Use of stratified cluster sampling for efficient estimation of solid waste generation at household level

Abstract

Relatively few studies have been performed to characterize municipal solid waste (MSW) at household level. This is due in part to the difficulties involved with collecting the data and selecting an appropriate statistical sample size. The previous studies identified in this paper have used statistical tools appropriate for analysing data collected at a material recovery facility or landfill site. This study demonstrates a statistically sound and efficient approach for characterizing MSW at the household level. Moreover, a household approach also allowed for consideration of the socio-economic conditions, level of waste generation, geography, and demography. The study utilized two-stage cluster sampling within strata in Kathmandu Metropolitan City (KMC) to measure MSW for 2 weeks. In KMC, the average household solid waste generation was 161.2 g capita( -1) day(- 1)with an average generation rate between 137.7 and 184.6 g capita(-1) day(-1) for a 95% confidence interval and 14.5% relative margin of error. The results show a positive relation between income and waste production rate. Organic waste was the biggest portion of MSW, and hazardous waste was the smallest of the total. Sample size considerations suggest that 273 households are required in KMC to attain a 10% relative margin of error with a 95% confidence interval.