Characterising indoor airborne particulate matter in Sharpeville, Gauteng

Abstract

Solid fuel use for domestic purposes is a significant air quality and health challenge within low-income settlements in South Africa. Fuels such as coal, wood and other “dirty” household energy carriers (including paraffin and crop waste) add to the problematic indoor airborne concentration of pollutants detrimental to the health of the residents in low-income communities. Sharpeville is no different to these other low-income areas; although the majority of low-cost houses have access to electricity, use of household solid fuel for domestic activities is still prevalent. The purpose of this research was to characterise indoor airborne respirable particulate matter (PM₄) in the low-income community of Sharpeville in Gauteng in order to inform health assessments in terms of household energy use. To reach this aim, the study took the direction to determining the dominant fuel in the community’s household energy mix; continuously measuring the variability of PM₄ over time, and evaluating the elemental characteristics of indoor respirable PM. A questionnaire based detailed energy use survey was carried out to determine the dominant household energy carrier used by the Sharpeville community. Following the survey, DustTrak 8530 fitted with a 10-mm Nylon Dorr-Oliver Cyclone to attain a 50% cut-off at 4 µm were installed at 16 randomly selected houses. The DustTraks were continuously measuring respirable airborne PM at 1.7 L.min-1. Gravimetric sampling using Gilian GilAir sampling pumps with a cyclone inlet offering the same PM cut-size and connected to a 37mm cassette containing Mixed Cellulose Ester (MCE) filters was conducted simultaneous with photometric samplers. Although this was done as a reference method to compensate for overestimation that comes with the laser photometers, a correction factor of 0.715 was adopted from Language et al (2016) and applied to the data. Moreover, indoor temperatures were continuously measured within all sample houses for both winter and summer sampling campaigns. Additional measurements of black carbon (BC) were taken in one of the sample houses that used solid fuels as their primary household energy carrier. The household energy mix within the community was mainly found to consist mainly of three “dirty” fuels which were wood, coal and paraffin (kerosene). Wood was found to be the predominant fuel used by the largest number of households in Sharpeville {winter: 2973 (24%); summer: 1122 (9%)}. Compared to coal and paraffin usage, wood use is higher by a factor of 1.5 in winter and 2 in summer. However, when calculated per coal-using and per wood-using household, coal was the main energy carrier in winter and wood in summer. Whereby an average 85 kg/month of coal is consumed whereas wood and paraffin usage was 47 kg/month and 14 litres/month in winter and only 8 47 kg/month, 13 47 kg/month and 9 litres/month respectively in summer. This is indicates coal as the preferred fuel for winter months whereas in summer, wood is more reliable due to the diminished need to burn for extended periods of time. The average indoor diurnal pattern of PM₄ shows a bi-modal distribution, with an extended period of elevated PM₄ concentrations during the evening peak. This pattern was observed in all houses in which measurements were taken. Houses that do not use solid fuels (non-solid fuel burning (NSFB)) combustion as a primary source of energy experienced lower concentrations throughout the day. Nevertheless, the concentrations were still found to exceed 24hr average PM₂․₅ National Ambient Air Quality Standards (NAAQS) in the winter during peak hours of the day. There are no existing NAAQS for PM₄ thus for the purpose of this study PM₁₀ and PM₂․₅ NAAQS are used as a proxy to measure the extent of the personal exposure to PM₄ within households. Summer concentrations in the NSFB households were mostly below both 24hr PM NAAQS. Although there was a ~47% decrease in average concentration levels of respirable PM in solid fuel burning (SFB) households, PM₄ was problematic during both seasons in these houses. In most cases, residents of SFB households were found to be exposed to PM₄ levels above the 24hr PM₂․₅ and PM₁₀ NAAQS throughout the year. This represents a health hazard for these communities. Moreover, additional black carbon (BC) measurements conducted continuously in one of the SFB houses during the summer campaign. The dataset supports the assumption of continued use of solid fuels during this period as it shows a bi-modal distribution and elevated concentrations during peak hours of the day. Housing physical structures need to ensure thermal properties are suitable for keeping the indoor environment thermally satisfactory to the dwellers. Indoor temperature profiles during winter seem to be below the minimum World Health Organisation (WHO) of value 15°C for over 60% of the day (14.4 hours). Manual indoor warming (using clean or solid fuel) was observed and assumed to occur in the late evening when ambient temperatures drop below 15 °C but indoor temperature remained on average above 17 °C although they rise to about 19 °C from 18h00 to 21h00. They then dropped again at about 22h00 but do not go below the WHO minimum value. Gravimetric mass concentrations confirms the results obtained from the photometric measurements. Concentrations in the NSFB were lower than SFB houses. Highest concentrations were measured in SFB houses in winter. In addition to mass, the collected filters have been analysed by X-Ray Fluorescence (XRF) to obtain the elemental composition of the airborne particles in houses. The elemental species composition was found to behave differently over both daily and seasonal timescales. These species (mostly soil-related) were found highly concentrated either during the day in summer or night in winter, or vice-versa. In an SFB household, combustion elements were mostly found during the night in winter, which is a sign of domestic burning in the evening which thus elevated PM4 concentrations during the evening peak in the SFB house. However, according to the Principal Components Analysis (PCA) multivariate factor analysis, in both household types, most of the elemental composition was dust-related. This signifies the impact of ambient dust infiltration to the indoor environment. Winter PCA results confirmed domestic burning because sulphur (S), potassium (K) and chlorine (Cl) were significant elements in the SFB household. The calculation and determining of non-crustal source enrichment show high impacts of ambient air on the indoor air of both household types, although most especially in NSFB households. The highly enriched S content in the PM samples in a house completely reliant on electricity was attributed to ambient source contribution such as from the industrial and energy sector. A further potential source was domestic SFB practices in the community. The study contributes to the overall aim of understanding pollution related health risks in low-income communities in South Africa. It also adds to the body of science in terms of showing the variability of low-income households and drivers to the individual use of solid fuels as household energy carriers. The study fosters an understanding of other air pollution sources to the indoor environment and the health of people thereof.