This paper presents an O (mn\ log U) O (mn log U) time algorithm to solve the feasibility problem, where n, m and U are the number of nodes, arcs, and the value of maximum upper bound, respectively. The merit of this paper in comparison with maximum flow algorithms, is that it presents some information that is useful to modeler in estimating the maximum cost of adjusting the infeasible network. Our algorithm improves upon the previous O (mn\ log (nU)) O (mn log (n U))-time method due to Ghiyasvand
