Data-driven method for estimating fixture use probability in simultaneous peak water flow calculations

Simultaneous peak water flow (SPWF) is a fundamental parameter calculated during the design phase to ensure the proper sizing of water systems in buildings. Over the years, overestimation has been a persistent problem in SPWF studies, leading to inefficiencies in energy and water use. This research aims to address this problem by developing a flexible data-driven approach to estimate water fixture use probability (FUP), a key variable in SPWF calculations. The proposed method uses artificial neural networks to model the water demand based on the type and FUP of water fixtures in the building, drawing on the Wistort model and Water Demand Estimation Model (WDEM). Differential evolution was used to optimise the FUP values based on empirical water flow data. The model was applied to two non-residential buildings of different sizes and uses, where results showed that the models closely estimate the empirical SPWF compared to building design codes. Due to its use of adaptable parameters, flexible underlying models, and real-world data, the developed approach can be applied to buildings of any type and size. By reducing the risk of SPWF overestimation, the method contributes to more efficient water and energy use, ultimately supporting sustainability goals in the built environment. Practical application This study presents a new methodology for estimating SPWF for buildings. The methodology is flexible and can be applied to different types and sizes of buildings. The use of real-world data aims to provide a more practical model that will closely estimate the SPWF of the buildings, as compared to the design flow rates from the design codes, for more cost-efficient, energy-efficient, and water-efficient building water supply systems.