In a gas transmission network (GTN), faults can easily propagate due to the interconnections of streams. The main objective of this paper is to develop a systematic framework for an online decision support system (DSS) in order to make the right decisions to get the GTN out of critical conditions (which cannot be handled by the plant controllers) smoothly. One of the key features of the proposed scheme is its lack of dependence on prior knowledge of the fault signals (e.g., number of faults, and their origin). In this article, the GTN is modeled by a fuzzy directed graph (FDG). The proposed approach utilizes a reasoning algorithm based on the deviations that exist in the process variables (attributes) of target nodes, in order to detect the most crucial attributes (equipment) whose manipulating variables can be changed appropriately (i.e., upward or downward) to get the GTN out of the abnormal condition gracefully. Thereafter, quantitative decisions are made from qualitative decisions, after which the forward reasoning is used to determine the impact of each quantitative decision on the attributes of target nodes. The attributes of the affected nodes are obtained based on their estimations by a fuzzy inference system (FIS) through traversing from affecting nodes to other nodes in a sequential manner. To keep the generality and flexibility of the proposed scheme, FIS uses preliminary intuitive rules between two connected nodes rather than empirical rules for the entire GTN. Eventually, all tentative feasible decisions are built. These feasible decisions get ranked and prioritized through an analytic hierarchy process method. The decision set with highest priority is then implemented whose effect on the GTN operation is assessed, and if it seems the GTN still requires some changes to get out of an abnormal situation, once again DSS is invoked based on the newly established condition and the appropriate set of decision is obtained accordingly. The performance of the proposed scheme is shown by its application for two GTNs as benchmarks. It is capable not only of offering decisions to alleviate an abnormal condition but also of leading the plant to a new desired state from a normal condition.