NETWORK LINEAR-PROGRAMMING CODES FOR WATER-SUPPLY HEADWORKS MODELING

There is a class of water-supply headworks models that uses network linear programming to assign water within the network. The choice of algorithm and computer code to solve the network linear programs formulated by such models becomes important when doing planning studies that may involve thousands of replicates of future stream-flow and demand conditions. The computational performance of two public-domain FORTRAN codes, NETFLO and RELAX, representing efficient implementations of the primal-simplex and primal-dual algorithms, were compared for two headworks systems. The results showed that for the two small network linear programs typical of headworks modeling, the RELAX code used without prior flow information was about 100% faster than the NETFLO code. When prior flow information was employed, further substantial gains in speed were realized. The second and subsequent iterations within a time step were solved in about one-third of the time for the first iteration. The actual overall speed gain depends on whether the prior information is used across time steps or just between the iterations in a time step.