Sectioning is one of the key mitigation strategies in pipeline transport of liquid hydrocarbons. The valves located along pipelines reduce the maximum volume that may be spilled decreasing economic, social and environmental losses. Defining the location and number of valves in a specific pipeline section is a challenging decision due to the countless combinations of these two design components (i.e., where and how many valves). In this work, we address the valve location problem (VLP) for sectioning using an optimization approach based on a shortest path problem, which assesses the possible location of valves to minimize environmental risk and guarantee a tolerable value of individual risk. To estimate and quantify both the environmental vulnerability and individual risk a new framework is proposed, employing the spill volume and ecological and social characteristics as inputs. We present two case studies for sectioning in Latin-American pipelines; the problem is solved using a Bellman-Ford algorithm with CPU times up to 40 s. The results show reductions on the order of 50%-85% of the maximum possible spill volume compared with the current situation, in each case respectively. The resulting valve configurations cover areas with high environmental vulnerability, reducing ecological losses and guaranteeing a level of individual risks lower than the acceptable value on all populated areas. (C) 2016 Elsevier Ltd. All rights reserved.