This study gives a MILP framework for solving a heat and water allocation problem. Both energy and water consumptions are taken into account for proposing numerous optimal solutions for a fixed number of connections between processes in the network. Results are represented on Pareto fronts for each number of connections. Networks proposed are optimal in terms of energy and water consumptions and can easily be applied at industrial scale because they involve low flow rates and their complexity is also topologically limited. In previous studies, the optimal network reaches the fresh water target but it is not always an optimal choice, as shows the economical study carried out here. A low increase in fresh water consumption leading to a significant decrease in energy consumption allows designing a cheaper network.