Stochastic Rainwater Harvesting System Modeling Under Random Rainfall Features and Variable Water Demands

A stochastic rainwater harvesting system modeling (StRaHaS) method is developed to enhance both accurateness and applicability of hydrologic modeling in guiding extensive applications of individual rainwater harvesting (RWH) systems (especially over large data-scarce regions). The reliability of a RWH system is characterized as the fraction of time water demands being satisfied by a rainwater storage unit (RWSU). The variations of water balances, the post-rainfall RWSU full-storage probability, and the system reliability with random rainfall features, variable water demands, and the other RWH system characteristics are derived as analytical, accurate, and easily applied models through stochastic integration. StRaHaS is applied to three domestic RWH systems in Toronto, Canada. Due to high accurateness and low complexity in modeling, designing, assessing, and analyzing the RWH systems, StRaHaS outperforms existing methods (e.g., inaccurate water-balance estimation, complicated continuous simulation, and simplified stochastic simulation). Based on StRaHaS, we quantitatively expound the increases in the reliability of the systems with higher rainwater supplies (corresponding to higher rainfall depths, and lower rainfall losses and first-flush depths), lower water demands (in shorter wet and dry periods), and higher RWSU capacities. Long dry periods are lengthened by climate change and play dominant roles in low, spatially heterogeneous reliability of the systems. Synchronic changes and extremely high values of rainfall features do not significantly affect the reliability. Overall, StRaHaS is a promising method in modeling RWH systems, revealing RWH mechanisms, scientizing RWH applications, and facilitating the Sustainable Development Goals for addressing global water issues.

Automated summary

The StRaHaS method aims to improve the accuracy and applicability of hydrologic modeling for individual rainwater harvesting systems. By analyzing random rainfall features, variable water demands, and other characteristics, it derives models for water balance variations, post-rainfall storage probability, and system reliability. The method outperforms existing methods in accuracy and simplicity, and shows promising results in enhancing the reliability and sustainability of rainwater harvesting systems.