Climate variability is reflected in water affluence that directly impacts the availability and level of water in reservoirs. Economic instruments, such as tariffs, are increasingly being used to promote the rational use of water. This study offers a floating rate model for charging raw water as a function of the reservoir level. The model was designed in the Jaguaribe–Metropolitan System, a place characterized by water-use conflicts during water scarcity. The model was developed as a function of the unit payment capacity (UPC) and drought states of the hydrographic regions. We applied the free software R and PSO optimization package to define the optimal UPC fraction and evaluated the model sensitivity using a synthetic series generated with the Markov chain method. Consequently, we observed that comfortable, alert, and critical situation levels predominated in the sensitivity analysis, and the average collections paid for administration, operation, and maintenance (AO&M) costs. In some cases, a collection surplus that can be managed through a financial fund to finance the system in periods of critical shortage is generated.