FIS-based approach to estimate probable flows for sizing pipes in building water supply systems

The paper presents Fuzzy Inference System (FIS) -based approach to estimating probable flows for sizing building water supply pipes. Instead of calculating the design flow by estimating the number of fixtures operating simultaneously using probabilistic or deterministic approach utilizing a large amount of high-resolution field data, the proposed methodology uses Fuzzy logic to estimate demand characteristics of fixtures, unlike conventional fixture unit method. The developed methodology is based on FISs derived from users' socio-demographic characteristics, their diurnal activities and water use habits. A MATLAB program was developed to estimate the probable flows through the pipe using inputs-type of accommodation, connected fixture types and fixture flow rates. An imaginary 7-storey residential building with a gravity tank system was considered for pipe sizing. The design flows obtained by the proposed methodology were 50-70% of the those produced by the fixture unit approach leading to smaller pipe sizes.