Methodology for rainwater reservoir dimensioning: a probabilistic approach

Resumo

The aim of this study was to propose a new methodology for reservoir rainwater dimensioning based on probabilistic modeling. Eucalyptus seedlings grown in a greenhouse were used to obtain a hypothetical water demand. Meteorological data were used to estimate the demand (evapotranspiration) and offer (rainfall over the greenhouse coverage). The probability distribution of Wakeby presented the best fit for the rainfall data; therefore, a Wakeby distribution was used to model the flow-duration curve of the greenhouse coverage. For a payback period (T) of 10 years of surplus water demand and water supply deficit, a reservoir with 13.60 m³ was obtained. The proposed methodology combined the simultaneous occurrence of the events to enable the scaling out of a reservoir with high safety to supply the required demand (T = 100 years) and therefore enables a lower cost of deployment compared to each approach separately (T = 10 years).