From end-use characterization to decentralized reuse design: Closing the data-to-system gap

Decentralized water reuse systems are increasingly deployed, yet their design standards and quantitative microbial risk assessment (QMRA) frameworks rely on assumed exposure parameters rather than empirical data. Concurrently, smart metering research has generated high-resolution, fixture-level demand profiles that remain disconnected from reuse practice. This review asks whether empirical end-use data support the exposure parameters assumed in non-potable reuse risk models, bridging three siloed literatures: residential end-use characterization, decentralized reuse system design, and QMRA. A systematic comparison of QMRA exposure assumptions against smart-meter measurements shows that the assumed toilet flush frequency (5.0/person/day) lies at the low end of smart-metering estimates (5.0-6.4/person/day), the upper value exceeding the assumption by 28 % and underestimating enteric exposure. Shower durations assumed in Legionella risk models (15 min) exceed the measured mean (7.8 min) by 92 % (a factor of 1.9), potentially overestimating inhalation risk. Accidental ingestion volume during toilet flushing, the parameter to which QMRA is most sensitive, has never been empirically measured, to the authors' knowledge. A sensitivity analysis shows that substituting measured for assumed values shifts modelled infection probabilities enough to cross the 10⁻⁴ annual infection benchmark. Current regulatory frameworks (ISO 20426:2018, EN 16941, US EPA 2012/2025) do not require empirically grounded exposure parameters. An integrated framework is proposed, linking fixture-level data to system design and risk assessment, and priority research needs are identified, including tracer-based ingestion measurements. The analysis is grounded in North American and European smart-meter data; applicability to regions with dual-flush toilets or absent metering infrastructure requires context-specific validation.