This paper studies effective algorithms for designing wireless sensor networks to be used for health monitoring of railway structures. The first step reviewed optimization problems arising in railway system and highlighted the importance of maximizing the network lifetime in a poor communication environment and minimizing the total cost of the wireless sensor network in a general communication environment. Basic ideas for a life span algorithm for the network lifetime maximization problem and a Lagrangian heuristic algorithm for the total cost minimization problem are introduced in this paper. Finally these algorithms were verified with randomly generated data for the former problem and data acquired from a real WSN used for railway structural health monitoring for the latter problem.