Real-Time Control of Rainwater Harvesting Systems: The Benefits of Increasing Rainfall Forecast Window

Abstract

 Four contrasting Real-Time Control strategies were applied to simulated rainwater harvesting systems  Long lead-time rainfall forecast (7-day) enhanced the ability to reduce flood risk and restore baseflow, with little impact on water supply efficiency  Using long lead-time rainfall forecast has the potential to holistically restore natural flow regimes.

Quantifying the performance of dual-use rainwater harvesting systems

“…This can result in either a tank that is too full (resulting in spills) or too empty (resulting in subsequent supply shortage) before a storm. Although the systematic bias applied to the Active system’s emptying volume did not significantly alter results, further strategies to improve the capacity of the Active system to mitigate flood risks and control flow rate could be employed, such as reducing the active release flowrate and utilising 7-day rainfall forecasts ( Xu et al., 2020 ) or adopting a minimum emptying time of 48 h before a storm ( Woods-Ballard et al., 2015 ). Such strategies are also impacted by issues of forecast accuracy.…”

“…The performance of active systems can be significantly affected by rainfall forecasting error. The main source of uncertainty is errors in rainfall intensity, which result in either over or under estimation of volume to be emptied ( Xu et al., 2020 ). To simulate this potential inaccuracy, additional sensitivity analyses were undertaken in which a systematic bias of ± 10% was applied to the emptying volume for every event.…”