Extending SIMDEUM: Stochastic Water Demand and Wastewater Modeling in the Educational and Food Service Sectors
Year: 2025
Type: article
Abstract: Water demand models aid water planning and management in buildings but are mainly limited to residential settings. This study extends simulation of water demand, and end-use model (SIMDEUM), a stochastic end-use water demand model, to educational and food service buildings through (1) developing water-use probability distribution functions (PDFs) based on occupancy patterns, and (2) integrating new end uses relevant to these sectors, namely, time-controlled urinals, ice machines, and steam equipment. An approach was implemented to estimate urine volume for frequent sanitary fixture use and allow different user types to share the same facilities simultaneously. Based on occupancy data from university lecture buildings and Google Popular Times data for food service businesses across the Republic of Ireland, representative occupancy patterns were identified via clustering analysis. These patterns were used to create water-use PDFs. The extended model was validated using high-resolution water-use data from a university lecture building and a campus restaurant, and benchmarked against the deterministic SIMDEUM Pattern Generator (SPG). Results show that the extended model predicts water demand profiles well in the educational building (R2 of 0.89–0.95) and the restaurant (R2 of 0.54–0.91), outperforming the SPG in both cases (R2 of 0.87–0.93 and −2.19 to 0.40, respectively). It predicts peak and total daily demand well for the educational building (R2 0.82–0.89 and 0.78, respectively), but is more nuanced for the restaurant (R2 of 0.04–0.46 and 0.37–0.77, respectively). The model produces realistic end-use water demand breakdown and wastewater discharge patterns and characteristics for both buildings. The findings demonstrate that (1) SIMDEUM can be adapted to various complex mixed-occupancy buildings when occupancy and end-use information is available, and (2) Google Popular Times data can be used for water demand modeling in diverse nonresidential settings. Further research could explore a broader range of buildings, refine water-use PDFs, and develop empirical design-demand equations for large-scale implementations.
Institution: University College Dublin
Language: English
Subfield: Ocean Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4416529657