The goal of a water system design is to maximize the value, or benefit, of the system while reducing the cost of the system. Minimizing cost alone may result in the smallest pipe sizes, which leads to the minimum-capacity design. The least capacity is not the preferable solution for long term system planning; some extra pipe capacity is beneficial to allow the supply to grow into its full capacity within a planning horizon to account for uncertainty in demands and to meet the need for reliability in case of outages.

The true benefit of water system design is to reliably supply service of adequate water quantity and quality. Provision of sufficient water supply must be ensured for a community not only at the present time but also in a reasonable planning horizon. During this planning period, the amount of water required for a system, or the demand, is estimated, and this is typically performed with some uncertainty. Thus, it is difficult to precisely forecast the demand. In order that a design is carried out for the maximum value or benefit for a water distribution system, engineers must be able to determine the maximum benefit within a budget.

The benefits of a design and rehabilitation may result from hydraulic performance improvement (hydraulic benefit), excess hydraulic capacity (capacity benefit), and pipe rehabilitation improvement (rehabilitation benefit). The hydraulic benefit is measured by using a surrogate of the junction pressure improvement. In this version of Darwin Designer, only pressure benefit is considered.

Pressure benefit is measured by the improvement of junction pressure of a design. If the pressure at a junction exceeds the minimum required, this shows the system has some extra capacity, which is considered a benefit. For some nodes, where the pressure is already high, you may want to exclude the node from the pressure benefit calculation because there is no value in increasing pressure at that node. (This is done in the Pressure Constraints tab.) For other nodes, the first unit of pressure is worth a great deal while subsequent units of pressure improvement are not worth as much. For example, if the minimum pressure is 20 psi, the increase from 20 to 21 psi is worth a great deal but an increase from 60 to 61 psi is not worth as much. To account for this effect, you can lower the exponent b in the benefit calculation from the default of 1 to a lower value, say 0.5.

With the definition of a benefit function as one of design objectives, the optimal design is no longer a single-objective (minimizing cost) optimization problem but a multi-objective (minimizing cost and maximizing benefit) one. A multi-objective optimization enables engineers to create a design that trades off between cost and benefit. The trade-off optimization problem is solved by using a competent genetic algorithm.

Darwin Designer concurrently optimizes two conflicting objectives and produces a set of Pareto optimal (i.e. non-dominated, non-inferior) solutions. One objective solution, such as cost, cannot be improved (minimized) without diminishing the other objective (reducing benefit). Therefore, a Pareto optimal solution set represents the best design solution for each cost range. Engineers can further justify the best design by other non-quantifiable criteria.