Evaluation of Homogeneous Exposure Grouping to Estimate Individual Workers' Exposure in South African Gold Mines, 2019–2022

Abstract

Background: The South African Mining Industry (SAMI) operates under the guidance of the Mine Health and Safety Act, which obligates employers to ensure the health and safety of mine workers. This legislation outlines the need for mining operations to develop and implement a mandatory code of practice (MCoP) to monitor airborne contaminant exposure, assess occupational health risks, and apply suitable control measures. Due to practical and resource constraints, conducting personal exposure monitoring for every worker is not feasible. Therefore, employees are organized into Homogeneous Exposure Groups (HEGs), where it is assumed that individuals performing similar duties in comparable environments are exposed to similar levels of airborne hazards. Exposure estimates for individuals are then based on the average exposure level of the group. However, this grouping method may not account for variations in specific job functions, task-related hazards, behavioral differences, or inconsistent use of protective equipment. Such factors can result in inaccurate exposure estimations, which may compromise the effectiveness of control strategies designed to reduce the risk of overexposure. Objectives: This study aimed to explore existing literature to identify standard methodologies used in defining HEGs. It also sought to assess how exposure metrics assigned to HEGs align with those of the individual occupations nested in them. Additionally, the research compared individual exposure characteristics with both their respective occupations and associated HEGs over five years, or a shorter duration where data availability was limited. Methods: This cross-sectional study analysed secondary data from two underground gold mines, focusing on respirable crystalline silica measurements recorded in an electronic database. Exposure data were collected through gravimetric sampling over four years (2019–2022). The study population consisted of individuals whose airborne pollutant exposure measurements were sampled while working in the Stope and Development End areas at different levels in underground mines. Ten HEGs were selected based on consistent exposure measurements and with at least more than two occupations recorded per year throughout the study period. To evaluate occupational exposure levels, a combination of statistical methods, including Independent t-tests, ANOVA, Welch’s ANOVA, Kruskal-Wallis tests, and Linear Mixed Models, was applied to assess variations in exposure within HEGs and across individual workers. Additionally, measures of central tendency (AM, SD, GM, GSD, and 90th percentile) were used to identify exposure trends over time. v Results: Statistical analyses indicate that exposure levels within HEGs were generally comparable across occupational groups, with no statistically significant differences (p > 0.05) in most cases. However, some occupational roles, particularly in Stoping activities, exhibited higher exposure variability. Though, some occupational roles, particularly in Stoping activities, exhibited higher exposure variability. However, some HEGs were accurately defined, with exposures generally comparable across occupational groups, whilst others showed clear deviations from expected homogeneity. Stoping activities, in particular, displayed high variability, with certain workers experiencing significantly higher or lower exposures than their HEG classification suggested. Development areas also remained a concern, with exposures in some years approaching or exceeding the RCS occupational exposure limit (OEL). Individual-level analysis revealed notable outliers, reinforcing that group-level estimates can mask extreme exposures. Over the four-year period, exposure patterns fluctuated, with spikes in 2021 and 2022 likely linked to operational or environmental changes. The GM values confirmed right-skewed exposure distributions, indicating that a small proportion of workers bore disproportionately high exposure risks. The 90th percentile metric proved effective in identifying these upper-bound risks. Overall, while HEG classification remains a useful tool, these results show that relying on it alone may lead to under- or overestimation of risks, highlighting the need for targeted individual monitoring, enhanced dust control, and periodic HEG reassessment. Despite overall declining exposure trends, some occupations in the Stoping and Development Ends under the study recorded RCS concentrations which, in certain years, remained close to or exceeded the RCS occupational exposure limit (OEL) of 0.1 mg/m³. Conclusion: The literature review and statistical findings underscore the significance of accurate HEG classification in occupational exposure assessment, particularly in mining environments. This study confirms that while HEG-based exposure monitoring offers a practical framework for occupational exposure assessment, it is not without limitations. HEGs generally grouped workers with similar exposure profiles; however, individual deviations and temporal fluctuations exposed critical gaps in accurately capturing personal exposure risk. The results advocate for a hybrid approach that combines traditional HEG methodologies with individualized or task-based monitoring to improve exposure classification. Additionally, reliance on AM alone for risk categorization may obscure peak exposures, reinforcing the need to incorporate complementary metrics such as GM, GSD, and the 90th percentile. To ensure worker protection and regulatory compliance, periodic reassessment of HEGs and the integration of real-time monitoring technologies are recommended. These refinements are essential for enhancing exposure estimation accuracy and improving health outcomes in dynamic mining environments.