Probabilistic design of storage capacity for rainwater cistern systems

The paper demonstrates the potential of a rainwater cistern design based on a probabilistic approach. Traditional cistern designs do not allow any water shortage. The tank-size design in this study permits the possibility of failure to account for the characteristics of the crop, which must sustain water shortage to a certain degree at infrequent intervals. Sequential peak analyses, combined with the failure probability concept, were adopted in the analytical procedure. A hilly area in northern Taiwan was selected as the study site, where water shortage is a severe problem for tea cultivation. Rainfall records and water-demand information were collected for analysis to determine an adequate tank size for a stable water supply. The determination of the tank size is related to the rainwater abstraction coefficient, the ratio of the cultivated area to the rainwater collecting area and the failure probability. The results are expressed as a regression equation to provide a convenient way for farmers to set up cistern systems in this area. (C) 2000 Silsoe Research Institute.