Soil polycyclic aromatic hydrocarbons in energy chemical industrial plants of Northwest China: Sources and health risk implications from a field‐based study

Abstract

Polycyclic aromatic hydrocarbons (PAHs) in soils of energy chemical industrial plants pose considerable ecological and health risks. However, comprehensive studies between petrochemical and coal chemical plants in arid Northwest China remain limited. In this study, 75 surface soil samples (0–20 cm) were collected from 15 large‐scale plants (four petrochemical and 11 coal chemical plants) across the region. Sixteen priority PAHs (∑PAHs) were quantified, and source apportionment was conducted using diagnostic ratios and principal component analysis‐multiple linear regression (PCA‐MLR). In addition, health risks were evaluated using toxic equivalence factors and incremental lifetime cancer risk (ILCR) models. Results showed a stark contrast in PAHs concentrations, with coal chemical plants averaging 4652.2 ng/g and reaching up to 188,506.0 ng/g at a coal tar processing site, compared to 225.6 ng/g in petrochemical plants. Four‐ring PAHs were predominant, accounting for 64.5% and 52.1% of total PAHs in coal chemical and petrochemical soils, respectively, indicative of high‐temperature combustion sources. PCA‐MLR analysis identified coal combustion as the dominant source at coal chemical sites (96.5%), while petrochemical sites exhibited mixed sources, including oil leakage (49.6%) and vehicular emissions (25.6%). Health risk assessment revealed that ILCR values exceeded the US EPA threshold of 10⁻⁶ at seven sampling points in coal chemical plants, with the highest risk observed at site BXT‐2 (1.99 × 10⁻⁴), primarily driven by dermal and oral exposure to benzo[a]pyrene. These findings highlight the urgent need for stricter control of high‐temperature industrial processes and improved management of hazardous waste in energy chemical industrial plants.

Core Ideas

• Coal chemical soil concentrations were significantly higher than in petrochemical sites, with extreme contamination observed near coal tar processing units.

• The predominance of four‐ring PAHs in soils points to high‐temperature combustion as a key contributing source.

• Coal combustion was the dominant contributor in coal chemical plants, while petrochemical plants were influenced by mixed sources

• Cancer risks exceeded safe limits at several coal chemical plants due to benzo[a]pyrene (BaP) exposure.

Plain Language Summary

Polycyclic aromatic hydrocarbons (PAHs) are toxic pollutants posing potential health risks near industrial facilities. This study investigated PAHs contamination in 75 soil samples collected from four petrochemical and 11 coal chemical plants in arid Northwest China. Results showed that PAHs levels in coal chemical plants were markedly higher, with extreme levels near coal tar processing sites. In both settings, PAHs profiles were dominated by high‐temperature combustion‐related PAHs. Source apportionment indicated coal combustion as the main contributor in coal chemical plants, while petrochemical plants showed mixed sources, including oil leakage and vehicular emissions. Health risk assessment showed that several coal chemical sites exceeded acceptable cancer risk thresholds, mainly due to benzo[a]pyrene exposure, highlighting the need for strengthened pollution control and soil management.

Abbreviations

Ace

acenaphthene

Acy

acenaphthylene

Ant

anthracene

BaA

benzo[a]anthracene

BaP

benzo[a]pyrene

BbF

benzo[b]fluoranthene

BghiP

benzo[g,h,i]perylene

BkF

benzo[k]fluoranthene

Chr

chrysene

DBA

dibenzo[a, h]anthracene

DRs

diagnostic Ratios

FlA

fluoranthene

Flu

fluorene

HMW

high‐molecular weight

ILCR

incremental lifetime cancer risk

InP

indeno[1,2,3‐cd]pyrene

LMW

low‐molecular weight

MDLs

method detection limits

MLR

multiple linear regression

MMW

medium‐molecular weight

Nap

naphthalene

PAHs

polycyclic aromatic hydrocarbons

PCA

principal component analysis

PCs

principal components

Phe

phenanthrene

Pyr

pyrene

TEFs

toxic equivalence factors

TEQ

toxic equivalency quantity

1. INTRODUCTION

Polycyclic aromatic hydrocarbons (PAHs) are persistent organic pollutants composed of two or more fused aromatic rings. They are mainly released from biomass combustion, fossil fuel pyrolysis, and vehicular emissions (Y. Li, Chen, et al., ²⁰²⁵; Liao et al., ²⁰²⁵; Z. Liu, Mu, et al., ²⁰²⁴). As a result, PAHs are widely detected in the environment, including soil, air, and aquatic systems (Bayer et al., 2025; Gao et al., ²⁰²⁵; W. Liu et al., ²⁰²⁵; H. Wang, Shu, et al., ²⁰²⁵). Their hydrophobicity promotes accumulation in soils, making terrestrial systems an important sink for industrial‐derived PAHs (Meierdierks et al., 2022). Humans are exposed to PAHs through multiple pathways, including ingestion of contaminated food, inhalation of particle‐bound compounds, and dermal contact with polluted soil and water (Rabieimesbah et al., 2024). Once absorbed, PAHs can persist and accumulate in the body, increasing the risk of adverse health effects (Deng et al., 2024; H. Jia et al., ²⁰¹⁷; Mallah et al., ²⁰²²; H. Wang, Shu, et al., ²⁰²⁵). Given the complexity of PAHs exposure pathways and the heterogeneity of industrial contamination, advanced risk assessment approaches are essential for quantitatively evaluating potential human health risks. Recent global meta‐analyses have systematically evaluated PAHs health risks through dietary pathways, particularly in high‐temperature cooked foods such as kebabs and fish (Gholami‐Borujeni et al., 2024, 2025; Mortezazadeh et al., ²⁰²³). These studies indicate that while dietary PAHs exposure generally remains below acceptable risk thresholds, the cumulative burden from multiple exposure routes, including occupational soil contact, warrants integrated assessment.

Coking and petroleum refining are among the most PAHs‐intensive industrial sectors in China. Studies have shown varying levels of PAHs contamination across industrial regions, with coal chemical and petrochemical sites often exhibiting the highest concentrations (You et al., 2024). For example, soil near a coal chemical facility in Chongqing showed ∑₁₆PAHs levels ranging from 0.3 to 6910 ng/g, while a petroleum refinery site in Shandong reported concentrations up to 2710 ng/g (Ai et al., 2024; N. Wang, Jiang, et al., ²⁰²⁵). Northwest China, a major base for energy and chemical industries, is characterized by an arid climate with low precipitation and high evaporation rates. The pollution profile in this region differs markedly from that of fine chemical enterprises in eastern coastal areas (Dai et al., 2024). Such conditions can influence the environmental fate of pollutants, particularly by enhancing soil dryness and dust generation, which may in turn affect the transport and human exposure pathways of PAHs (J. Wang, Yu, et al., ²⁰²⁵). However, previous studies have mainly focused on individual industrial sectors, urban soils, or specific contamination characteristics, while comparative investigations between coal chemical and petrochemical sites remain limited. In particular, few studies have systematically evaluated the combined effects of industrial activities, arid climatic conditions, source apportionment, and multi‐pathway health risks of PAHs in Northwest China.

Coal chemical and petrochemical processes, including pyrolysis, coking, and catalytic refining, are important anthropogenic sources of PAHs. For instance, in Shanxi Province, coke oven emissions were identified as the primary source of high‐molecular weight (HMW) PAHs (Ai et al., 2024), while in a large petrochemical complex in Yanshan, petroleum‐derived PAHs accounted for 41.6% of the total burden (Z. Liu, Mu, et al., ²⁰²⁴). PAHs in industrial environments are often characterized by high concentrations, mobility, ecotoxicity, and complex mixtures. Occupational exposure through inhalation and skin contact increases the risk of cancer and respiratory conditions among workers (Dalvand et al., 2024; B. Wu, Lin, et al., ²⁰²³). Understanding the environmental behavior of PAHs and their occupational exposure pathways is essential for advancing cleaner industrial practices and minimizing human and ecological risks.

In this study, 75 surface soil samples (0‐ to 20‐cm depth) were collected from 15 large‐scale chemical enterprises in Northwest China, including four petrochemical and 11 coal chemical plants. The analysis focused on 16 priority PAHs with three main objectives: (1) to compare PAHs contamination between petrochemical and coal chemical plants; (2) to characterize the occurrence, source profiles, and health risks of PAHs in arid industrial areas of Northwest China; and (3) to examine the relationships among industrial functional zones, PAHs compositional patterns, and localized health‐risk hotspots. This study provides a comparative understanding of PAHs contamination characteristics in petrochemical and coal chemical plants under arid environmental conditions. The findings offer new insights into industry‐specific pollution patterns and support targeted risk management and remediation strategies for contaminated industrial sites in Northwest China.

2. MATERIALS AND METHODS

2.1. Materials and reagents

A total of 16 priority PAHs were analyzed (Table S1), including naphthalene (NaP), acenaphthene (Ace), acenaphthylene (Acy), fluorene (Flu), phenanthrene (Phe), anthracene (Ant), fluoranthene (FlA), pyrene (Pyr), benzo[a]anthracene (BaA), chrysene (Chr), benzo[b]fluoranthene (BbF), benzo[k]fluoranthene (BkF), benzo[a]pyrene (BaP), dibenzo[a, h]anthracene (DBA), Benzo[g,h,i]perylene (BghiP), indeno[1,2,3‐cd]pyrene (InP). Decafluorobiphenyl was used as a surrogate standard, while naphthalene‐D8, acenaphthene‐D10, phenanthrene‐D10, chrysene‐D12, and perylene‐D12 were used as internal standards. All PAHs standard solutions (purity > 98%) were obtained from Sichuan Zhongbiao Standardization Technology Co., Ltd. HPLC‐grade acetonitrile, n‐hexane, dichloromethane, and acetone were purchased from Shanghai Anpu Laboratory Technology Co., Ltd.

Core Ideas

• Coal chemical soil concentrations were significantly higher than in petrochemical sites, with extreme contamination observed near coal tar processing units.

• The predominance of four‐ring PAHs in soils points to high‐temperature combustion as a key contributing source.

• Coal combustion was the dominant contributor in coal chemical plants, while petrochemical plants were influenced by mixed sources

• Cancer risks exceeded safe limits at several coal chemical plants due to benzo[a]pyrene (BaP) exposure.

2.2. Sample collection

In June 2024, a total of 75 surface soil samples (0‐ to 20‐cm depth) were collected from 15 large chemical enterprises in Northwest China, including four petrochemical and 11 coal chemical plants. This composition reflected the actual distribution of major chemical facility types in the study region. Sampling locations (three to eight per plant) were selected based on three criteria: (1) operational scale, (2) process complexity, and (3) terrain accessibility. At each site, samples were obtained using the five‐point composite sampling method (J. Wu, Yang, et al., ²⁰²³). Collected samples were immediately sealed in aluminum foil bags and stored at 4°C prior to analysis. Detailed information on the sampling sites is provided in Table S2 and Figure S1.

2.3. PAHs analysis

Pretreatment of soil samples was conducted according to the Chinese national standard HJ805‐2016 (Ministry of Environmental Protection). The procedure involved removing impurities, followed by moisture removal using a freeze dryer. Dried samples were ground and passed through an 80‐mesh sieve. Approximately 40 µg/mL decafluorobiphenyl was added as a surrogate standard before transferring the sample into a Soxhlet extraction thimble. Extraction was performed in a Soxhlet apparatus using 100 mL of a dichloromethane‐acetone mixture for 16–18 h, with a reflux rate maintained at 4–6 cycles per hour. After extraction, the solvent was concentrated to about 2 mL at 40°C. The extract was transferred to a rotary evaporation flask, and the extraction vessel was rinsed twice with a small volume of the solvent mixture; these rinses were combined with the main extract and further concentrated to approximately 1 mL under nitrogen blowing. Quantification was performed using a Trace 1300 gas chromatography‐mass spectrometry system (Thermo Fisher Scientific) equipped with an HP‐5MS capillary column (30 m × 0.25 mm × 0.25 µm) and operated in selected ion monitoring mode. Detailed extraction, cleanup, and instrumental conditions are provided in Text S1 and Figure S2.

2.4. Quality control and quality assurance

During sample pretreatment, quartz sand blanks and quartz sand samples spiked with standard solutions were included every 10 samples as procedural blanks and spiked blanks to monitor potential contamination and analytical recovery. Decafluorobiphenyl was used as a surrogate standard to evaluate recovery throughout the extraction and pretreatment procedures. Quantification of PAHs was conducted using the internal standard method. Before extraction, internal standards, including naphthalene‐D8, acenaphthene‐D10, phenanthrene‐D10, chrysene‐D12, and perylene‐D12, were added to all samples. Calibration curves were established using at least five concentration levels, and satisfactory linearity was obtained for all target compounds (r ² > 0.99). No target PAHs were detected in procedural blanks above the method detection limits (MDLs). Recovery rates ranged from 59.9% to 102.0%, while relative standard deviations for replicate analyses were generally below 10%, indicating acceptable analytical precision and accuracy. Concentrations below the MDLs were treated as zero in subsequent statistical analyses. Detailed quality assurance/quality control information was provided in Table S3.

2.5. Source apportionment

2.5.1. Diagnostic ratios

Four diagnostic ratios (DRs) were employed: Ant/(Ant + Phe), FlA/(FlA + Pyr), InP/(InP + BghiP), and BaA/(BaA + Chr) (J. Li, Xue, et al., 2025; Suman et al., ²⁰¹⁶). The threshold values for source identification based on these ratios are summarized in Table 1.

TABLE 1.

Threshold criteria of diagnostic ratio.

Source	Criteria
Petroleum sources	Ant/(Ant + Phe) < 0.1, FlA/(FlA + Pyr) < 0.4, BaA/(BaA + Chr) < 0.2
Petroleum combustion sources	Ant/(Ant + Phe) > 0.1
Combustion sources	0.4 < FlA/(FlA + Pyr) < 0.6, InP/(InP + BghiP) < 0.4, 0.2 < BaA/(BaA + Chr) < 0.35
Wood, coal, and biomass combustion sources	FlA/(FlA + Pyr) > 0.6, InP/(InP + BghiP) > 0.4, BaA/(BaA + Chr) > 0.35

Abbreviations: Ant, anthracene; BaA, benzo[a]anthracene; BghiP, benzo[g,h,i]perylene; Chr, chrysene; FlA, fluoranthene; InP, indeno[1,2,3‐cd]pyrene; Phe, phenanthrene; Pyr, pyrene.

2.5.2. Principal component analysis‐multiple linear regression

To reduce uncertainties associated with single‐method source apportionment, principal component analysis combined with multiple linear regression (PCA‐MLR) was employed for quantitative source apportionment. PCA‐MLR was a receptor modeling approach that integrates dimensionality reduction with linear regression to identify independent principal components (PCs) representing distinct pollution sources while preserving the maximum variance of the original dataset (Zhou et al., 2024). Prior to analysis, PAHs concentration data were standardized using Z‐score normalization to ensure equal weighting across variables. PCs with eigenvalues greater than 1 were retained and subjected to varimax rotation to improve interpretability. The factor loading matrix (L) and score matrix (X) were derived according to the following equation:

$$Y=X\times L$$ (1)

where Y was the normalized PAHs concentration matrix, X was the factor score matrix, and L was the factor loading matrix.

Subsequently, the MLR method was used to quantify the identified source contributions of PAHs in the sample with the PCA factor scores (X) and normalized total PAHs concentrations as the independent and dependent variables, respectively. The basic equation of a multiple linear model is as follows.

Subsequently, MLR was applied to quantify the contributions of the identified sources to total PAHs. In this model, the PC scores (X) obtained from PCA served as the independent variables, while the normalized total PAHs concentrations were used as the dependent variable. The basic form of the MLR model is expressed as:

$$Z=\sum B_i\times X_i$$ (2)

where Z was the standardized normal deviates of ∑PAHs, X_i was the factor score matrix, and B_i was the MLR regression coefficient.

The contribution of source k can be obtained by the following equation:

$$k=B_k/\sum B_i$$ (3)

where B_k represents the MLR regression coefficient of source k.

2.6. Risk assessment

Each PAHs exhibited a different degree of toxicity. Generally, low‐molecular weight (LMW) PAHs are associated with acute toxicity, while HMW PAHs tend to be carcinogenic. To facilitate a unified assessment of their toxic potential, toxic equivalence factors (TEFs) were applied. The total toxicity of a PAH mixture was quantified using the following formula:

$$\mathrm{TE}{\mathrm{Q}}_{\mathrm{BaPi}}=\sum _{i=1}^{16}(C_i\times \mathrm{TE}{\mathrm{F}}_i)$$ (4)

where C_i was the concentration of PAHs i, and TEF _i was the corresponding toxic equivalency factor of PAHs i related to BaP. The TEF value for individual PAHs were obtained from previously published studies (He et al., 2020). More detailed information was shown in Table S4.

Incremental lifetime cancer risk (ILCR) represented the probability of an individual developing cancer over a lifetime as a result of long‐term exposure to carcinogenic pollutants (Shukla et al., 2022). We used ILCR to evaluate the potential health risks associated with PAHs contamination at both petrochemical and coal chemical sites, following the methodology recommended by the US EPA (1992) and adapted in recent studies (H. Wang, Shu, et al., ²⁰²⁵).

PAHs in soil may pose carcinogenic risks primarily through three exposure pathways: dermal contact, inhalation, and oral ingestion (Luo et al., 2025). Accordingly, the total ILCR was calculated using the following equation (Y. Wu et al., ²⁰²⁵):

$$\mathrm{ILCR}=\frac{\mathrm{CS}\times \mathrm{IR}\times \mathrm{EF}\times \mathrm{ED}\times \mathrm{SF}}{\mathrm{BW}\times \mathrm{AT}}\times 10^{-6}$$ (5)

where CS represented the toxic equivalent concentration of the 16 individual PAHs in soil (mg·kg⁻¹), EF was the exposure frequency (day·year⁻¹), SF was the cancer slope factor ((mg·kg⁻¹·day⁻¹)⁻¹), BW was the body weight (kg), ED was the exposure duration (years), and IR is the intake rate (g·person⁻¹·day⁻¹). Following the Exposure Factors Handbook of Chinese Population (Duan, 2013), AT represented the average exposure time (days). These standard exposure parameters for the Chinese population were adopted to ensure the context‐specific relevance of the health risk assessment. Detailed parameter values used in the calculations are summarized in Table S5.

3. RESULTS AND DISCUSSION

3.1. Occurrence of PAHs in two types of chemical sites

In petrochemical plants, Nap and Chr exhibited the highest detection frequencies, both at 72% (Figure 1; Table S6). The detection frequencies of the remaining 14 PAHs were considerably lower, with BbF at 44%, BaA at 40%, Ace at 36%, BkF at 32%, Pyr at 28%, and Phe at 24%. All others were detected in less than 20% of the samples. In contrast, coal chemical plants showed generally higher detection rates, with FlA, Nap, and Pyr detected in 92%, 86%, and 86% of samples, respectively (Figure 1; Table S6). Only DBA and Ace had detection frequencies below 20%. Notably, Nap and Chr were detected in more than 70% of samples across both types of sites, highlighting their widespread presence.

FIGURE 1.

Occurrence of polycyclic aromatic hydrocarbons (PAHs) in petrochemical and coal chemical plants. (a) Ranking of detection frequency of PAHs. (b) Mean concentrations of detected PAHs. (c) Average concentrations of PAHs on two different plants. (d) Spatial distribution of PAHs with different ring numbers.

As shown in Figure 1, the total concentrations of the 16 PAHs (∑PAHs) in soils from petrochemical plants ranged from 10.1 to 1276.2 ng/g, with an average of 225.6 ng/g. Chr recorded the highest mean concentration (68.69 ng/g), followed by Nap (22.46 ng/g), aligning with their high detection frequencies. These concentration levels are comparable to those reported in previous studies. For example, Tommasi et al. (2023) observed ∑PAHs ranging from 0 to 1118 ng/g (mean: 217.1 ng/g) in soils near Italian refineries, while H. Wang et al. (2024) reported concentrations from 60.2 to 1678.0 ng/g (mean: 278.9 ng/g) in refinery‐adjacent soils in China.

In coal chemical sites, ∑PAHs exhibited a wide range, from below detection limits to 188,506.1 ng/g, with a mean concentration of 4652.2 ng/g (Figure 1). The dominant PAHs were all four‐ring compounds, including FlA (931.9 ng/g), Chr (864.7 ng/g), and Pyr (676.1 ng/g). Compared with other coking‐related studies, the mean ∑PAHs concentration observed in our study was approximately twice that reported for two coking plants in Shanxi Province (1957.8 ng/g), yet lower than the levels recorded at a facility in Shaanxi (8160.0 ng/g) (X. Liu, Tan, et al., ²⁰²⁴). It was also substantially lower than those reported at two heavily contaminated coking sites in Shanxi, which reached 12,600 ng/g and 28,984 ng/g, respectively (Du et al., 2022; S. Zhang, Li, et al., ²⁰²³). These differences may be attributed to variations in production processes, types of chemical products, and on‐site hazardous waste management practices.

Among the ∑PAHs, 13 exhibited higher average concentrations in soils from coal chemical sites compared to petrochemical sites. The only exceptions—Ace, Flu, and DBA—were detected only at trace levels (Figure 1c). Notably, the contrast in PAHs composition between the two industry types was especially evident for medium‐molecular weight (MMW) and HMW PAHs (MMW‐ and HMW‐PAHs, rings number ≥4), which were up to 10 times more concentrated in coal chemical soils. This pattern likely reflects the dominant high‐temperature, incomplete combustion processes in coal chemical production, such as coking and coal tar distillation, which are major sources of MMW‐ and HMW‐PAHs (Qu et al., 2020).

According to the Maliszewska‐Kordybach's soil contamination classification, 76% of sampling points at petrochemical plants fell below 200 ng/g, indicating no contamination. Only two sites (WSH‐2 and WSH‐4) were mildly contaminated (200–600 ng/g), three (DSH‐5, WSH‐6, and TSH‐5) were moderately contaminated (600–1000 ng/g), and one site (TSH‐2) exceeded 1000 ng/g, indicating heavy contamination. In contrast, coal chemical sites showed a lower proportion of uncontaminated samples (62%), with nine sites classified as mildly contaminated and 10 exceeding 1000 ng/g, indicating heavy contamination. Overall, these results suggested more severe PAHs contamination at coal chemical sites than at petrochemical sites. Notably, although most sampling points were classified as “uncontaminated” under this scheme, PAHs concentrations exhibited a highly skewed distribution. A small number of hotspots showed exceptionally high levels, which disproportionately increased the mean concentrations and dominated the overall health risk.

Figure 1d presented the distribution of PAHs by ring number. In soils from the four petrochemical sites, four‐ring and three‐ring PAHs were predominant, comprising 52.1% and 21.4% of ∑PAHs, respectively. The average concentration of four‐ring PAHs was 116.7 ng/g, while three‐ring PAHs were present at lower levels (Figure 1d). The persistence of three‐ring PAHs was likely due to their slow biodegradation rates, as reported by Esfandiar et al. (2021), and their frequent occurrence in contaminated industrial soils has been widely documented. Four‐ring PAHs, categorized as MMW PAHs, tend to accumulate in soils due to their low volatility and strong hydrophobicity, which facilitates adsorption onto soil particles (Jesus et al., 2022). In contrast, soils from the 11 coal chemical sites showed an even greater dominance of four‐ring PAHs, accounting for 64.5% of ∑PAHs, followed by five‐ring PAHs at 18.4%. These HMW PAHs were typically associated with high‐temperature industrial activities such as coal combustion and coke production (T. Jia et al., ²⁰²¹).

Among all sites, BXT‐2 exhibited the highest average ∑PAHs concentration (69,638.7 ng/g), with four‐ring PAHs contributing 45,931.3 ng/g. At this site, all PAHs except two‐ring compounds exceeded 4000 ng/g. In particular, BXT‐2 recorded an extreme ∑PAHs concentration of 188,506.1 ng/g, far exceeding the 1000 ng/g threshold for severe soil contamination. According to the verified site information, BXT‐2 is located in the hazardous waste temporary storage area of a coal chemical site. Therefore, the pronounced PAHs hotspot at this location was more reasonably associated with site‐specific hazardous waste storage and handling activities, together with the broader influence of coal chemical production processes at the BXT facility. Given that the BXT site was involved in the production of asphalt, Ant oil, and tar, the elevated PAHs burden at BXT‐2 may reflect the accumulation, transfer, or leakage of PAHs‐rich residues generated during coal chemical operations, as well as potential spillage and seepage during hazardous waste storage. Similar PAHs hotspots have been reported in soils surrounding coal chemical and coking‐related industrial areas, indicating that waste management units within such facilities can act as important localized sinks of PAH contamination (X. Zhang et al., ²⁰²⁵).

3.2. Source apportionment of PAHs in petrochemical and coal chemical plants

3.2.1. Diagnostic ratio method

Previous studies had demonstrated that PAHs with varying ring numbers can serve as indicators of different emission sources. DR method was widely used to distinguish these sources based on the emission characteristics of PAHs. LMW PAHs (two to three rings number) generally associated with petrogenic sources such as petroleum spills or natural seepage. In contrast, HMW (≥4 rings number) were commonly associated with high‐temperature pyrolytic processes, including the incomplete combustion of coal, biomass, or other organic matter (H. Wang et al., ²⁰²⁴). The ratio of LMW to HMW PAHs offered a preliminary means of inferring dominant sources across the 15 industrial plants (Table 2). A LMW/HMW ratio greater than 1 generally indicates a petrogenic origin, while a ratio less than 1 suggests a pyrogenic (combustion‐related) source (Chen et al., 2022). At the petrochemical sites, LMW/HMW ratios ranged from 0.12 to 2.60. DSH and KSH exhibited ratios above 1, suggesting the possible influence of petroleum‐derived PAHs, whereas the remaining sites appeared to be influenced mainly by combustion‐related inputs. Similarly, at the coal chemical sites, the ratios ranged from 0.10 to 2.34 (Table 2). Only GH and XTM had ratios above 1, suggesting some contribution from petroleum, while the majority of sites showed combustion as the primary PAHs source.

TABLE 2.

Spatial distribution of polycyclic aromatic hydrocarbon (PAH) ring‐based diagnostic ratios.

Site	Average LMW/HMW	Average HMW/∑PAHs	Types
WSH	0.1200	0.8929	Petrochemical
DSH	1.0161	0.4960
KSH	2.5990	0.2779
TSH	0.5610	0.6406
GN	0.7133	0.5837	Coal chemical
GK	0.4473	0.6909
ZN	0.7872	0.5595
XLX	0.3444	0.7439
BXT	0.0963	0.9122
ZT	0.7654	0.5664
YH	0.8046	0.5542
GH	1.6240	0.3811
JJ	0.4595	0.6852
XTM	2.3441	0.2990
QH	0.4280	0.7003

The ratio of HMW/∑PAHs also served as a valuable indicator for differentiating PAH sources. Values greater than 0.5 typically reflect petroleum combustion, while ratios below 0.4 are indicative of petrogenic sources such as oil spills or geological inputs (Yan et al., 2005). As shown in Table 2, the petrochemical site KSH had an HMW/∑PAHs ratio below 0.4, suggesting a petrogenic origin, whereas the remaining sites (WSH, DSH, and TSH) showed ratios above 0.5, consistent with combustion‐related sources. A similar trend was observed in the coal chemical sites: two locations (GH and XTM) had ratios below 0.4, while the other nine exceeded 0.5. Overall, four out of the 15 sampling sites (DSH, KSH, GH, and XTM) showed evidence of petrogenic or geological sources of PAHs, whereas the remaining 11 were primarily influenced by high‐temperature combustion processes (Table 2). These findings further support the conclusion that combustion‐related activities, particularly coal‐related processes, likely represent an important contributor to PAHs pollution in most of the studied sites.

Due to thermodynamic stability differences among PAHs isomers, source attribution can be refined using DRs in addition to molecular weight classifications (H. Zhang et al., ²⁰⁰⁶). Figure 2a showed the distribution of selected DRs across all sampling points. Notably, 46.7% of the samples exhibited FlA/(FlA + Pyr) ratios between 0.4 and 0.6. When focusing specifically on coal chemical sites, this proportion increased to 60%, suggesting a stronger contribution from combustion‐related sources. Further analysis revealed that most samples from coal chemical plants had Ant/(Ant + Phe) ratios above 0.1, consistent with pyrogenic origins, while samples from petrochemical sites typically showed ratios below 0.1 or were undetermined due to low concentrations (Figure 2a). These findings suggested that PAHs in petrochemical sites were likely influenced by petroleum spills or leakage during processing and transport, whereas PAHs in coal chemical sites are likely derived from high‐temperature combustion of coal or petroleum. The result was also supported by the BaA/(BaA + Chr) ratio patterns (Figure 2b). In petrochemical plants, most sites fell below 0.4, primarily between 0.2 and 0.3, indicating petroleum‐related sources. In contrast, coal chemical plants exhibited BaA/(BaA + Chr) ratios predominantly above 0.35, pointing to combustion of coal or biomass. Collectively, the DRs indicated that PAHs in petrochemical plants mainly originated from petroleum‐related inputs and combustion processes, whereas PAHs in coal chemical plants were more strongly associated with combustion sources.

FIGURE 2.

Polycyclic aromatic hydrocarbon (PAH) isomeric ratios in soils from two type chemical plants. (a) Ant/(Ant + Phe) to FlA/(FlA + Pyr). (b) BaA/(BaA + Chr) to InP/(InP + BghiP). Triangles indicate petrochemical plants, whereas squares indicate coal chemical plants.

Overall, DR analysis of PAHs concentrations indicated that PAHs in the four petrochemical areas originate from mixed sources. These were primarily associated with petroleum leaks and combustion processes, with possible contributions from natural geological background levels. In contrast, PAHs detected in the 11 coal chemical sites are predominantly attributed to the combustion of coal, petroleum, or biomass. These findings were consistent with previous studies (Y. Zhang, Niu, et al., ²⁰²³). Both petrochemical and coal chemical facilities typically involved complex processes with a wide range of chemical feedstocks and deliberate high‐temperature operations, which often resulted in unorganized and irregular emissions. Additionally, leaks of various oils and raw materials were commonly observed, further contributing to the release of PAHs into the surrounding environment. However, it should be noted that DRs may be influenced by environmental weathering processes such as volatilization, degradation, and sorption, which can alter the original PAHs compositional signatures. Therefore, the identified sources should be interpreted as indicative patterns rather than definitive source assignments.

3.2.2. Principal component analysis‐multiple linear regression

To address the limitations of the DRs in source identification, PCA‐MLR was employed to identify the major sources of PAHs in both petrochemical and coal chemical plants (Tables S7 and S8). At the petrochemical sites, five PCs were identified, explaining 88.6% of the total variance in PAHs composition (Table S7). PC1 (25.6%), dominated by LMW PAHs (e.g., Phe, Ant, BaA, Chr, and NaP), was characteristic of uncombusted petroleum inputs and was therefore likely associated with oil leakage (Ribeiro et al., 2012). PC2 (21.2%), dominated by HMW PAHs such as BaP, DBA, BghiP, and InP, was indicative of high‐temperature combustion processes and was interpreted as reflecting petroleum combustion‐related emissions (Kwon & Choi, 2014). PC3 (18.2%), characterized by relatively high loadings of FlA, BbF, and BkF, was commonly associated with vehicular emissions, particularly diesel exhaust, and was likely linked to traffic‐related emissions (Liao et al., 2025). PC4 (18.2%), primarily represented by Pyr, may be linked to mixed combustion sources, including biomass and fossil fuel burning, as Pyr was frequently reported as a general combustion marker (Cao et al., 2020). PC5 (8.5%), marked by Ace, may reflect emissions from industrial processes such as coke production, although this attribution is subject to uncertainty due to the limited number of indicator compounds (Shi et al., 2021).

At the coal chemical sites, five PCs explained 95.9% of the total variance in PAHs composition (Table S8). Among them, PC1 was the dominant component, accounting for 68.4% of the variance. It was characterized by both LMW PAHs, such as Acy and Phe, and several HMW PAHs, suggesting a strong influence of coal combustion‐related activities on PAHs accumulation at these sites. The remaining components (PC2–PC5) contributed comparatively less to the overall variance and were associated with different emission sources. PC2 was linked to biomass burning, PC3 reflected fossil material leakage, PC4 was associated with coke oven operations, and PC5 was attributed to vehicular emissions.

The quantitative source apportionment results derived from MLR analysis further suggested differences between petrochemical and coal chemical plants (Figure 3). At petrochemical plants, oil leakage was the predominant source, contributing 49.6% of total PAHs, followed by traffic emissions (25.6%) and petroleum combustion (13.1%). Four major sources each contributed over 15%, indicating a diverse range of contamination pathways. In contrast, coal combustion–related processes were estimated to contribute substantially to PAHs contamination (96.5%), with other sources contributing only marginally. These findings highlight fundamental differences in source profiles, with petrochemical facilities exhibiting greater variability and coal chemical sites being dominated by high‐temperature combustion processes.

FIGURE 3.

Factor loading of different polycyclic aromatic hydrocarbons (PAHs) in petrochemical area (a) and coal chemical area (b).

To further investigate spatial variability, we compared source contributions with PAHs concentration hotspots. At petrochemical sites, elevated PAHs concentrations were primarily found near storage tanks, transportation pipelines, and equipment maintenance plants (Table S2). These areas closely corresponded with the dominant sources of oil leakage and traffic emissions, suggesting that localized leaks and frequent vehicular activity significantly impacted soil contamination. Similarly, plants adjacent to flaring units and combustion chambers showed higher levels of HMW PAHs, aligning with petroleum combustion as a key source and reinforcing the role of on‐site thermal processes (Table S2; Figure 1). At coal chemical sites, the spatial distribution of PAHs was more uniform, but concentrations were notably higher around coking units and coal tar distillation facilities. This pattern was broadly consistent with the influence of combustion‐related activities inferred from PCA analysis (PC1, 96.5%; Table S8), and suggested that emissions from centralized, high‐temperature units result in broader, more diffuse contamination of surrounding soils. Minor contributions from diesel emissions and fossil material leakage were detected near transportation corridors and chemical transfer pipelines, though their impact was limited. These spatial and compositional patterns indicated that the nature of industrial processes and site layouts may influence PAHs contamination profiles. The greater source diversity observed at petrochemical plants reflected the complexity of their operations, which typically involve a combination of open and closed systems, dynamic material handling, and frequent vehicular movement (X. Wang et al., ²⁰¹³). In contrast, coal chemical facilities were characterized by centralized thermal operations, resulting in a more uniform but combustion‐heavy pollution signature (S. Li et al., ²⁰²⁴).

Comparison with prior studies further validates these findings. For instance, Ren et al. (2022) reported that in petrochemical industrial clusters in Northwest China, significant PAH pollution originates from the leakage of petroleum products during oil production processes. Cai et al. (2018), whose study on the emission characteristics of thermally generated PAHs further confirmed the role of pyrolysis‐related processes as key contributors as well. Similarly, X. Zhang et al. (2025) identified localized PAH hotspots around tar and bitumen production tanks. These concentrations were primarily driven by emissions from processing leaks and transportation activities. These external observations corroborate our PCA‐MLR findings and underscore the influence of operational activities on PAH distribution. In coal‐intensive regions such as Inner Mongolia (Xu et al., 2024) and Shanxi Province (Z. Liu, Mu, et al., ²⁰²⁴), PAHs contamination was likewise dominated by HMW species from coal combustion, with contribution rates exceeding 90%. The presence of coking and tar‐processing units in those studies mirrors the layout of our coal chemical sites, further supporting the possible link between process characteristics and pollutant profiles.

In summary, our results emphasize the value of integrating chemical source profiles with site‐specific operational data to improve source apportionment and inform risk assessment. While DR methods offer a basic screening tool, advanced approaches such as PCA‐MLR can provide useful insights into spatial heterogeneity and source complexity. The general agreement between inferred PAHs sources and site operational characteristics offers practical guidance for remediation, including leak control, traffic pathway redesign, and the optimization of combustion systems to mitigate emissions at their source.

3.3. Risk assessment of PAHs

3.3.1. Health risk assessment using toxic equivalence factors

Toxic equivalence factor (TEF) method was applied to evaluate the carcinogenic potential of 16 priority PAHs (Agarwal et al., 2009; Santodonato et al., ¹⁹⁸¹). Detailed calculated toxic equivalent concentrations were summarized in Figure 4a and Table S9. In the petrochemical plants, the toxic equivalent quantities (TEQs) of the 16 priority PAHs ranged from 0.01 to 209.83 ng/g, with an average of 18.14 ng/g. Notably, seven carcinogenic PAHs (BaP, BaA, BbF, BkF, InP, Chr, and DBA) contributed an average of 17.97 ng/g, accounting for approximately 99% of the total TEQs. Among them, BaP had the highest average TEQ at 7.45 ng/g, representing 41.1% of the total, followed by DBA at 5.94 ng/g, contributing 32.8%. In the coal chemical plants, the TEQs of the 16 PAHs ranged from 0 to as high as 14,864.99 ng/g, with an average total TEQs of 378.99 ng/g, approximately 20 times higher than that observed in the petrochemical plants. As with the petrochemical plants, the seven carcinogenic PAHs accounted for nearly the entire TEQ (374.65 ng/g). BaP remained the predominant contributor, representing 63.0% of the total TEQs, followed by BaA at 12.9%.

FIGURE 4.

(a) TEQ_BaP of 16 polycyclic aromatic hydrocarbons (PAHs). Different colors represent PAHs with different ring numbers. (b) Spatial distribution of incremental lifetime cancer risk (ILCR), where the color gradient from green to red represents increasing risk levels. BaP, benzo[a]pyrene; TEQ, toxic equivalence quantity.

These results consistently indicated that BaP was the primary contributor to the overall carcinogenic toxicity in both petrochemical and coal chemical plants. This dominance was largely attributed to BaP's high toxic equivalency factor (TEF = 1.0) and its relatively elevated concentrations in contaminated soils (Downs et al., 2024). BaP was a well‐known genotoxic carcinogen produced during the incomplete combustion of fossil fuels, and was closely associated with high‐temperature industrial activities such as oil refining, coal coking, and hazardous waste incineration(Yu et al., 2023). The potential ecological and human health risks associated with BaP were further evaluated using the Canadian Soil Quality Guidelines for the Protection of Environmental and Human Health, which set a threshold of 600 ng/g for BaP toxic equivalent quantity (TEQ_BaP) in soil. At the petrochemical plants, TEQ_BaP values at all 25 sampling locations were below this threshold, indicating a low level of carcinogenic risk. In contrast, two sampling points at the coal chemical plants exceeded the guideline, both located at site BXT, specifically within the hazardous waste storage and production plants. These areas also exhibited unusually high PAHs concentrations, suggesting localized contamination likely caused by historical waste accumulation, leakage, or accidental spillage during industrial operations. In summary, the dominance of BaP in both plants underscored its critical role in driving the overall toxicity of PAHs‐contaminated soils. Its sources were closely linked to high‐temperature industrial operations, and its persistence and strong carcinogenicity necessitate targeted monitoring and remediation efforts, particularly in hotspots like the coal chemical site BXT.

3.3.2. Cancer risk assessment

In addition to using TEF‐based assessment, the ILCR model was applied to further evaluate the potential carcinogenic risks of PAHs exposure at both petrochemical and coal chemical sites (Z. Zhang et al., ²⁰²²). Because all sampling locations were situated within industrial park boundaries, the assessment focused on adult workers potentially exposed to PAHs through dermal contact, oral ingestion, and inhalation pathways. As shown in Figure 4b, the calculated ILCRs ranged from 3.41 × 10⁻¹¹ to 1.27 × 10⁻⁴ for dermal exposure, 1.29 × 10⁻¹¹ to 7.16 × 10⁻⁵ for ingestion, and 1.27 × 10⁻¹⁹ to 3.41 × 10⁻¹⁵ for inhalation exposure. According to the US EPA risk guidelines (U.S. Environmental Protection Agency, ¹⁹⁸⁹; Moya et al., ²⁰¹¹), ILCR values exceeding 1 × 10⁻⁶ indicate potential carcinogenic concern. While inhalation posed negligible risks across all sites, several sampling locations exceeded the acceptable threshold through dermal and ingestion pathways, indicating localized health risks associated with soil exposure.

At the petrochemical plants, total ILCR values ranged from 1.35 × 10⁻¹⁰ to 2.81 × 10^−6, with an average value of 2.43 × 10⁻⁷, generally remaining within the acceptable risk range. However, two sampling points (WSH‐6 and DSH‐5) exceeded the US EPA acceptable cancer risk threshold (1 × 10⁻⁶). WSH‐6 was located adjacent to a 50,000‐L crude oil storage tank, while DSH‐5 was situated within the waste treatment area of the petrochemical facility. The elevated risk at these locations were likely associated with oil leakage, surface runoff, and the handling of petroleum‐derived by‐products.

In contrast, coal chemical plants exhibited substantially higher carcinogenic risks, with ILCR values ranging from 5.34 × 10⁻¹¹ to 1.99 × 10⁻⁴. The average ILCR reached 5.07 × 10⁻⁶, whereas the median value was much lower (5.44 × 10⁻⁸), indicating a highly skewed distribution dominated by several contamination hotspots. Seven sampling locations exceeded the acceptable threshold, and site BXT‐2 surpassed 1 × 10⁻⁴, indicating a high carcinogenic risk. Most high risk points were concentrated near production units and hazardous waste storage areas, highlighting the importance of localized industrial activities in driving soil PAHs exposure.

Pathway‐specific analysis revealed that dermal contact and oral ingestion were the dominant routes of exposure. This aligned with the physicochemical properties of soil‐bound PAHs, which were characterized by low volatility and strong particle affinity (J. Liu et al., ²⁰²²). The arid climate of the Northwest China likely amplifies these risks, as frequent wind erosion and low soil moisture can increase dust generation and surface deposition, enhancing the chances of skin contact and unintentional ingestion (Lian et al., 2025). These conditions were especially problematic in areas lacking effective dust suppression or adequate personal protective equipment (Aquilina & Harrison, 2023). Further correlation with source apportionment results reinforced these findings. High risk sites generally overlapped with plants exhibiting strong contributions from pyrogenic sources such as petroleum combustion and industrial discharge. For example, BXT‐2, the site with the highest ILCR, was dominated by combustion‐derived PAHs, particularly BaP and DBA, both of which have high TEFs and therefore contributed substantially to the overall risk level. This spatial convergence between pollution sources and elevated ILCR highlighted the urgent need for targeted risk mitigation near production and waste management areas. In summary, PAH‐related health risks in both petrochemical and coal chemical plants were primarily driven by dermal and ingestion exposure pathways. High risk plants were closely linked to localized pollution from industrial operations. Moreover, the region's arid climate may exacerbate exposure risks, emphasizing the importance of site‐specific mitigation strategies in such environments.

Overall, the health risks associated with PAHs in industrial soils were primarily driven by dermal contact and incidental ingestion pathways rather than inhalation exposure. High risk sites were closely associated with localized pollution from industrial activities. In addition, the region's arid climate may further increase exposure, highlighting the need for site‐specific mitigation strategies. To place these findings in context, the estimated risks were compared with those reported for other exposure pathways, particularly dietary intake, which is widely recognized as a major source of human PAHs exposure. Previous meta‐analyses of PAHs in high‐temperature cooked foods (e.g., meat and fish) indicate that, although certain compounds such as BaP may pose noncarcinogenic risks at high intake levels, the associated lifetime cancer risks are generally below the acceptable threshold (1×10⁻⁶) (Gholami‐Borujeni et al., 2024, 2025).

In contrast, the ILCR values observed in this study (up to 1.99 × 10⁻⁴) substantially exceed the dietary PAH risk estimates reported in recent meta‐analyses, where lifetime cancer risks from meat kebab consumption were consistently below 1 × 10⁻⁶ (Gholami‐Borujeni et al., 2025; Mortezazadeh et al., ²⁰²³). This disparity underscores that occupational soil exposure in industrial settings may pose a greater carcinogenic threat than dietary intake in contaminated regions, challenging the conventional assumption that diet is the dominant PAHs exposure pathway.

4. CONCLUSION AND FURTHER OUTLOOK

This study provides a comparative evaluation of PAHs contamination characteristics, source profiles, and associated health risks in petrochemical and coal chemical industrial plants located in the arid region of Northwest China. The results demonstrated that coal chemical plants exhibited substantially higher PAHs accumulation and carcinogenic risk than petrochemical plants, largely due to intensive high‐temperature industrial processes such as coal combustion, coking, and tar distillation. In contrast, petrochemical plants were characterized by more complex mixed sources, including oil leakage and traffic‐related emissions. By integrating DRs, PCA‐MLR source apportionment, TEQ analysis, and ILCR assessment, this study revealed strong links between industrial operational characteristics, PAHs composition patterns, and human exposure risks. The findings further suggest that arid environmental conditions, including low soil moisture and enhanced dust generation, may increase dermal and ingestion‐related exposure risks in industrial areas. The results provided practical guidance for pollution control strategies, including emission reduction from high‐temperature industrial processes, hazardous waste management optimization, and targeted remediation of contamination hotspots in coal chemical parks.

Despite these findings, several limitations should be noted. First, the health risk assessment relied on standard exposure parameters, which may not fully reflect site‐specific exposure conditions or differences among exposed populations. Second, the analysis focused only on the 16 priority PAHs, whereas other emerging organic pollutants may also contribute to the overall environmental and health risks. Third, the PCA‐based source apportionment contains inherent uncertainty and may not completely distinguish mixed emission sources, particularly in weathered industrial soils where volatilization, degradation, and sorption processes may alter PAHs compositional signatures. In addition, the sampling design has several limitations that may introduce uncertainty into sector‐level comparisons. The number of investigated petrochemical and coal chemical plants was not fully balanced, and all samples were collected during a single sampling campaign, which may not fully capture temporal variability in PAHs distribution. Furthermore, background/control soils and detailed soil physicochemical properties were not systematically included in the present study. Therefore, the observed differences between petrochemical and coal chemical plants should be interpreted as representative patterns of the investigated sites rather than universal characteristics of all industrial facilities.

Future studies should incorporate site‐specific exposure scenarios and include a wider range of contaminants to provide a more comprehensive risk evaluation. In addition, advanced source apportionment approaches, such as isotopic analysis and high‐resolution receptor models, should be applied to improve source identification and resolution. Integrating multi‐pathway exposure assessments, including dietary intake, would also enable a more comprehensive evaluation of PAHs‐related health risks.

AUTHOR CONTRIBUTIONS

Mengcheng Wu: Methodology; writing—original draft. YunZhou Tan: Writing—original draft. Guodong Kang: Resources. Ping Wu: Project administration; resources. Mengjiao Li: Visualization. Liangliang Liu: Methodology. Ning Wang: Formal analysis. Ling Wang: Methodology. Shenghu Zhang: Formal analysis; supervision. Leilei Lu: Funding acquisition; writing—review and editing.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Supporting information

Supplementary Material

Wu, M. , Tan, Y. Z. , Kang, G. , Wu, P. , Li, M. , Liu, L. , Wang, N. , Wang, L. , Zhang, S. , & Lu, L. (2026). Soil polycyclic aromatic hydrocarbons in energy chemical industrial plants of Northwest China: Sources and health risk implications from a field‐based study. Journal of Environmental Quality, 55, e70258. 10.1002/jeq2.70258

DATA AVAILABILITY STATEMENT

Data will be made available on request.