Main content area

Analysis of water consumption in toilets employing Shewhart, EWMA, and Shewhart-EWMA combined control charts

FREITAS, Lucas Lepinski Golin, Henning, Elisa, Kalbusch, Andreza, Konrath, Andrea Cristina, Walter, Olga Maria Formigoni Carvalho
Journal of cleaner production 2019 v.233 pp. 1146-1157
equipment, females, males, models, monitoring, process control, statistical analysis, water analysis, water supply
The objective of this article is to statistically analyze and apply process control on water consumption in toilet flush devices in a public university building using Shewhart, EWMA and combined Shewhart-EWMA control charts. Data such as water volume, number of activations and flush-time duration were collected utilizing Smart-Meter sensors. A script was developed in R language to perform the appropriate statistical analysis, such as finding control charts control limits and transforming the obtained raw data into key variables: total volume per day, average flush volume, flush duration, and number of flushes per day. During the data gathering period, the regular single-flush equipment was replaced by a dual-flush model. The control charts Phase 1 is based on the collected data before the installation of the dual-flush devices, while Phase 2 is based on after replacement. After installing the new dual-flush devices, all control charts displayed visible changes based on the key variables. The replacement of the flush valves brought about a considerable reduction in water consumption in both male and female restrooms. There was a 21.72% reduction in average water consumption and a 12.09% average reduction in flushing time duration. The equipment replacement resulted in an evaluated Internal Rate of Return of 0.73%. Periodic monitoring of Shewhart Control Charts proved useful for observing large shifts of data, while EWMA control charts were more useful in locating small but constant data shifts. All chart models proved useful in identifying water consumption reduction and events such as leakages and water supply cutoffs occurring during the data gathering period.