A numerical optimization procedure for complex pipe and duct network design

Abstract

The design of optimum pipe and duct networks with available procedures is difficult, if not impossible. A more efficient procedure that will automatically produce the optimum design is required. Such a procedure is presented in this article. The design is formulated as a constrained nonlinear optimization problem. This problem is solved using a unique numerical optimization algorithm. The solution entails the calculation of the cross sectional dimensions of the ducts and pipes so that the life cycle cost of the network is minimized. The topology equations are derived using graph theory thereby allowing complex networks with loops to be designed numerically. A duct network consisting of a fan and 35 duct sections is designed according to certain specifications. Using the proposed procedure optimum designs were obtained within seconds on a 33 MHz 486 micro‐computer. The procedure was further applied to the optimization of a coal pipeline. It is shown that the optimized solution will cost 14% ($8 million) less than the previous design with conventional design techniques.