In this work the determination of optimally located pipeline networks has been proposed by means of the implementation of a metaheuristic algorithm called Simulated Annealing with GAMS (SAG) in order to find the best pipeline layout together with a subset of locations to install concentrating nodes. The strategy essentially consists of a hybridization of Simulated Annealing, combined with the well-known GAMS package. In particular, the sample cases consisted of finding the most convenient routes so as to transport natural gasoline from Santa Cruz (Argentina) gas fields to the processing plants. The SAG algorithm behaved satisfactorily because it proved to be efficient and flexible.