Performance comparison of the Nanozen Dustcount personal real-time monitor with a conventional respirable dust exposure sampling method

Abstract

Title: Performance comparison of the Nanozen DustCount personal real-time monitor with a conventional respirable dust exposure sampling method Introduction: The mining industry, a vital contributor to many economies, particularly in South Africa, poses significant health risks due to dust exposure. Dust from mining operations varies in composition and particle size, influencing its health impacts. Safeguarding workers' health has been a legislated priority since 1911, with legislation mandating exposure monitoring and the implementation of control measures. Conventional gravimetric sampling is the most widely used and accepted standard for air monitoring, using a pump and filter system to collect samples in a worker's breathing zone over a shift. Although reliable, this method involves gravimetric analysis of the filter in a laboratory, which delays results and limits timely interventions. Advancements in technology and the development of direct-reading instruments, such as the Nanozen DustCount® 9000-Z1 personal real-time monitor (Nanozen Industries Inc., Vancouver, Canada), provide faster, real-time dust exposure measurements. The accuracy of these devices must, however, be validated against conventional gravimetric methods, such as the Methods for the Determination of Hazardous Substances (MDHS14/4) method, to ensure their reliability. A limited number of published studies have evaluated the performance of the DustCount, and even fewer assessed the performance of the DustCount through collecting samples side-by-side with conventional gravimetric sampling methods in a work environment. Method: This study evaluated the performance of the DustCount device compared to a conventional gravimetric sampling method through quantitative analyses of secondary data. The data was collected by a company contracted to conduct exposure monitoring in various areas of an open-cast iron ore mine in the Northern Cape, South Africa. A dataset of 35 paired results from side-by-side monitoring was analysed. The real-time respirable dust concentration results obtained from the DustCount’s optical particle counter were compared to those obtained using a cyclone following MDHS14/4, a conventional gravimetric sampling method. Additionally, the respirable dust concentration results from the DustCount’s optical particle counter were compared to those collected by the DustCount’s impactor and filter. Lastly, the respirable dust concentrations determined by the DustCount’s impactors were compared to those obtained using the cyclone following the MDHS14/4 method. Results: The geometric mean of the conventional gravimetric concentration (CGC) dataset was 0.416 mg/m³, with a coefficient of variance (CV) of 175%. In contrast, the DustCount real-time concentration (DRC) dataset had a lower geometric mean of 0.172 mg/m³ and a CV of 111%.Across all samples, the CGC method consistently reported higher respirable dust concentrations than the DRC method. Bland-Altman (BA) analysis confirmed that the DustCount underestimated respirable dust concentrations, with a mean bias of −71.41%, standard deviation (SD) of 35.17%, and 95% limits of agreement ranging from −166.6% to 11.84%. This significant underestimation highlights a key limitation of the DustCount for precise exposure quantification. Conclusion: The significant bias of −71.41% suggests that the DustCount instrument is inaccurate in measuring respirable dust concentrations, as it consistently underestimates values compared to the reference method. The large standard deviation of 35.17% and the wide limits of agreement (ranging from −166.6% to 11.84%) indicate a lack of precision, suggesting that the DustCount provides inconsistent measurements when compared to the reference method. Despite these limitations, the findings suggest that the DustCount has potential value as a real￾time monitoring tool, as it does reflect temporal changes in respirable dust concentrations. With further refinement — particularly regarding the ease with which correction factors can be established — the instrument’s real-time capabilities could facilitate immediate interventions and enhance workplace safety by providing near instantaneous feedback. Word count: 535