The present study proposes a new methodology that estimates the acceptable injection rate of a hydrate inhibitor considering the random failure and subsequent maintenance of the topside system followed by the cooldown of the flowlines and risers where hydrate can potentially form. The proposed methodology consists of four steps: system description, topside system simulation, subsea system simulation, and construction of the exceedance curve for the inhibitor injection. As the system is defined in the first step, Monte Carlo simulation (MCS) is performed in the topside system simulation step to predict the system failure frequency and subsequent downtime that leads to the exceedance curve of the downtime. With the exceedance curve of the downtime, the longest downtime, equivalently the longest cooldown time, tau(LC), is estimated. In the subsea system simulation step, the cooldown path of the inventory is obtained by multiphase simulation and tau(TD), the time that the inventory touch down the hydrate formation region is defined. The last step combines the MCS with multiphase simulation to yield the frequency of the required MEG injection to avoid the hydrate formation. The exceedance curve provides risk-based information for determining the required MEG injection rate based on risk acceptance criteria, which may vary with different points of view. Four case studies are performed for several installations that have different extent of process configurations and insulation. (C) 2016 Elsevier Ltd. All rights reserved.