Linearized optimization model for reliability-based design of water systems

Optimal network design constrained by reliability is an improvement over the traditional methods for water-distribution system analysis. But the large computer time for reliability-optimization restricts its practical use for design. In this paper, a new technique is developed that applies a linear programming algorithm to optimal design constrained by reliability. The objective function and constraints are expressed by a first-order Taylor series expansion, and three submodels (a steady-state simulation model, a reliability model, and a linear optimization model) are linked in the algorithm. The optimal search is done in two steps. The first step is to search for an optimal solution and then the reliability and pressure heads, evaluated on this configuration, are compared with the constraint values. If there still exists a possibility that the cost of the system can be further decreased a new iteration is repeated. The computing burden is significantly decreased and, therefore, the proposed reliability-optimization technique can be used in planning large networks.