The startup of distillation columns is one of most difficult operations in the chemical industry. In order to decrease the cost of operation, the startup period is to be shortened through optimizing startup operation policies. The conventional direct setting operating policy for startup is pre-defined but it may be neither optimal in terms of startup time nor does it consider the feasibility of the operation. Moreover, time optimal control does not consider the variability due to the disturbances during the startup which can affect the startup time as well as the energy consumption. This work presents the startup of the distillation column based on economic model predictive control and considers the economics of startup. With the consideration of the dynamic disturbances in the startup of the process, the optimization is evaluated under a probabilistic framework. Case studies on startup control show the effectiveness of the proposed method in terms of economic saving. (C) 2018 Elsevier Ltd. All rights reserved.