2.1 Description of typical industries
Soil sand mining, ferrous metal casting, ship repair, equipment repair, and gasoline station were selected as typical industries for the following reasons. According to the “Management catalogue of occupational hazard risk classification of construction projects” issued by the State Administration of Work Safety of China (2012 edition),24 the inherent risk (IR) for occupational hazards in the soil sand mining and ferrous metal casting industries was classified as “severe.” The IR for the ship repair and equipment repair industries was classified as “medium”, while that for the gasoline station industry was classified as “low.” Thus, these five industries represent a range of IR levels in China (severe, medium, and low IR). Among the five industries, IR for occupational hazards decreases in the following order: IR mining and IR ferrous > IR ship and IR equipment > IR gasoline.
A total of 151 enterprises in Zhejiang Province in eastern China were selected as typical enterprises. These included three large enterprises, eight medium-sized enterprises, 29 small enterprises and 111 micro-enterprises.25 A total of approximately 16,000 workers exposed to hazard factors were involved. Basic information is shown in Table 1.
2.2 Identification and detection of hazard factors
The hazard factors and levels of exposure were identified through occupational health field investigations, air sampling, and laboratory testing. Air sampling and laboratory testing were carried out in accordance with the Chinese standard “Specifications of air sampling for hazardous substances monitoring in the workplace (GBZ 159)” and “Determination of toxic substances in workplace air (GBZ/T160 and 300).” Table 1 shows the basic information and levels of exposure to hazard factors (e.g., silicon dust, welding dust, manganese and inorganic compounds, grinding wheel dust, xylene, and iron ore powder) in each industry. The exposure levels of hazard factors at some locations in the soil sand mining, ferrous metal casting, and ship repair industries exceeded the permissible concentration-time weighted average (PC-TWA) permitted by China or the threshold limit values-time weighted average (TLV-TWA) permitted by ACGIH. This was not the case for the equipment repair and gasoline station industries.
2.3 Introduction to the six commonly used OHRA methods22
The six common OHRA methods (EPA, COSHH, Singaporean, ICMM, Australian, and Romanian) have similar assessment frameworks. The main assessment framework is based on the degree of hazard, exposure level, and probability of occurrence and includes hazard identification, hazard characteristic assessment, exposure assessment, and risk description. The detailed principles of the six methods have been reported previously5-10 and are briefly described below.
- EPA method (quantitative evaluation). The EPA inhalation risk assessment consists of two parts: carcinogenic and non-carcinogenic risk assessments. The non-carcinogenic risk assessment was mainly applied in this study and involves two primary steps:
A) Estimating exposure concentration (EC, in μg/m3):
EC = (CA × ET × EF × ED) / AT (1)
where CA (μg/m3) is the concentration of hazard factor in the air; ET (h/d) is the exposure time; EF (days/year) is the exposure frequency; ED (years) is the exposure duration; AT [ED (years) × 365 days/year × 24 h/day] is the average exposure time.
B) Non-carcinogenic risk assessment:
The hazard quotient (HQ), which indicates the risk level, is defined as
HQ = EC/RfC × 1000 (μg/mg) (2)
where RfC (mg/m3) is the reference concentration of inhalation toxicity.
The EPA model can calculate the occupational health risk level of chemical toxicants with relative accuracy, but can only assess the health risk caused by inhalation route, and is limited to chemical toxicants with reference concentration (RfC) and inhalation unit risk (IUR), which can only be retrieved from the EPA website poison database.
(2) COSHH model for qualitative evaluation. In this method, the health hazard levels and exposure levels of chemical substances (solid or liquid) are considered comprehensively, and the control level is provided by a matrix method. The health hazard level of a chemical is determined according to a hazard band using risk phrases or OELs. The exposure level is determined according to the dustiness of a solid or the volatility of a liquid and the scale of use. While this method is simple and feasible, it may not always be accurate because it does not consider protection measures or on-site toxicant concentrations.
(3)Singaporean method (semi-quantitative evaluation). The risk level is calculated according to the hazard ratings (HR) and exposure ratings (ER), and the formula is as follows:
Risk = (HR×ER)1/2 (3)
The HR is determined based on carcinogenicity classifications from the ACGIH and the
International Agency for Research on Cancer, or on the acute toxicity data of chemicals (LD50 and LC50). The ER is classified according to the ratio of field exposure concentration to occupational exposure limits.
(4) ICMM method (qualitative evaluation). This method comprehensively considers the possible health hazards, probability of exposure, and exposure time. The risk level is determined using a quantitative method or matrix method.
(5) Australian method (qualitative evaluation). In this method, the risk levels are determined manually using a diagram or a calculator based on the likelihood of occurrence, frequency of exposure, and severity of consequences. This method is simple and easy to apply and is suitable for a wide range of assessments (e.g., risk assessments carried out by occupational health management personnel in small- and medium-sized enterprises.26
(6) Romanian method (qualitative assessment). In this method, the risk level is evaluated using a matrix method based on the severity and probability of consequences resulting from hazard factors. This method can be used to calculate the overall risk level of the workplace and has obvious advantages in comprehensive risk assessment.
2.4 Quantitative analysis model
2.4.1 Risk ratio (RR)22
The six OHRA methods produced different levels of risk. To compare the results of each method, the risk levels obtained using the six methods were converted into RRs for quantitative comparison.
(1) Conversion of risk level: The EPA method produces quantitative data. The output of the COSHH method is control method classification. The risk assessment results of the other four methods are classifications of risk level. Thus, to compare the assessment results among different methods, the EPA non-carcinogenic risk assessment results (HQ) were converted into risk level by referring to the classification standard of exposure concentration of the Singaporean method, which includes five levels. The results of the COSHH method were converted by referring to the risk level of the Singaporean method (Table 2).
(2) RR calculation: After risk level conversion, the results for the six methods were converted to the classification of risk level. The risk assessment results of the EPA, Australian, Singaporean, and ICMM models were divided into five levels, while those of the Romanian and COSHH models were divided into seven and four levels, respectively. The concept of RR was introduced to allow comparison among the risk assessment results of different methods. RR was defined as the ratio of the risk level of an occupational hazard factor assessed by a method to the highest risk level of the model. The RR represents the relative risk level of hazard factors derived from a certain method.
2.4.2 Concentration ratio (CR)22
To make the exposure concentration of hazard factors of different positions comparable, CR was defined as the ratio of the exposure concentration of a hazard factor to the OEL of the hazard factor. CR represents the relative exposure level of a certain hazard factor in a certain position; thus, CR can be used to compare the exposure levels of different hazard factors or different positions. CR > 1 indicates that the exposure to a hazard factor exceeds the OEL for that factor.
2.4.3 Quantitative analysis
2.4.3.1 Comparison of RRs among the six OHRA methods
The statistical differences among the RRs evaluated by the six OHRA methods reflect the differences among the evaluation results of the OHRA methods for the same occupational hazard factors.
2.4.3.2 Correlations among the RRs of the six OHRA methods
The correlations among the RRs obtained by the six methods were statistically analyzed.
2.4.3.3 Accuracy verification based on the RRs of the six OHRA methods
(1) The accuracy of the OHRA results obtained using the six OHRA methods in different industries was verified by comparing the consistency in RR values for different industries and inherent risks (IR) levels. Refer to Section 2.1, Description of typical industries, for the classification of inherent risks in each industry.
(2) The accuracy of the evaluation result of each method was verified by evaluating the consistency in the RRs for different chemical toxicants and IRs. We selected four chemical toxicants (manganese and inorganic compounds, benzene, xylene, and ethyl acetate) to evaluate the accuracy of each method. The IR of a hazard factor depends on its inherent hazardous consequences and exposure probability. The IR increases as the inherent hazardous consequences become more severe and as the exposure concentration increases. In this study, the inherent hazardous consequences of a hazard factor were determined based on the RfC value of the EPA method. A larger RfC indicates less severe inherent hazard consequences. Table 4 shows the RfC values and exposure concentrations for each hazard factor. The IR values of the four hazard factors in the five industries decreases in the following order: IR manganese > IR benzene ≈ IR xylene > IR ethyl acetate.
2.5 Statistical analysis
The Kruskala–Wallis H(K) method was used to analyze the RRs and CRs of multiple independent samples. The Mann–Whitney U method was used to compare the RR or CR between two independent samples. The correlations between RR values were analyzed by Spearman correlation analysis (abnormal distribution).