholars around the world have extensively studied the synergistic effects of reducing air pollutant and GHG emissions to explore effective strategies [6]. At the regional scale, Bollen et al. applied the MERGE model and confirmed that localized air pollutant and GHG emission reductions had significant synergistic benefits [7]. At the national scale, Brendemoen and Vennemo found that GHG emission reduction is important in improving the local ambient air quality [8]. Mardones et al. used a computable general equilibrium (CGE) model to study the impacts of introducing an environmental tax in Chile, and the results showed that a significant reduction in net carbon dioxide (CO₂) emissions provided co-benefits for local air pollutant reductions [9]. At the city scale, Mexico conducted a study on the co-benefits of air pollution control programs in urban agglomerations in 2009 with the support of the Integrated Environmental Strategies (IES) program organized by the U.S. Federal Environmental Protection Agency [10]. At the sectoral level, the Greenhouse Gas and Air pollution Interactions and Synergies (GAINS) model developed by the International Institute for Applied Systems Analysis (IIASA) was used to evaluate 3500 end-of-pipe technologies for air pollutants and 350 end-of-pipe technologies for structural CO₂ reduction [11]. Wagner et al. applied the GAINS model to derive the marginal abatement cost curves of GHGs and air pollutants for three scenarios, confirming that GHG abatement policies can simultaneously contribute to local air quality improvement [12]. Dandan Liu et al. proposed a method for the evaluation of the synergistic effects of technologies in pollution reduction and carbon mitigation in the petroleum refining industry. This model evaluates the synergistic control efficacy of technologies from multiple perspectives, including pollution–carbon synergy, the synergistic reduction of pollutant–carbon emissions, cost–benefit analysis, and environmental benefits. The developed methodology was validated through its application to typical petroleum refining technology. The results confirm the synergistic benefits of pollution and carbon reduction through five petroleum refining technologies [13].

Chinese scholars have focused on the synergistic effects of policy measures such as energy use reduction, energy efficiency improvement, and clean energy substitution on air pollutants and GHG emission reduction, but the scale of research on the synergistic effects of end-of-pipe measures is still insufficient [14,15]. Studies have shown that many localized air pollutant abatement measures, especially structural abatement, can synergistically reduce GHG emissions, but the high energy consumption of end-of-pipe treatment facilities may also lead to an increase in GHG emissions [16,17,18]. Researchers combining life cycle assessment (LCA) and scenario modeling have demonstrated that LCA not only enhances our understanding of environmental impacts but also enables a deeper analysis of the synergies between energy efficiency and emission reduction policies through scenario simulations [19].

In general, research in this field remains active, especially at the meso–micro level, where quantitative studies on synergistic emission reduction in specific industries or enterprises, as well as the development of synergistic control programs, are still scarce [20]. Additionally, data on the synergistic coefficients are scarce, making it difficult for scholars to accurately assess the synergistic emission reduction potential and actual effects of various emission reduction measures and technologies. (The synergistic coefficient is the amount of greenhouse gases that may be emitted or synergistically abated by abatement of local air pollutants per unit mass. If the synergistic coefficient is positive, it indicates that the abatement technology can synergistically abate greenhouse gases at the same time as the abatement of pollutants, with a positive synergistic effect; if the synergistic coefficient is negative, it indicates that the abatement technology will increase the emission of greenhouse gases alongside the abatement of pollutants, with a negative synergistic effect.) Currently, domestic and foreign studies on the synergistic effects of reducing air pollutant and GHG emissions mainly focus on the synergistic analysis of emissions. Critically, the synergistic effects are not only reflected in the levels of emissions but also in the extent of ecological and environmental impacts. Furthermore, quantitative analyses of the synergistic effects of reducing of air pollutant and GHG emissions in industries and enterprises are still relatively rare.

Taking the Chinese chemical industry as an example, we selected three typical enterprises, namely oil refining, ethylene, and chlor-alkali manufacturing, to quantitatively study the synergistic effects of reducing conventional air pollutant and GHG emissions. By carefully accounting for air pollutant and GHG emissions, we analyzed the synergistic effects of different air pollutant end-of-pipe measures on GHG emissions to provide specific data support and guidance for the industry and enterprises in optimizing emission reduction technologies. Moreover, we adopted the life cycle assessment methodology to carry out an environmental impact assessment and a synergistic effect analysis based on the LCA of the main products of the three typical enterprises (producing 1 ton of refined diesel fuel, 1 ton of ethylene, and 1 ton of chlor-alkali). We also expanded the synergistic effect analysis to the level of ecological and environmental impacts, providing scientific evidence and data support to aid in the development of targeted emission reduction measures and policies.

2. Research Subjects and Methods

2.1. Subjects of Study and Data Sources

Three typical enterprises were selected for this research: SH Petrochemical (in the oil-refining sector), SK Ethylene, and HS Chlor-Alkali. SH Petrochemical produces gasoline, diesel, and other products derived from crude oil as feedstock. SK Ethylene produces petrochemicals, such as ethylene, propylene, polyethylene, polypropylene, and styrene, derived from naphtha and propane as cracked feedstocks. HS Chlor-Alkali project produces chlorine and caustic soda, and it also produces dichloromethane from the byproducts of hydrochloric acid and ethylene supplied by SK as feedstocks to produce ethylene dichloride. The scope of this study regarding emissions accounting for the three enterprises is shown in Figure 1, Figure 2 and Figure 3.

The basic data used for emissions accounting regarding SH Petrochemical (part of the refining sector) are derived from the EIA report of this enterprise’s technological transformation project, with 2021 serving as the base year; the air pollutant and GHG emissions of SK Ethylene are derived from this company’s retrospective EIA report, with the base year of 2017; and the data used for accounting for HS Chlor-Alkali’s emissions are derived from its retrospective EIA report, with the base year of 2017.

2.2. Methodology

2.2.1. Methods Used to Account for Emissions of Air Pollutants and GHGs

The air pollutants released by SH Petrochemical and SK Ethylene primarily consist of conventional pollutants (NO_x, TSP, and SO₂) and volatile organic compounds (VOCs). HS Chlor-Alkali utilizes cleaner energy sources such as natural gas and hydrogen, with its emissions mainly comprising process exhausts that include specific pollutants (Cl₂ and HCl) and VOCs. The conventional air pollutants (NO_x, TSP, and SO₂) and specific pollutants (Cl₂ and HCl) emitted by these enterprises are categorized as organized emissions, whereas the VOCs are classified into organized and unorganized emissions. According to the 2017 air pollutant accounting guidelines released by the Ministry of Ecology and Environment, the methods for the calculation of enterprises’ actual emissions of major pollutants include the direct measurement method, the material balance method, and the production and emission coefficient method [21].

Table 1. Accounting methods for organized and unorganized emissions used by the chemical enterprises studied.

Enterprise	Category	Source of Pollution	Pollutant	Accounting Method
SH Petrochemical (refining segment)	Organized emissions	Process-heating furnaces	SO2	Material balance algorithm
			NOx, particulate matter, NMHC	Analogy
		Sulfur recovery unit off-gas	SO2	Field measurement
		Other organized emissions	SO2	Material balance algorithm
			NOx, particulate matter	Analogy
			VOCs	Field measurement
	Unorganized emissions	Equipment and piping dynamic and static sealing points	VOCs	Equation method
SK Ethylene	Organized emissions	Combustion flue gas emissions	VOCs	Field measurement
		Organized emissions from processes	VOCs	Field measurement
		Flare emissions	VOCs	Material balance algorithm
		Other organized emissions	SO2, NOx, particulate matter	Field measurement
	Unorganized emissions	Volatile losses in the storage and blending of organic liquids	VOCs	TANKS model
		Equipment static and dynamic sealing leakage	VOCs	Equation method + coefficient method
		Cooling tower, circulating water cooling system release	VOCs	Equation method
HS Chlor-Alkali	Organized emissions	Organized emissions from processes	Cl2, HCl, VOCs	Field measurement
	Unorganized emissions	Volatile losses in the storage and blending of organic liquids	VOCs	TANKS model
		Equipment static and dynamic sealing leakage	VOCs	Equation method

details the accounting methods used by SH Petrochemical (in the refining sector), SK Ethylene, and HS Chlor-Alkali, with specific accounting equations provided in Appendix A.

In accordance with the “Shanghai Municipal Chemical Industry Greenhouse Gas Emission Accounting and Reporting Methods” promulgated by the Shanghai Development and Reform Commission, petrochemical enterprises primarily account for CO₂ emissions, barring exceptional circumstances [22]. CO₂ emissions from net purchased electricity and heat are calculated using the coefficient method, while other emission sources are assessed via the carbon balance method, with detailed equations provided in Appendix A.

2.2.2. Synergy Analysis Methodology

(1)

Synergistic coefficients

The end-of-pipe treatment process for air pollutants has synergistic effects on GHG emissions. These effects can be categorized into direct effects and indirect effects [23]. The direct effect refers to GHG emissions generated by applying the end-of-pipe treatment process to air pollutants or via chemical reactions in this process within the geographical boundaries of an enterprise [24]. According to its chemical reaction equation, the limestone–gypsum method of desulfurization generates direct carbon dioxide emissions; in other words, there is a negative synergistic emission effect, and every 1 kg of SO₂ removed brings 0.68 kg of CO₂ emissions.

The indirect effects are mainly reflected in the GHG emissions indirectly generated by the end-of-pipe treatment facilities via power consumption. In both theoretical and on-site research methods, the values of electricity consumption per unit of desulfurization, denitrification, and dedusting are comprehensively used as the key indicators to measure the electricity consumption of chemical enterprises’ end-of-pipe treatment facilities [24,25]. Using the known motor power and the annual amount of desulfurization, denitrification, and dust removal equipment used, the electricity consumption coefficient can be calculated via the following equation:

$$\begin{array}{c}C{E}_{ec}={\displaystyle \frac{P\times t}{R_a}}\end{array}$$

(1)

where the subscript a denotes the type of air pollutant; $C{E}_{ec}$ is the coefficient of electricity consumption (kWh/t); $P$ is the average power of the end-of-pipe treatment equipment (kw); t is the annual operation time of the end-of-pipe treatment facility (h); and $R$ is the amount of air pollutant removed (t).

The coefficient of synergy is defined as the amount of additional CO₂ emitted per ton of air pollutant removed (t, both direct and indirect).

(2)

Cross-elasticity coefficients based on emissions

The pollutant reduction cross-elasticity coefficient can be used to quantitatively analyze the synergistic effect between end-of-pipe reductions in air pollutant and CO₂ emissions. It is calculated using the following equation:

$$\begin{array}{c}ElsC{O}_{2/a}={\displaystyle \frac{-\left(EC{O}_{2,a}+E^{\prime }C{O}_{2,a}\right)/EC{O}_{2,total}}{R_a/P_a}}\end{array}$$

(2)

where Els denotes the cross-elasticity coefficient; $EC{O}_{2,a}, E^{\prime }C{O}_{2,a}$ denote the direct CO₂ emissions (t) and indirect emissions (t) in the end-of-pipe treatment process, respectively; $EC{O}_{2,total}$ is the total amount of CO₂ emissions within the accounting scope of the enterprise (t); $R_a$ is the quantity of air pollutants removed (t); and $P_a$ is the quantity air pollutants generated (t).

The cross-elasticity coefficient is mainly used to measure the degree of response of changes in the emission reduction of one pollutant to changes in the emission reduction of another pollutant. When ElsCO₂/a < 0, it indicates that changing the end-of-pipe treatment facility or technology has a negative synergistic effect and the CO₂ emissions increase while air pollutant emissions decrease [26]. For each end-of-pipe treatment technology employed, the larger the absolute value of ElsCO₂/a, the greater its impact on CO₂ emissions [26,27,28].

(3)

Life cycle environmental impact assessment methodology

Life cycle assessment (LCA) is a comprehensive assessment method that considers a product’s resources, energy consumption, and environmental emissions throughout its life cycle. The “Life Cycle Assessment—Principles and Framework” (ISO 14040:2006), issued by the International Organization for Standardization (ISO) in 2006, provides the principles of and a framework for LCA, allowing the extensive adoption of this method worldwide [29,30].

Life cycle impact assessment (LCIA) is the core component of LCA, and its goal is to transform the data collected in a life cycle inventory into comparable environmental impact indicators. In LCIA, two main evaluation models exist: midpoint and endpoint models [31]. The midpoint model is an environmental problem-oriented approach, widely used in China because of its universality and relative simplicity. The endpoint model, on the other hand, is more concerned with impacts on human health, but the evaluation results yielded by the endpoint model are usually characterized by higher uncertainty. Thus, there are some limitations in its application [32]. Owing to a lack of localized research on life cycle environmental impact assessment models, Chinese scholars have more often adopted the midpoint assessment model when conducting life cycle environmental impact assessments regarding chemical enterprises.

SimaPro 9.4.0.2, a software product with a robust life cycle database and rich environmental impact assessment models, is currently one of the mainstream assessment tools. This software contains several authoritative databases, such as Ecoinvent, Agri-footprint, Industry Data 2.0, etc., which provide sufficient upstream data for researchers and greatly facilitate life cycle assessment and environmental impact analysis. In this study, the CML-IA midpoint model in the SimaPro software was used to carry out a life cycle environmental impact assessment, taking the emission characteristics of chemical enterprises into account [33,34]. Midpoint environmental impact indicators were selected for calculation and analysis (for details of the types of indicators and standardization factors, please refer to Appendix B,

Table A1. Names of environmental impact midpoint indicators and standardization factors.

Environmental Impact Indicator	Unit	Abbreviation	Standardized Factor
Abiotic depletion	kg Sb eq	ADP element	1.80 × 10−8
Abiotic depletion (fossil fuels)	MJ	ADP fossil	3.18 × 10−14
Global warming (GWP100a)	kg CO2 eq	GWP	1.99 × 10−13
Ozone depletion	kg CFC-11 eq	ODP	1.12 × 10−8
Human toxicity	kg 1,4-DB eq	HTP	1.29 × 10−13
Fresh water aquatic ecotoxicity	kg 1,4-DB eq	FAETP	1.93 × 10−12
Marine aquatic ecotoxicity	kg 1,4-DB eq	MAETP	8.57 × 10−15
Terrestrial ecotoxicity	kg 1,4-DB eq	TETP	2.06 × 10−11
Photochemical oxidation	kg C2H4 eq	POCP	1.18 × 10−10
Acidification	kg SO2 eq	AP	3.55 × 10−11
Eutrophication	kg PO4-eq	EP	7.58 × 10−11

). The indicators included the abiotic resource depletion potential (ADP element), fossil fuel depletion potential (ADP fossil), global warming potential (GWP), ozone depletion potential (ODP), human toxicity potential (HTP), freshwater aquatic ecotoxicity potential (FAETP), marine aquatic ecotoxicity potential (MAETP), terrestrial ecotoxicity potential (TETP), photochemical smog potential (POCP), acidification potential (AP), eutrophication potential (EDP), and eutrophication potential (EP).

(4)

Cross-elasticity coefficients based on ecological impacts

The use of standardized methods for indicators enables a comparison of the levels of environmental impact between different indicators [29]. Furthermore, selecting region-specific environmental impact standardization factors eliminates the differences in magnitude between midpoint indicators [26]. After applying these approaches, all standardized indicators were summed to derive the overall environmental impact load value of the product, which was calculated as follows:

$$\begin{array}{c}E{I}_{total}={{\displaystyle \sum }}_{c}E{I}_c={{\displaystyle \sum }}_c\left(I{R}_c\times N{F}_c\right)\end{array}$$

(3)

where the subscript c denotes the environmental impact indicator; $E{I}_{total}$ is the total environmental impact load value; $E{I}_c$ is the environmental impact load value of an indicator; $I{R}_c$ is the environmental impact indicator characterization value; and $N{F}_c$ is the standardization factor.

The synergistic effect of emission reduction policies is reflected in the changes in various indicators. Based on the characterization method used for the CML-IA model, CO₂ is characterized as a GWP factor, while SO₂, NO_x, and particulate matter are characterized as AP, HTP, and POCP, indicating that these pollutants have potential impacts on global climate change, human health, acidification, and photochemical smog formation [35]. Thus, four environmental impact indicators were selected: GWP, HTP, AP, and POCP. The cross-elasticity coefficients facilitate the development of a quantitative equation for the synergistic effect of atmospheric pollutant and GHG emission reductions within the future low-carbon development scenario based on the life cycle environmental impacts:

$$\begin{array}{c}S{E}_{{\displaystyle \frac{A}{C}}}={\displaystyle \frac{△E{I}_{AP}}{△E{I}_{GHG}}}={\displaystyle \frac{\frac{\sum E{I}_{AP,s}-E{I}_{AP,BAU}}{\sum E{I}_{AP,BAU}}}{\frac{\sum E{I}_{GHG,s}-E{I}_{GHG,BAU}}{\sum E{I}_{GHG,BAU}}}}\end{array}$$

(4)

where the subscript BAU denotes the baseline scenario (the policy-as-usual scenario); subscript s denotes the low-carbon development scenario; SE_A/C is the cross-elasticity coefficient based on ecological and environmental impacts; ΔEI_AP and ΔEI_GHG are the relative magnitudes of change in the environmental impact loads (in percent) associated with atmospheric pollutants and GHGs; EI_AP is the value of the environmental impact load associated with the emission of air pollutants; and EI_GHG represents the GWP load associated with GHG emissions.

A positive ${\mathrm{SE}}_{\mathrm{A}/\mathrm{C}}$ indicates that GHG emission reduction has a positive synergistic effect with respect to air pollutant emission reduction; on the contrary, if it is negative, it has a negative synergistic effect. If SE is equal to 0, it indicates that the change in GHG emissions will not have an impact on air pollutant emissions and their environmental impacts. When |SE_A/C| = 1, it indicates that, under a certain policy and measure scenario, the two factors above both have the same role and effect within that scenario [14,26]. For the GHG emission reduction policy within the low-carbon development scenario, if |SE_A/C| > 1, the policy has a more significant effect on air pollutant emissions while reducing GHG emissions. Conversely, |SE_A/C| < 1 indicates that the policy has a more significant effect on GHG emissions and a relatively weak synergistic effect on air pollutant reduction [14].

2.2.3. Scenario Analysis Methodology

Scenario analysis is a forecasting methodology in which researchers systematically assess potential situations and their consequences by assuming that particular phenomena or trends will persist in the future [3]. Based on a detailed study of three enterprises, and in light of the goal of the Chinese chemical industry to achieve peak carbon emissions by 2030, we made reasonable assumptions about the changes that may occur in this industry’s production processes under the influence of relevant policies. The three chemical companies’ baseline and low-carbon development scenarios were designed by adjusting the energy structure, process, and air pollution emissions in the life cycle inventory. The base year was set to 2021, and the baseline scenario was developed accordingly. The study’s time frame was from 2021 to 2035.

In developing the baseline scenario, we relied on existing measures, excluding potential future policy adjustments or technological innovations. The enterprises’ levels of utilization of energy reduction and efficiency technologies were maintained at the current level in the base year, serving as a reference point in assessing the emission reduction potential of the low-carbon development scenario [36]. In the model, the source structure of enterprises’ purchased electricity was based on installed capacity data for various power sources in the base year published by the National Bureau of Statistics [26] and on the upstream power structure above. The purchased electricity was proportionally allocated to five types of power: thermal power, hydropower, nuclear power, wind power, and solar power. The life cycle inventory data for these five types of electricity production were obtained from the Ecoinvent database in the SimaPro software.

The low-carbon development scenarios were mainly based on China’s 2030 carbon peak policies. Carbon emissions from refining and chemical enterprises can be categorized into two main groups based on their sources: carbon emissions from fossil fuel combustion and carbon emissions from processes [3]. The relevant parameters of the life cycle evaluation under the different scenarios were established by adjusting the input and output data of the life cycle inventories of the three chemical enterprises. The upstream power structures of the enterprises changed significantly, and the proportion of power consumed by the enterprises changed when adjusting the upstream power structure according to the planning targets for the five types of installed power capacity for 2030 defined in China Energy Outlook 2030 [37]. These enterprises will realize technological improvements and upgrades under the low-carbon development scenario. We considered initiatives developed by oil refineries to adopt green hydrogen instead of grey hydrogen, ethylene, or chlor-alkali in order to reduce their energy consumption, such as that of electricity and steam, through technological advancement and management optimization. Based on these adjusted data, the life cycle inventory was updated to reflect the environmental impacts of the enterprises under different scenarios. Appendix B (

Table A2. Chemical industry future (2035) scenario setting and related parameters.

Type	Name	Unit	Refinery Project		Ethylene Project		Chlor-Alkali Project
			Baseline Scenario	Low-Carbon Development Scenario	Baseline Scenario	Low-Carbon Development Scenario	Baseline Scenario	Low-Carbon Development Scenario
Upstream power structure	Thermal power	%	66.5	42.39	66.5	42.39	66.5	42.39
	Hydroelectricity	%	15.28	18.7	15.28	18.7	15.28	18.7
	Nuclear energy	%	4.72	5.65	4.72	5.65	4.72	5.65
	Wind power	%	8.61	18.7	8.62	18.7	8.62	18.7
	Solar power	%	4.83	14.55	4.83	14.55	4.83	14.55
Process improvement			Substitution of green hydrogen for grey hydrogen		Total energy consumption of electricity and steam decreased by 1.5%		Total energy consumption of electricity and steam decreased by 1.5%

Note: In the baseline scenario, the power generation structure of externally purchased electricity by enterprises is based on the installed capacity data of various power sources for the benchmark year as published by the National Bureau of Statistics. The installed capacity proportions for thermal, hydropower, nuclear, wind, and photovoltaic power are 66.5%, 15.28%, 4.72%, 8.61%, and 4.83%, respectively.

) details the scenario settings and related parameters.

3. Results

3.1. Accounting for Air Pollutant and GHG Emissions

3.1.1. Accounting for Air Pollutant Emissions

(1)

Refinery Enterprises

Table 2. Summary of SH Petrochemical’s organic air pollutant emission accounting results (t/a).

Pollutant	SO2	NOx	Particulate Matter	NMHC	VOCs
Process emissions	17.33	41.83	2.13	13.96	/
Combustion process emissions	26.48	355.44	43.83	/	76.70
Total	43.81	397.27	45.96	13.96	76.70

summarizes the total emissions of organized atmospheric pollutants from the processes and combustion activities of SH Petrochemical (in the refining sector). Detailed control measures and corresponding emission calculations are provided in Appendix C (

Table A3. SH Petrochemical’s process devices, organized air pollutants, control measures, and emission accounting.

Device	Serial Number	Source of Pollution	Governance Measures	Exhaust Volume/ (Nm3/h)	Generation (t/a)				Emissions (t/a)
					SO2	NOx	Particulate Matter	NMHC	SO2	NOx	Particulate Matter	NMHC
Boiling Bed Residue Hydrogenation Plant	G1-2-1	Heating furnace flue gas	Low-sulfur fuel gas, ultra-low-NOx burners	60,000	2.02	20.16	1.01	1.01	2.02	20.16	1.01	1.01
Hydrocracker	G1-3-1	Heating furnace flue gas	Low-sulfur fuel gas, ultra-low-NOx burners	64,500	2.17	21.67	1.08	1.08	2.17	21.67	1.08	1.08
Sulfur Recovery Combined Unit	G1-6-1	Incinerator tail gas	RASOC	62,590	600.94	/	/	/	13.15	/	/	/
POX Hydrogen Plant	G1-7-1	Petroleum rubber silo exhaust	Bag filter	1000	/	/	8.40	0.17	/	/	0.04	0.17
	G1-7-2	Slag pond air release	/	162	/	/	/	0.03	/	/	/	0.03
	G1-7-3	Acid gas Stripping tail gas	Washing tower	13,056	/	/	/	54.67	/	/	/	11.68
Total					605.12	41.83	10.49	56.95	17.33	41.83	2.13	13.96

Note: POX: petroleum coke to hydrogen; RASOC: regenerable absorption process for SO_x cleanup, i.e., renewable wet flue gas desulfurization technology.

and

Table A4. SH Petrochemical’s combustion processes, organic air pollutant control measures, and emission accounting.

Type	Serial Number	End-of-Pipe Process	SO2 Generation (t/a)	SO2 Emissions (t/a)	NOx Generation (t/a)	NOx Emissions (t/a)	Particulate Matter Generation (t/a)	Particulate Matter Emissions (t/a)	VOC Generation (kg/a)	VOC Emissions (kg/a)
Oil boiler	MF0342	Desulfurization—sodium hydroxide; Denitrification—selective catalytic reduction; Dedusting—Venturi	129.9	0.73	369.24	71.41	*	11.77	0	0
	MF0343		135.57	0.7	313.86	72.53	*	11.42	0	0
	Subtotal		265.47	1.43	683.1	143.94	*	23.19	0	0
Heating crude oil	MF0922	/	4.13	4.13	34.84	34.84	3.4	3.4	12,619.98	12,619.98
	MF0923		4.13	4.13	34.84	34.84	3.4	3.4	12,619.98	12,619.98
	MF0924		4.13	4.13	34.84	34.84	3.4	3.4	12,619.98	12,619.98
	MF0938		6.33	6.33	53.49	53.49	5.22	5.22	19,418.95	19,418.95
	MF0939		6.33	6.33	53.49	53.49	5.22	5.22	19,418.95	19,418.95
	Subtotal		25.05	25.05	211.5	211.5	20.64	20.64	76,697.84	76,697.84
Total			290.52	26.48	894.6	355.44	*	43.83	76,697.84	76,697.84

Note: Data for each sector are from the EIA report, and data for oil boilers are from the 2017 EMS data. “*” indicates missing data.

). Table 2 clearly reveals that the annual organized emissions of SO₂, NO_x, and particulate matter from SH Petrochemical amount to 43.81 tons, 397.27 tons, and 45.96 tons, respectively. Notably, NO_x emissions occur across multiple production units, with fuel combustion making the largest contribution, accounting for 89.47% of the total emissions (Figure 4). Particulate matter emissions also predominantly stem from fuel combustion, constituting 95.37% of the total particulate emissions in the refining sector. The generation and emission of SO₂ are primarily concentrated in the sulfur recovery units and fuel combustion processes, with annual production amounts of 600.94 tons and 290.52 tons, respectively. After desulfurization using the regenerable absorption process for SO_x cleanup (RASOC) and sodium hydroxide desulfurization, the removal efficiency exceeds 90%. NMHC, CO, and H₂S emissions are mainly associated with processing operations, likely due to SH Petrochemical’s use of asphalt and petroleum coke gasification for hydrogen production, generating CO and H₂S as major byproducts. Additionally, during the removal of acidic gases in the partial oxidation (POX) hydrogen production units, significant amounts of NMHC are generated in the exhaust. Organized VOC emissions are primarily concentrated in the combustion of fuel during the heating of crude oil, while process-related VOC emissions are mainly unorganized, originating from sealing leaks (

Table 3. SH Petrochemical’s unorganized VOC emissions accounting.

Device	Sealing Points (pcs)								VOC Emissions /(t/a)
	Connectors	Open-Ended Valves or Open-Ended Lines	Valves	Compressors, Agitators, Pressure Relief Equipment	Pump	Flange	Other	Total
Boiling bed residue hydrogenation unit	0	0	9500	6	30	16,500	0	26,036	50.73
High-pressure hydrocracker	0	0	5230	238	31	7401	0	12,900	24.78
POX hydrogen generation unit	0	0	3620	229	54	2734	0	6637	12.22
PSA hydrogen generation unit	0	0	75	1	0	165	0	241	0.48
Sum								45,814	88.21

).

Table 3 summarizes the calculated results regarding the unorganized VOC emissions generated by SH Petrochemical (refining sector), with an annual VOC generation capacity of 88.21 tons. The primary emission source is leakage from sealing points, with the hydrocracking unit for residual oil having the most sealing points and the most significant VOC emissions, reaching 50.73 tons per year. As there is no VOC removal process in the facility used, SH Petrochemical’s total VOC generation and emission degree from organized and unorganized sources for 2021 amounts to about 164.91 tons. Among these sources, unorganized emissions slightly exceed organized emissions, accounting for 53.49%.

(2)

Ethylene Producers

Table 4. SK Ethylene’s conventional air pollutant accounting results (t/a).

Source of Pollution	NOx			TSP			SO2
	Generation	Reduction	Emission	Generation	Reduction	Emission	Generation	Reduction	Emission
Process installations	3784.26	1435.82	2348.44	2002.01	1902.26	99.75	675.3	492.83	182.46
Power center and cogeneration	2806.49	1364.91	1441.57	19.13	6.31	12.82	115.20	62.21	52.99
Total	4203.80	1965.86	2237.94	2021.13	1908.57	112.56	790.49	555.04	235.44

and

Table 5. SK Ethylene’s VOC emission accounting results (t/a).

Type	VOCs
	Generation	Reduction	Emission
Organized emissions	32,597.31	31,969.57	627.74
Unorganized emissions	780.84	153	627.83
Total	33,378.15	32,122.57	1255.57

present the results regarding the emissions of conventional air pollutants and VOCs by SK Ethylene.

Table A5. SK Ethylene’s air pollutant emission accounting.

Device	NOx				TSP				SO2
	Governance Measures	Generation (t/a)	Emission Reduction (t/a)	Emissions (t/a)	Governance Measures	Generation (t/a)	Emission Reduction (t/a)	Emissions (t/a)	Disposal Measures	Generation (t/a)	Emission Reduction (t/a)	Emissions (t/a)
Ethylene plant	PCC’s patented technology for nitrogen removal	2386.95	834.87	1552.08	Not present	90.19	0	90.19	Not present	2.26	0	2.26
Ethylbenzene/styrene (EB/SM) plant	Low-nitrogen burners	384.66	134.63	250.03	Not present	0	0	0	Not present	5.68	0	5.68
Polystyrene (PS) units	Low-nitrogen burners	11.79	4.13	7.66	Not present	0	0	0	Not present	0	0	0
Acrylonitrile (I and II) plant	AOGI PCC denitrification	861.66	424.34	437.32	Bag filter	1911.82	1902.26	9.56	Not present	8.82	0	8.82
Sulfuric acid recovery (SAR) unit	Organic catalytic flue gas desulfurization and denitrification	139.20	37.85	101.35	Not present	0	0	0	Organic catalytic flue gas desulfurization and denitrification	658.54	492.83	165.70
Butadiene (BEU1 and 2) plant	Incorporation into acrylonitrile (I and II) unit				Incorporation into acrylonitrile (I and II) unit				Incorporation into acrylonitrile (I and II) unit
Power center	SCR denitrification	1125.42	776.54	348.88	Ammonia desulfurization and dedusting	19.13	6.31	12.82	Ammonia desulfurization and dedusting	115.20	62.21	52.99
Combined heat and power supply	Low-nitrogen burners	1681.07	588.37	1092.69	Not present	0	0	0	Not present	0	0	0
Total	/	4203.80	1965.86	2237.94	/	2021.13	1908.57	112.56	/	790.49	555.04	235.44

and

Table A6. SK Ethylene’s VOC emission accounting.

Type	Segment	Disposal Measures	VOCs
			Generation (t/a)	Emission Reduction (t/a)	Emissions (t/a)
Organized emissions	Combustion flue gas emissions	Absorption tower + AOGI thermal incineration	9187.33	9028.53	158.80
	Organized emissions from processes	Not present	0.3	0	0.3
	Flare emissions	Flare thermal incineration	23,409.68	22,941.04	468.64
	Subtotal		32,597.31	31,969.57	627.74
Unorganized emissions	Organic liquid storage and reconciliation of volatilization losses	AOGI thermal incineration	162.14	153	9.13
	Leakage at static and dynamic sealing points of equipment	Not present	259.72	0.00	259.72
	Cooling tower, circulating water, cooling system release	Not present	358.98	0.00	358.98
	Subtotal		780.84	153	627.83
Total			33,378.15	32,122.57	1255.57

in Appendix C provide a comprehensive overview of the pollutant emission accounting results for each compound. As illustrated in Table 4, the aggregate annual emissions of NO_x, TSP, and SO₂ are 2237.94 tons, 112.56 tons, and 235.44 tons, respectively. NO_x is generated and emitted by multiple production units, with processing units generating and emitting the highest amounts. TSP is predominantly generated by processing units, with a generation rate of 2002.01 tons. However, these emissions decreased to 99.75 tons after efficient dust removal, primarily for units lacking dedusting equipment. SO₂ generation and emission are also primarily concentrated in the processing units, primarily via sulfuric acid recovery. SO₂ generated and emitted by processing units accounted for 85.43% and 77.50% of ethylene production, respectively.

SK Ethylene emitted 1255.6 tons of VOCs in 2017. Among the various pollution sources, the generation and emission of VOCs from organized emissions were the largest, mainly concentrated in combustion flue gas emissions and flare emissions.

(3)

Chlor-Alkali Producers

The atmospheric pollutant results for HS Chlor-Alkali are presented in

Table 6. HS Chlor-Alkali’s air pollutant accounting (t/a).

Device	Cl2				HCl
	Disposal Measures	Generation	Reduction	Emissions	Disposal Measures	Generation	Reduction	Emission
Caustic soda plant	Alkali absorption	10.71	10.70	0.01	Falling film absorption + water jet	2.08	2.00	0.08
Ethylene dichloride (EDC) plant	/	0.04	0.00	0.04	/	0.03	0.00	0.03
Hydrogen boiler	/	0.00	0.00	0.00	/	0.00	0.00	0.00
Total		10.75	10.70	0.05		2.11	2.00	0.11

. The specific pollutants emitted by this company are Cl₂ and HCl, with annual emissions of 0.05 tons and 0.11 tons, respectively. The caustic soda unit is the primary source of Cl₂, accounting for 99.63% of the total Cl₂ generated by this company. As a result of efficient end-of-pipe treatment, this company’s Cl₂ removal efficiency is 99.91%, so the caustic soda unit only represents a 20.00% contribution to the company’s total Cl₂ emissions. Of all plants, caustic soda plants produce the highest amounts of HCl, accounting for 98.58% of the entire company’s HCl emissions. Falling film absorption and water jet ejector end treatments significantly reduce the HCl emissions.

Table 7. HS Chlor-Alkali’s VOC emission accounting (t/a).

Type	VOCs
	Generation	Reduction	Emission
Organized emissions	9.20	9.11	0.09
Unorganized emissions	166.46	0.00	166.46
Total	175.66	9.11	166.55

summarizes the organized and unorganized VOC emission accounting results for HS Chlor-Alkali (detailed in Appendix C,

Table A7. HS Chlor-Alkali’s VOC emission accounting.

Type	Segment	Device	VOCs				Note
			Disposal Measures	Generation (t/a)	Emission Reduction (t/a)	Emissions (t/a)
Unorganized emissions	Organic liquid storage and reconciliation of volatilization losses	Storage tank area	Not present	156.88	0.00	156.88
	Leakage at static and dynamic sealing points of equipment	/	Not have	9.58	0.00	9.58
Organized emissions	Organized emissions from processes	Caustic soda plant	Not present	/	/	/
		EDC device	High-temperature incineration	9.20	9.11	0.09
		Hydrogen boiler	Not present	/	/	/
	Combustion flue gas emissions	/	/	/	/	/	Use of hydrogen
	Flare emissions	/	/	/	/	/	No flare
Total				175.66	9.11	166.55

). The annual VOC emissions for HS Chlor-Alkali amount to 166.55 tons. The highest unorganized VOC emissions result from volatilization losses during the storage and blending of organic liquids and leaks from dynamic and static equipment seals, accounting for 99.95% of the company’s total VOC emissions. The organized VOC emissions are limited to process-related emissions, which are minimal after applying efficient high-temperature incineration end-of-pipe treatment technology.

3.1.2. Accounting for Carbon Dioxide Emissions

Data on the electricity and heat, fueled by natural gas, produced by SH Petrochemical (refining sector), SK Ethylene, and HS Chlor-Alkali were supplied by the Shanghai Chemical Industry Zone (SCZ) Cogeneration Project, and these companies’ CO₂ emission factors were calculated according to the emission factor method specified in the “2020 Average Carbon Dioxide Emission Factor of China’s Regional and Provincial Electricity Grids” document. The CO₂ emissions of SH Petrochemical (refining sector), SK Ethylene, and HS Chlor-Alkali are presented in

Table 8. Summary of carbon dioxide emissions generated by SH Petrochemical (refining sector), SK Ethylene, and HS Chlor-Alkali (t/a).

Enterprise	Sources of CO2 Emissions	CO2 Emissions
SH Petrochemical	CO2 emissions from fuel combustion	4,282,108
	Implied CO2 emissions from net purchases of electricity and heat	128,148
	Total	4,410,256
SK Ethylene	CO2 emissions from fuel combustion	304,915
	CO2 emissions from flare combustion	209,024
	CO2 emissions from processes	2011
	Implied CO2 emissions from net purchases of electricity and heat	485,950
	Direct carbon emissions from end-of-pipe management	28,818
	Total	1,030,719
HS Chlor-Alkali	CO2 emissions from fuel combustion	3139
	CO2 emissions from processes	14,773
	Implied CO2 emissions from net purchases of electricity and heat	707,580
	Direct carbon emissions from end-of-pipe management	139
	Total	725,631

. The parameters and results are described in Appendix C (

Table A8. SH Petrochemical’s fuel combustion CO₂ emissions.

Type of Fuel	Annual Use	Unit	Calorific Value (GJ/t, GJ/million m3)	Carbon Content Per Unit Calorific Value (tC/GJ)	Carbon Oxidation Rate	CO2 Emissions (tCO2/a)
Fuel oil	46,386	t/a	41.86	0.015	0.99	317,723.08
Petroleum	183,026.15	million m3/a	390	0.015	0.99	3,964,384.78
Total						4,282,107.86

Note: Energy consumption data are from 2017 environmental statistics.

,

Table A9. SK Ethylene project’s fuel combustion CO₂ emissions.

Type of Fuel	Annual Use	Unit	Calorific Value (GJ/t, GJ/million m(3))	Carbon Content per Unit Calorific Value (tC/GJ)	Carbon Oxidation Rate	CO2 Emissions (tCO2/a)	Other Information
Fuel oil	95,700	t/a	42	0.0211	0.98	304,748.23	Self-production
Fuel gas	7.9	Million m3/a	380	0.0153	0.99	166.73	Blending of plant-produced hydrogen methane with purchased natural gas, with hydrogen methane components similar to natural gas
Total						304,914.96

,

Table A10. SK Ethylene project’s flare CO₂ emissions.

Flare Model		Flare Gas Flow Rate (Million Nm3/a)	Total Carbon Content of Carbon-Containing Compounds (Tons of Carbon/10,000 Nm3)	Carbon Oxidation Rate	CO2 Emissions (tCO2/a)
HP-1		2639.34	6.52	0.98	61,812.06
HP-2		1376.69	9.71	0.98	48,009.87
LP-A		484.04	12.50	0.98	21,736.93
LP-B		673.34	4.70	0.98	11,377.80
LP-C		2330.20	7.85	0.98	65,768.37
Low-temperature flare	Inlet pipeline 1	0.00	13.82	0.98	0.48
	Inlet pipeline 2	0.13
Acrylonitrile flare	Craft torch	7.38	10.53	0.8	318.80
	Hydrocyanic acid Norming furnace	131.40
Total					209,024.31

,

Table A11. SK Ethylene project’s process CO₂ emissions (ethylene cracker).

Serial Number	Average Tail Gas Flow (Nm3/h)	Annual Cumulative Coking Time (h/a)	Volume Concentration of CO2 in Exhaust Gas (%)	CO2 Emissions (tCO2/a)
1	90,100	216	0.20%	76.68
2	272,000	216	0.20%	231.48
3	272,000	216	0.20%	231.48
4	252,000	216	0.20%	214.46
5	258,000	216	0.20%	219.57
6	276,000	216	0.20%	234.89
7	220,000	216	0.20%	187.23
8	249,000	216	0.20%	211.91
9	240,000	216	0.20%	204.25
10	120,000	216	0.20%	102.12
11	114,000	216	0.20%	97.02
Total				2011.09

,

Table A12. Direct CO₂ emissions related to end-of-pipe treatment in SK Ethylene project.

Pollutant	Department	End-of-Pipe Management Equipment	Direct CO2 Emissions from End-of-Pipe Treatment (t/a)
VOCs	Vertical fixed roof tanks	AOGI thermal incineration	480.22
	Acrylonitrile I	AOGI thermal incineration	28,337.55
	Total		28,817.76

,

Table A13. HS Chlor-Alkali project’s fuel combustion CO₂ emissions.

Type of Fuel	Annual Use	Unit	Source	Low-Level Heat Generation (GJ/million Nm3)	Carbon Content per Unit Calorific Value (tC/GJ)	Carbon Oxidation Rate	CO2 Emissions (tCO2/a)
Petroleum	145.2	million Nm3/a	Natural gas pipeline in chemical zone	389.31	0.0153	0.99	3139.50
Total							3139.50

,

Table A14. HS Chlor-Alkali’s process CO₂ emissions.

Device	CO2 Emissions from Raw Material Consumption					CO2 Emissions from Carbonate Use Processes
	Carbon-Containing Raw Materials	Dosage (104t/a)	Carbon-Containing Product	Production (10,000 t/a)	CO2 Emissions (tCO2/a)	Carbonate Type	Consumption (t/a)	CO2 Emission Factor (Tons CO2/Tons Carbonate)	Purity of Carbonate (%)	CO2 Emissions (tCO2/a)
Caustic soda plant	/	0	Not present	0	0	Na2CO3	5000	0.4149	100	2074.5
EDC device	Ethylene	12	Refined EDC	41	12,698.41	/	0	/	/	0
Total	14,772.91

and

Table A15. Direct CO₂ emissions associated with end-of-pipe management in HS Chlor-Alkali project.

Pollutant	Device	Exhaust Port	End-of-Pipe Management Equipment	Direct CO2 Emissions from End-of-Pipe Treatment (t/a)
VOCs	EDC device	1# Incinerator	High-temperature incineration	138.99
		2# Incinerator	High-temperature incineration
	Total			138.99

).

According to our calculations, the total carbon dioxide emissions generated by SH Petrochemical (refining sector) amount to 4.41 million tons annually, with fuel combustion accounting for 97.1% of these emissions.

SK Ethylene emits 1.03 million tons of CO₂ annually, with its emission sources including fuel combustion, flare combustion, process production, direct carbon emissions from end-of-pipe treatment, and CO₂ emissions embedded in net purchased electricity and heat, accounting for 29.6%, 20.3%, 0.2%, 2.8%, and 47.1%, respectively. Indirect CO₂ emissions from the end treatment of ethylene projects—that is, carbon emissions from end-treatment electricity consumption—are included in the CO₂ emissions implied by the net purchased electricity and heat.

HS Chlor-Alkali emits 0.73 million tons of CO₂ annually, with its emission sources comprising fuel combustion, process production, direct carbon emissions from end-of-pipe treatment, and CO₂ emissions embedded in net purchased electricity and heat, totaling four categories. CO₂ emissions embedded in net purchased electricity and heat account for 97.5%, while direct carbon emissions from end-of-pipe treatment, fuel combustion, and process production account for 0.019%, 0.43%, and 2.03%, respectively. Indirect CO₂ emissions from end-of-pipe treatment due to electricity consumption are included in the CO₂ emissions embedded in the net purchased electricity and heat.

Among the three enterprises, SH Petrochemical has the highest CO₂ emissions by far, followed by SK Ethylene, with HS Chlor-Alkali having the lowest emissions. Summing the CO₂ emissions from different sources across these enterprises reveals that fuel combustion is the most significant contributor, accounting for 74.4% of the total emissions. Indirect CO₂ emissions from purchased electricity and heat are also substantial, comprising 21.4% and representing a major source of carbon emissions for these enterprises.

3.2. Air Pollution End-of-Pipe Management Synergies

End-of-pipe technologies have been widely employed to reduce the emission of air pollutants during chemical production. Although technologies can significantly reduce air pollutant emissions, the chemical reactions and power consumption accompanying the removal of these pollutants may increase CO₂ emissions. Therefore, when utilizing end-of-pipe technologies, their synergistic effects with respect to air pollutant level reduction and GHG emission need to be comprehensively considered. The synergistic coefficients of the corresponding end-of-pipe treatments for different air pollutants for SH Petrochemical (refining sector), SK Ethylene, and HS Chlor-Alkali are shown in

Table 9. SH Petrochemical’s synergistic coefficients.

Pollutant	Source of Emissions	Emission Reduction Technologies/End-of-Pipe Equipment	Synergistic Coefficient: Amount of Additional CO2 Emitted per Ton of Air Pollutant Reduced (t, Both Direct and Indirect)
SO2	Sulfur recovery combined unit	RASOC renewable wet flue gas	7.4
NOx	Oil boilers	SCR	1.9
Particulate matter	POX hydrogen generation plant	Bag filter	0.004

Notes: RASOC, regenerable absorption process for SO_x cleanup; SCR, selective catalytic reduction; POX, partial oxidation.

,

Table 10. Synergistic coefficients for SK Ethylene.

Pollutant	Source of Emissions	Emission Reduction Technologies/End-of-Pipe Equipment	Synergistic Coefficient: Amount of Additional CO2 Emitted per Ton of Air Pollutant Reduced (t, Both Direct and Indirect)
NOx	Power center	SCR denitrification	1.2
Particulate matter	Acrylonitrile I, II	Bag filter	0.004
VOCs	Storage tank, acrylonitrile I and II, power center	AOGI thermal incineration	3.8
	Ground torch, cryogenic torch, acrylonitrile torch	Flare thermal incineration	9.2
SO2, particulate matter	Power center	Ammonia desulfurization and dedusting	53.1
SO2, NOx	Sulfuric acid recovery	Organic catalytic flue gas desulfurization and denitrification	4.2

Notes: AOGI, absorber off-gas incinerator.

, and

Table 11. Synergistic coefficients for HS Chlor-Alkali.

Pollutant	Source of Emissions	Emission Reduction Technologies/End-of-Pipe Equipment	Synergistic Coefficient: Amount of Additional CO2 Emitted per Ton of Air Pollutant Reduced (t, Both Direct and Indirect)
Cl2	Caustic soda unit	Alkali absorption	26.9
HCl	Caustic soda unit	Falling film absorption + water jet	425.9
VOCs	EDC unit	High-temperature incineration	70.7

Notes: EDC, 1,2-dichloroethane.

, respectively, and are detailed in Appendix D (

Table A16. SH Petrochemical’s synergy factors.

Pollutant	Department	Emission Reduction Technologies/ End-of-Pipe Equipment	Equipment Power/ Electricity Consumption Factor	Emission Reduction (t/a)	Electricity Consumption (kWh/a)	Indirect CO2 Emissions (t/a)	Synergistic Coefficient: Amount of Additional CO2 Emitted per Ton of Air Pollutant Reduced (t, both Direct and Indirect)
SO2	Boiling bed residue hydrotreating unit	Low-sulfur fuel gas	/	/	/	/	/
	Hydrocracker		/	/	/	/	/
	Sulfur recovery combined unit	RASOC renewable wet flue gas	1280 kW	587.79	10,752,000.00	4351.97	7.40
	Oil boiler	NaOH	/	264.04	/	/	/
NOx	Boiling bed residue hydrotreating unit	Ultra-low-nitrogen burners	/	/	/	/	/
	Hydrocracker		/	/	/	/	/
	Oil boiler	SCR	295.5 kW	539.16	2,511,750.00	1016.65	1.89
Particulate matter	POX hydrogen plant	Bag filter	10 kWh/tTSP	8.36	83.58	0.03	0.0041
Total				1399.35	13,263,833.58	5368.66	3.84

,

Table A17. SK Ethylene project’s synergistic coefficients.

Pollutant	Department	Emission Reduction Technologies/ End-of-Pipe Equipment	Equipment Power/Electricity Consumption Factor	Emission Reduction (t/a)	Electricity Consumption (kWh/a)	Direct CO2 Emissions (t/a)	Indirect CO2 Emissions (t/a)	Synergistic Coefficient: Amount of Additional CO2 Emitted per Ton of Air Pollutant Reduced (t, both Direct and Indirect)
NOx	Acrylonitrile I	AOGI PCC denitrification	/	236.38	/	0	/	/
	Acrylonitrile II		/	187.97	/	0	/	/
	Power center	SCR denitrification	295.5 kW	776.54	2,364,000	0	956.85	1.23
	Subtotal	/	/	776.54	2,364,000	0	956.85	1.23
Particulate matter	Acrylonitrile I	Bag filter	10 kWh/tTSP	1628.59	16,285.9	0	6.59	0.00405
	Acrylonitrile II		10 kWh/tTSP	273.66	2736.6	0	1.11	0.00405
	Subtotal	/	/	1902.25	19,022.5	0	7.70	0.00405
VOCs	Vertical fixed roof tanks	AOGI thermal incineration	1538.32 kWh/t (VOC)	100.27	154,247.35	480.22	62.43	3.76
	Floating roof tanks			52.73	81,115.61		32.83
	Acrylonitrile I			4876.14	7,501,058.64	28,337.55	3036.12	3.76
	Acrylonitrile II			4114.13	6,328,841.37		2561.66
	Power center			36.03	55,425.92		22.43
	Ground flare	Flare thermal incineration	611.41 kW	22,938.15	4,891,307.57	209,024.31	1979.80	9.20
	Low-temperature flare			0.91	1395.99		0.57
	Acrylonitrile flare			1.98	3027.78		1.23
	Subtotal	/	/	32,120.33	19,016,420.23	237,842.08	7697.07	7.64
SO2, particulate matter	Power center	Ammonia desulfurization and dust removal	1124.3 kW	62.21 t SO2; 6.31 t TSP	8,994,400	0	3640.56	53.13
SO2, NOx	Sulfuric acid recovery	Organocatalytic flue gas desulfurization	697.32 kW	492.83 t SO2; 37.85 t NOx	5,578,560	0	2257.97	4.25
Total				35,398.32	35,972,402.73	237,842.08	14,560.15	7.13

,

Table A18. HS Chlor-alkali project’s synergies.

Pollutant	Device	Exhaust Port	End-of-pipe Management Equipment	Equipment Power/ Electricity Consumption Factor	Emission Reduction (t/a)	Electricity Consumption (kWh/a)	Direct CO2 Emissions (t/a)	Indirect CO2 Emissions (t/a)	Synergistic Coefficient: Amount of Additional CO2 Emitted per Ton of Air Pollutant Reduced (t, both Direct and Indirect)
Cl2	Caustic soda plant	1# Waste chlorine absorption tower	Alkali absorption	66,458.98 kWh/(ton Cl2)	8.73	711,111.11	0	287.83	26.90
		2# Waste chlorine absorption tower			0.67
		3# Waste chlorine absorption tower			1.30
	Subtotal				10.70	711,111.11	0.00	287.83
HCl	Caustic soda plant	1# Hydrochloric acid tail gas treatment (A furnace)	Falling film absorption + water jet	1,052,380.95 kWh/(ton HCl)	0.20	2,103,634.59	0	851.47	425.95
		1# Hydrochloric acid tail gas treatment (B furnace)			0.00
		1# Hydrochloric acid tail gas treatment (C furnace)			0.00
		1# Hydrochloric acid tail gas treatment (D furnace)			0.11
		1# Hydrochloric acid tail gas treatment (E furnace)			0.24
		1# Hydrochloric acid tail gas treatment (F furnace)			1.44
	Subtotal				2.00	2,103,634.59	0.00	851.47
VOCs	EDC device	1# Incinerator	High-temperature incineration	136,915.60 kWh/(tonVOCs)	4.56	1,247,287.57	138.99	504.85	70.67
		2# Incinerator			4.56
	Subtotal				9.11	1,247,287.57	138.99	504.85
Total					21.81	4,062,033.27	138.99	1644.15	81.76

,

Table A19. SH Petrochemical’s life cycle environmental impact assessment results for the production of one ton of refined diesel fuel.

Environmental Impact Indicator	Unit	Abbreviation	$\mathsf{I}{\mathsf{R}}_{\mathsf{c}}$ Standardized Factor	Standardized Environmental Load Value	Percentage of Environmental Load
Abiotic depletion	ADP element	kg Sb eq	7.28 × 10−5	1.31 × 10−12	0.02%
Abiotic depletion (fossil fuels)	ADP fossil	MJ	118,867.92	3.78 × 10−9	69.48%
Global warming (GWP100a)	GWP	kg CO2 eq	1035.18	1.71 × 10−10	3.79%
Ozone depletion	ODP	kg CFC-11 eq	3.60 × 10−5	4.07 × 10−13	0.01%
Human toxicity	HTP	kg 1,4-DB eq	182.95	2.32 × 10−11	0.43%
Fresh water aquatic ecotoxicity	FAETP	kg 1,4-DB eq	1.58	2.97 × 10−12	0.06%
Marine aquatic ecotoxicity	MAETP	kg 1,4-DB eq	1.31 × 105	9.89 × 10−10	20.59%
Terrestrial ecotoxicity	TETP	kg 1,4-DB eq	0.16	2.97 × 10−12	0.06%
Photochemical oxidation	POCP	kg C2H4 eq	1.04 × 10−1	1.20 × 10−11	0.23%
Acidification	AP	kg SO2 eq	1.20	3.89 × 10−11	0.78%
Eutrophication	EP	kg PO4- eq	3.27	2.47 × 10−10	4.56%
Abiotic depletion	/	/	/	5.44 × 10−9	100.00%

,

Table A20. SK Ethylene’s life cycle environmental impact assessment results for the production of one ton of ethylene.

Environmental Impact Indicator	Unit	Abbreviation	$\mathsf{I}{\mathsf{R}}_{\mathsf{c}}$ Standardized Factor	Standardized Environmental Load Value	Percentage of Environmental Load
Abiotic depletion	ADP element	kg Sb eq	3.05 × 10−4	3.60 × 10−12	0.05%
Abiotic depletion (fossil fuels)	ADP fossil	MJ	1.19 × 105	3.80 × 10−9	53.94%
Global warming (GWP100a)	GWP	kg CO2 eq	1.83 × 103	3.63 × 10−10	5.15%
Ozone depletion	ODP	kg CFC-11 eq	1.53 × 10−3	1.71 × 10−11	0.24%
Human toxicity	HTP	kg 1,4-DB eq	261	3.37 × 10−11	0.48%
Fresh water aquatic ecotoxicity	FAETP	kg 1,4-DB eq	39.5	7.63 × 10−11	1.08%
Marine aquatic ecotoxicity	MAETP	kg 1,4-DB eq	2.58 × 105	2.21 × 10−9	31.37%
Terrestrial ecotoxicity	TETP	kg 1,4-DB eq	0.921	1.90 × 10−11	0.27%
Photochemical oxidation	POCP	kg C2H4 eq	0.438	5.17 × 10−11	0.73%
Acidification	AP	kg SO2 eq	10.6	3.77 × 10−10	5.35%
Eutrophication	EP	kg PO4- eq	1.24	9.40 × 10−11	1.33%
Abiotic depletion	/	/	/	7.05 × 10−9	100.00%

and

Table A21. HS Chlor-Alkali’s life cycle environmental impact assessment results for the production of one ton of chlor-alkali.

Environmental Impact Indicator	Unit	Abbreviation	$\mathsf{I}{\mathsf{R}}_{\mathsf{c}}$ Standardized Factor	Standardized Environmental Load Value	Percentage of Environmental Load
Abiotic depletion	ADP element	kg Sb eq	0.00504	5.95 × 10−11	9.22%
Abiotic depletion (fossil fuels)	ADP fossil	MJ	6.94 × 103	2.21 × 10−10	34.23%
Global warming (GWP100a)	GWP	kg CO2 eq	805	1.60 × 10−10	24.78%
Ozone depletion	ODP	kg CFC-11 eq	9.72 × 10−6	1.09 × 10−13	0.02%
Human toxicity	HTP	kg 1,4-DB eq	41.2	5.32 × 10−12	0.82%
Fresh water aquatic ecotoxicity	FAETP	kg 1,4-DB eq	1.21	2.34 × 10−12	0.36%
Marine aquatic ecotoxicity	MAETP	kg 1,4-DB eq	1.68 × 104	1.44 × 10−10	22.31%
Terrestrial ecotoxicity	TETP	kg 1,4-DB eq	1.47 × 10−1	3.03 × 10−12	0.47%
Photochemical oxidation	POCP	kg C2H4 eq	4.70 × 10−2	5.55 × 10−12	0.86%
Acidification	AP	kg SO2 eq	1.08	3.83 × 10−11	5.93%
Eutrophication	EP	kg PO4- eq	8.47 × 10−2	6.42 × 10−12	0.99%
Abiotic depletion	/	/	/	6.46 × 10−10	100.00%

).

As shown in Table 9, the synergistic coefficients of NO_x, TSP, and SO₂ emission reduction are 1.89, 0.0041, and 7.40 tCO₂/ton of pollutants, respectively; in other words, every ton of NO_x, TSP, and SO₂ emission reduction leads to 1.89, 0.0041, and 7.40 tons of carbon dioxide emissions. The synergistic coefficients of SO₂ emission reduction are much higher than those of the other three pollutants, while the synergistic coefficients of TSP emission reduction are the lowest. Further analysis shows that the RASOC renewable wet flue gas desulfurization technology used in the combined sulfur recovery unit consumes more power and has a higher power consumption factor, resulting in greater indirect emissions of CO₂.

As listed in Table 10, the synergistic coefficients for VOC abatement in SK Ethylene’s end-of-pipe treatment approach are considerably higher than those for NO_x and particulate matter. Conversely, the synergistic coefficients for particulate matter abatement are the lowest. In the abatement of VOCs, the direct and indirect CO₂ emissions generated by the flare thermal incineration method are higher, amounting to 2.4 times the level observed for the AOGI thermal incineration method. The technologies employed for end-of-pipe treatment in the power center and the sulfuric acid recovery sector include ammonia flue gas desulfurization (FGD) (the simultaneous removal of SO₂ and particulate matter), selective catalytic reduction (SCR) (the removal of NO_x), and organocatalytic flue gas desulfurization and denitrification (the simultaneous removal of SO₂ and NO_x), with synergistic effects. The synergistic coefficient of ammonia FGD is notably higher, reaching 53.1 tCO₂/t, 1.2 tCO₂/t, and 4.2 tCO₂/t. Evidently, the synergistic coefficients of ammonia desulfurization and dust removal are the highest, followed by organic catalytic flue gas desulfurization and denitrification—a result that may be due to the great amounts of power that these two methods consume, resulting in high indirect CO₂ emissions.

As indicated in Table 11, the pollutants associated with the chlor-alkali process are relatively unique, with a high electricity consumption coefficient for end-of-pipe treatment, resulting in increased electricity usage during the pollutant reduction process. The synergistic coefficients for the reductions in the emissions of Cl₂, HCl, and VOCs are 26.90, 425.96, and 70.67 tCO₂ per ton of pollutant, respectively, implying that, for every ton of Cl₂, HCl, and VOCs removed, there is an increase of 26.9, 426.0, and 71.0 tons in carbon dioxide emissions, respectively.

Table 12. Cross-elasticity coefficients for pollutant reductions for end-of-pipe air pollutant management technologies at SH Petrochemical, SK Ethylene, and HS Chlor-Alkali.

Enterprise	Pollutant	Emission Reduction Measures/Technologies	Air Pollutant Emission Reduction (t)	Direct Synergistic CO2 Emissions (t)	Indirect Synergistic CO2 Emissions (t)	Cross-Elasticity Coefficient
SH Petrochemical	SO2	RASOC renewable wet flue gas	587.8	/	4352.0	−0.001
	NOx	SCR	539.2	/	1016.6	−0.0004
SK Ethylene	NOx	SCR	776.5	0	956.8	−0.001
	Particulate matter	Bag filter	1902.2	0	7.70	−7.5 × 10−6
	VOCs	AOGI thermal incineration	9179.3	28,817.8	5715.5	−0.03
		Flare thermal incineration	22,941.0	209,024.3	1981.6	−0.2
	SO2, TSP	Ammonia desulfurization and dedusting	68.5	0	3640.6	−0.007
	SO2, NOx	Organic catalytic flue gas desulfurization and denitrification	530.7	0	2258.0	−0.003
HS Chlor-Alkali	Cl2	Alkali absorption	10.7	/	287.8	−0.0004
	HCl	Falling film absorption + water jet	2.0	/	851.5	−0.001
	VOCs	High-temperature incineration	9.1	139.0	504.8	−0.002

presents the cross-elasticity coefficients for atmospheric pollutants and the associated end-of-pipe control technologies for all three enterprises. All coefficients exhibit negative values, confirming that, while end-of-pipe treatments reduce the atmospheric pollutant levels, they concurrently increase the CO₂ emissions. Critically, the lower absolute value of the cross-elasticity coefficient indicates a weaker synergistic effect, indicating that GHG emissions increase at a slower rate per unit reduction in atmospheric pollutants.

Regarding SH Petrochemical, the RASOC wet flue gas desulfurization method led to a 97.81% reduction in SO₂ emissions (587.79 tons), while also generating 4351.97 tons of CO₂, with a growth rate (the proportion of CO₂ emissions related to end-of-pipe treatment with respect to the total CO₂ emissions of the company) of 0.1%. This method had the largest absolute value of the cross-elasticity coefficient (0.00101) among the three emission reduction approaches used by SH Petrochemical, indicating that this method results in a relatively high growth rate for carbon dioxide emissions. Regarding SK Ethylene, flare thermal incineration removed 22,900 tons of VOCs (with a reduction rate of 98%), while directly and indirectly emitting 211,000 tons of CO₂, with an increase rate of 20.47%. Its cross-elasticity coefficient’s absolute value was 0.209, the largest among the emission reduction approaches employed by SK Ethylene. In contrast, the bag filter approach yielded the smallest absolute value of the cross-elasticity coefficient, amounting to only 7.51 × 10⁻⁶. For HS Chlor-Alkali, reducing 10.7 tons of Cl₂ (with a reduction rate of 99.53%) resulted in only 287.83 tons of CO₂ emissions, with an increase rate of 0.04%—the smallest absolute value of the cross-elasticity coefficient in this project—indicating that the Cl₂ end-of-pipe reduction technology leads to a relatively small rate of growth in CO₂ emissions. For this company, reducing 9.11 tons of VOCs (with a reduction rate of 5.19%) resulted in 643.82 tons of CO₂ emissions, with an increase rate of 0.09% and an absolute value of the cross-elasticity coefficient of 0.0171, which is relatively large, indicating that the VOC end-of-pipe reduction technology has a relatively significant impact on CO₂ emissions.

3.3. Life Cycle Environmental Impact Assessment for Typical Products

The CML-IA model in the SimaPro software was used to characterize and standardize the life cycle inventory data. The environmental impact loads of the 11 indicators were cumulated to derive the life cycle environmental impact assessment results corresponding to each ton of product produced by SH Petrochemical (refining sector), SK Ethylene, and HS Chlor-Alkali. The characterization and standardization results obtained are shown in Appendix E.

Normalization eliminated the dimensional disparities between the environmental impact indicators, allowing for a comparison between the different indicators [37]. The calculations indicated that the environmental impact loads per ton of product for SH Petrochemical (refining sector), SK Ethylene, and HS Chlor-Alkali were 5.44 × 10⁻⁹, 7.05 × 10⁻⁹, and 6.46 × 10⁻¹⁰, respectively. SK Ethylene exhibited the highest environmental impact per ton of ethylene produced, whereas HS Chlor-Alkali exhibited the lowest impact per ton of chlor-alkali produced. Figure 5 delineates the proportions of the environmental impact loads for each indicator per unit product produced by these three enterprises. The figure reveals that the predominant environmental impacts for the production of one ton of product for these chemical enterprises are ADP fossil, MAETP, GWP, and AP, exhibiting a convergent trend. Notably, the impact on ADP fossil is the most pronounced, constituting 69.48%, 53.94%, and 34.23% of each enterprise’s total environmental impact load, respectively. The extensive utilization of fossil raw materials and fossil energy consumption during production and upstream electricity generation significantly contributes to the elevated proportion of the environmental impact load attributed to ADP fossil. The environmental impacts of the global warming potential and acidification potential predominantly originate from the emissions of carbon dioxide, sulfur dioxide, and nitrogen oxides during production. The marine aquatic ecotoxicity impact primarily results from the discharge of wastewater containing toxic substances during production.

3.4. Synergistic Effects Based on Impacts on Ecological Environment Under Different Future Scenarios

Table 13. Proportion of change in life cycle environmental impact load between low-carbon development scenario and baseline scenario.

Sector	Scenario Comparison	Type of Indicator	Value
Refining industry	Low-carbon development scenario vs. baseline scenario	ΔEIAP	−5.5%
		ΔEIGHG	−17%
Ethylene industry	Low-carbon development scenario vs. baseline scenario	ΔEIAP	−2.1%
		ΔEIGHG	−8.8%
Chlor-alkali industry	Low-carbon development scenario vs. baseline scenario	ΔEIAP	−0.29%
		ΔEIGHG	−15.0%

presents the incremental ratios of the life cycle environmental impacts of producing one ton of refined diesel in the oil-refining industry, one ton of ethylene in the ethylene industry, and one ton of chlor-alkali in the chlor-alkali industry under the low-carbon development scenario, compared with the environmental impacts in the baseline scenario. The results reveal 5.5% and 2.1% reductions in the environmental impact loads of air pollutants in the oil-refining and ethylene industries, respectively, suggesting that optimizing the upstream electric power structure and utilizing “green hydrogen” and analogous measures effectively reduce their environmental impact loads. In contrast, the upstream power structure adjustment and process-related energy-saving measures exhibit a comparatively diminished effect in terms of reducing air pollutants in the chlor-alkali industry, with a mere 0.29% decrease in the environmental impact load of air pollutants associated with producing one ton of product. This decline can be attributed to the nature of air pollutants in the chlor-alkali industry, primarily consisting of specific pollutants such as chlorine gas and hydrogen chloride, making them less directly associated with energy consumption.

Furthermore, the GHG-related environmental impact loads of the three industries are 17%, 8.8%, and 15% lower in the low-carbon development scenario, with relatively substantial emission reductions, suggesting that optimizing the upstream electric power structure and process improvement can reduce GHG emissions. This reduction in environmental impact is particularly pronounced in the oil-refining and chlor-alkali sectors, which substantially impact GHGs. The findings underscore the efficacy of strategies such as optimizing the upstream electric power structure and implementing process improvements to achieve substantial GHG emission reductions and significantly mitigate the environmental impacts of the production of these goods.

Table 14. Cross-elasticity coefficients based on life cycle environmental impacts for low-carbon development scenario compared to baseline scenario.

Sector	Scenario	SEA/C
Refining industry	Low-carbon development scenario	0.32
Ethylene industry	Low-carbon development scenario	0.24
Chlor-alkali industry	Low-carbon development scenario	0.019

presents the cross-elasticity coefficients (based on life cycle environmental impacts) for the low-carbon development scenario relative to the baseline scenario. In the low-carbon development scenario, the SE_A/C of the oil-refining, ethylene, and chlor-alkali industries is greater than zero, with all values being less than 1. This result indicates that the environmental impacts of reducing GHG and air pollutant emissions have a positive synergistic effect and exhibit positive synergies. The SE_A/C values for the oil-refining, ethylene, and chlor-alkali industries are 0.32, 0.24, and 0.019, respectively. It is evident that the rate of change in the environmental impact loads associated with GHG emission reductions is greater than that of atmospheric pollutant emission reductions, suggesting that, in the low-carbon development scenario, the synergistic effect of policy measures targeting carbon peaking on air pollutant emissions is relatively weak. The oil-refining industry has the most considerable SE_A/C value, indicating that air pollutant and GHG abatement have relatively strong synergistic effects. By contrast, the chlor-alkali industry has the smallest SE_A/C value and the weakest synergistic effect. Likely because of the distinct production processes and raw materials used in this sector, the chlor-alkali industry emits primary air pollutants such as chlorine and hydrogen chloride, which bear less relevance to energy consumption.

4. Discussion

(1) Comparison of emission accounting methods and analysis of data discrepancies in chemical industry: Accounting for localized air pollutant and GHG emissions is foundational for quantitative research and synergistic emission reduction practices. In the chemical industry, the emission of major pollutants can be accounted for via field measurement, material balance, and analogical methods. Field measurement involves the on-site monitoring of pollutant emissions, offering high accuracy and enabling the analysis of the emission characteristics of the source. However, this method is limited by its expensiveness and instrument dependence [38]. The material balance method, based on the law of conservation of matter, is highly accurate and inexpensive. However, it requires precise knowledge of material usage and alterations in the production process—information that is often challenging to obtain, particularly in complex production processes [39]. The analogical method is a technique used to indirectly calculate the emissions of air pollutants generated by a target enterprise. This method has a wide range of applications but is less accurate. The quality of the raw materials used by different enterprises, differences in the production process, and the operational status will affect the accounting results [39]. The field measurement and material balance methods were adopted in this study. We compared the accounting results obtained in this study with the Second National Census on Pollution Sources (2017) (SNCPS) database and the Environmental Statistics Database. According to the Second Census, the NO_x emissions of SH Petrochemical (in the refining sector) amounted to 414.6 tons, while the figure obtained in this study was slightly lower. This discrepancy can be attributed, at least in part, to our use of 2021 data, as China further strengthened its efforts to reduce emissions from the petrochemical and other heavily polluting industries between 2017 and 2021, resulting in a decrease in the annual emissions of pollutants from all refineries in the country. This phenomenon has resulted in a consistent decline in the annual emissions of pollutants from all refineries across the country. According to the 2017 annual environmental statistics, the NO_x, particulate matter, and SO₂ emissions generated by SK Ethylene boilers amounted to 391.0 tons, 7.8 tons, and 37.5 tons, respectively. The emissions of NO_x in the power center unit found in this study amounted to 348.9 tons, which is slightly lower than in the environmental statistics, while the emissions of particulate matter and SO₂ amounted to 12.8 tons and 53.0 tons, slightly higher than in the environmental statistics. This is possibly due to the fact that, with the increase in ethylene production, the power center’s increased load resulted in elevated particulate matter and SO₂ emissions, concurrently augmenting nitrogen oxide generation. However, in 2017, this ethylene company appended a novel SCR denitrification apparatus to all three boilers within the power center, culminating in a substantial reduction in emissions.

The main reason behind the Chinese chemical industry’s high contribution to GHG emissions is its use of the carbon balance and emission factor methods. The carbon balance method requires comprehensive and precise data, including detailed material inputs and outputs, energy use data, and carbon content parameters that align with an enterprise’s actual circumstances [26]. The accuracy of the emission factor method is contingent on the emission factors employed. Therefore, the utilization of emission factors by enterprises in China should be prioritized when accounting for these entities [40]. A comparison of the environmental statistics from 2021 revealed that the CO₂ emissions of SH Petrochemical’s refining segment amounted to 5.2 million tons per year. This figure exceeds the result reported in the present study. The potential causes of this discrepancy may include the process of CO₂ emission, which is associated with the hydrocracking process, which, in turn, may have been underestimated due to the absence of relevant data. This potential oversight could have led to inaccurate calculations of the process emissions, leading to a lower result.

The synergistic coefficient of air pollutant emission reduction for chlor-alkali projects is much higher than that of oil-refining and ethylene production enterprises. Evidently, the synergistic coefficients of different manufacturers and the end-of-pipe technologies for different pollutants can be diverse. Thus, the synergistic coefficients of end-of-pipe technologies are affected by a variety of factors, such as the products, pollutant types, and treatment technologies. According to the “dual-carbon” goal, when selecting end-of-pipe technologies, in addition to considering their emission reduction effects on specific pollutants, it is also necessary to comprehensively consider their effects on GHG emissions, their technical feasibility, and the synergistic benefits, among other factors. The research in this paper can help the chemical industry and enterprises to select decarbonized pollution treatment technologies and realize simultaneous air pollution and carbon reduction. In the foreseeable future, the end-of-pipe treatment of localized air pollutants will remain an important means of reducing pollutant emissions and improving the air quality. Therefore, conducting a systematic study on the synergistic coefficients of end-of-pipe treatment technologies for various pollutants in different industries and enterprises, and establishing a corresponding database, will provide data support and a basis for the selection of decarbonized end-of-pipe treatment technologies in different industries and enterprises.

(2) Quantification and implications of synergistic effects of pollution and carbon reduction in end-of-pipe treatment technologies for chemical enterprises: The synergistic effects of localized air pollutant and GHG emission reduction are evident in the changes in emissions and their ecological impacts. Quantitative research on synergistic effects, based on ecological and environmental impacts, can assist industries and enterprises in comprehensively considering their production characteristics and environmental impacts, thereby helping them to formulate targeted pollution and carbon reduction strategies. In this study, we incorporated the life cycle approach into the quantitative analysis of synergies, combined it with simulations of low-carbon development scenarios, and quantitatively evaluated the synergistic effects of emission reduction in terms of environmental impacts. The study of synergistic effects can be expanded from the accounting of emissions to the assessment of environmental impacts, thus providing a technical reference for the development of relevant policies and the assessment of implementation effects. Our findings indicate that the oil-refining industry, through its use of crude oil as a primary raw material, is a significant consumer of fuel oil and electricity. China’s thermal-power-based energy structure is characterized by substantial fossil fuel consumption, contributing to the depletion of fossil resources and ensuring that it has the greatest environmental impact as a proportion of the total load. The oil-refining industry must prioritize optimizing its energy structures, enhancing its energy efficiency, and achieving energy conservation and emission reduction. Secondly, the three enterprises examined in this study also exhibited higher values of MAETP loads. Notably, SK Ethylene demonstrated the highest impact load value for this indicator, underscoring the need for this enterprise to address potential impacts on the marine ecosystem that may arise from its production processes. The ADP element and GWP of HS Chlor-Alkali accounted for the largest proportions of their total environmental impact loads, a fact that may be attributed to the substantial consumption of salts and abiotic resources in the chlor-alkali production process. Moreover, this finding underscores the need for HS Chlor-Alkali to implement continuous improvements in its utilization of abiotic resources and regulation of carbon dioxide emissions. An examination of the synergistic effects, grounded in the life cycle environmental impacts in future scenarios, reveals that the oil-refining industry has the highest cross-elasticity coefficient, indicating strong synergies between air pollutant and GHG emission reduction. In contrast, the chlor-alkali industry exhibits the lowest value and the weakest synergistic effect. In the future, this research is anticipated to be extended to various sectors, including diverse chemical manufacturing enterprises. The findings will serve as a reference with which government management departments can formulate more targeted and synergistic emission reduction policies across different industries.

(3) Absence of real-time monitoring data: The absence of real-time monitoring data constitutes a limitation of this study. Because of the availability of real-time CEMS data and its focus on major pollution sources and discharge outlets, it lacks comprehensive coverage. Furthermore, issues such as sensor aging and data loss can compromise data reproducibility. While real-time CEMS data remain limited in our study context (particularly for VOC fugitive sources), we maximized the data robustness through cross-validation methods (e.g., material balance calculations). We believe that, with broader deployment and technological refinements of real-time CEMS, such monitoring data will see wider application in the future.

(4) Use of static electricity emission factors: At present, real-time CEMS data mainly focus on the major pollution sources and major discharge outlets, and they do not cover all of them. Moreover, there are problems such as sensor aging and data loss, which will affect the reproducibility of the data as well. While real-time CEMS data remain limited in our study context (particularly for VOC fugitive sources), we maximized the data robustness through cross-validation (e.g., using material balance calculations). However, the embodied CO₂ emissions from the net purchased electricity were calculated using the emission factor approach. As a fundamental tool in current greenhouse gas accounting systems, this methodology is designated as the standard method in the IPCC Guidelines and China’s Provincial Greenhouse Gas Inventory Compilation Guidelines. The emission factors, uniformly published annually by the state (Study on CO₂ Emission Factors for China’s Regional Power Grids), prevent reporting deviations arising from enterprise-selected factors, thus enhancing the comparability of emission data across enterprises and industries. Furthermore, the emission factors applied in the calculations corresponded to activity-level data, thereby improving the reliability of the results. However, the annual average power emission factor adopted lacks the dynamic nature of spatiotemporal differentiation. Therefore, in the future, research should fully consider spatiotemporally refined accounting—for instance, a hierarchical factor system can be established.

(5) Promotion of research results: In the study of the co-benefits between atmospheric pollutant and greenhouse gas emission reductions, the accuracy of the results is highly dependent on the precision of the accounted emissions. Future research will include the accounting of emissions data from various enterprises in the chemical industry over multiple years. By comparatively analyzing the synergy coefficients and cross-elasticity coefficients, we aim to derive more generalizable patterns from these data.

5. Conclusions

The emission inventory analysis indicates that refining and ethylene production predominantly emit SO₂, NO_x, particulate matter, and VOCs. In contrast, chlor-alkali production primarily discharges specific pollutants such as Cl₂, HCl, and VOCs. Regarding CO₂ emissions, the principal contributors in refining processes are emissions generated by fuel combustion and those associated with net purchased electricity and heat. In ethylene production, the key emission sources include fuel combustion, flare combustion, and emissions linked to net purchased electricity and heat. For chlor-alkali production, the emission sources encompass fuel combustion, process emissions, direct carbon emissions from end-of-pipe treatments, and implied emissions from net purchased electricity and heat.

The results of our synergy analysis indicate significant variability in the synergistic coefficients for emission reduction in all three enterprises and across different end-of-pipe technologies. The synergistic coefficients for one air pollutant treatment vary significantly across chemical industries and production sectors. Overall, for each ton of air pollutant reduced, SH Petrochemical, SK Ethylene, and HS Chlor-Alkali increase their CO₂ emissions by 3.8, 7.1, and 81.8 tons, respectively.

The life cycle environmental impact assessment results indicate that the environmental impact load values per ton product for SH Petrochemical (oil-refining sector), SK Ethylene Project, and HS Chlor-Alkali are 5.4 × 10⁻⁹, 7.0 × 10⁻⁹, and 6.5 × 10⁻¹⁰, respectively. SK Ethylene exhibited the highest environmental impact per unit product, while HS Chlor-Alkali showed the lowest. The primary environmental impacts associated with the production processes of these three companies include ADP fossil, GWP, MAETP, and AP, with some convergence in these areas. Additionally, ADP fossil presented the highest environmental impact loading ratio.

The scenario analysis results reveal that implementing low-carbon development policies can substantially mitigate the environmental impacts of the chemical industry, particularly concerning AP, GWP, and MAETP. Under a low-carbon development scenario, the total environmental impact load values for SH Petrochemical (in the refining sector), SK Ethylene, and HS Chlor-Alkali decrease by 3.2%, 3.9%, and 4.9%, respectively, compared to the baseline scenario. Consequently, optimizing the upstream power structure, substituting “green hydrogen” for “grey hydrogen” in refinery hydrogenation units, and reducing energy consumption, such as that of electricity and steam, in ethylene and chlor-alkali production processes are effective strategies for the achievement of low-carbon development in the chemical industry. Under the scenario of low-carbon development, the environmental impact loads resulting from atmospheric pollutant and greenhouse gas emissions can be synergistically reduced for all three enterprises mentioned above. In particular, the refinery enterprise demonstrates strong synergistic effects, whereas the chlor-alkali enterprise exhibits relatively weaker synergistic effects.