Rainwater harvesting systems (RHS) have been increasingly considered an important adaptive measure to improve water security in urban areas and water conservation motivates their dissemination as a public policy. However, the performance of this public policy includes the assessment of benefits for both the community and the water utility. The literature has showed the variability of RHS feasibility according to climate (rainfall regime) and technical parameters (RHW demand and roof area). However, the impact of the deep uncertain factors (such as water service tariff, discount rate, operational and maintenance costs) in the performance of the system are still of great concern. This work aims to propose a method for assessing the performance of RHS under the context of deep uncertainty both for households and for water utilities. We focused the analysis on three cities in the State of Goias, Brazil and used economic (Net Present Value) and technical performance indicators (Percent of Satisfied Demand-SD and Reliability-REL) in an adaptive and robust decision making framework. Given the selected configuration with best performances under an ensemble of deeply uncertain scenarios (State of the World), we further evaluated the impact of the RHS operation on the technical and economic sustainability of the water utilities. The method uses a historical data of technical and economic indicators available in the SNIS (National Sanitation Information System). The general methodology is broadly applicable to RHS evaluation as public policy to improve the water security in urban areas facing demand pressure and water availability restriction.
