In this paper, we define and formulate the shelter location-allocation problem considering both the network vulnerability of the affected area and the shelter’s disruption during the disaster management’s response phase. We capture the vulnerability metric using the traveling cost and location vulnerability for shelter disruption using the shelter’s operating cost. We formulate the problem as a mixed-integer linear programming (MILP) model and present an evacuee-allocation plan considering vulnerable network connectivities between the populated areas and the shelter locations. We finally apply and solve the problem using real-life case data obtained during the Nepal earthquake in 2015 and compare our models with Rahman’s same data [10]. We demonstrate using the case example the usefulness of our modelling approach and show that we can achieve better results compared to a simulation study.