Terry Thomas (2002)
In every roofwater harvesting system there is a trade-off obtainable between increasing system performance and increasing system cost. The medium of this trade-off is normally tank size and hence at the centre of RWH design is the optimisation of that size. Many methodologies for tank sizing have been published. Fortunately, as with all optimisation, the plot of performance (e.g. cost:benefit) against tank size is ‘flat-topped’ in the area of interest, so that a ±10% variation in size there has little influence on economic performance. The cost of a tank of given size can usually be readily assessed, but the performance of the system containing that tank cannot. We therefore seek a methodology for predicting performance over a system’s expected life: one that suits both the surrounding constraints (e.g. of data availability and access to computing facilities) and the RWH system use envisaged.
Suitable meteorological information is actually, at least in developing countries, rarely available and affordable in the right form and for the exact location of interest. This note addresses the specific problem of inadequately rainfall data.
The method was deemed sufficiently accurate to justify offering an open-access RWH system modelling service which is driven by user’s monthly rainfall data and proposed system details and which yields reliability, satisfaction and efficiency estimates.