This paper presents a new framework for congestion management utilizing a reliability model of demand response resources under the smart grid environment. Demand response resources modeling is highly appertaining to uncertainty of customer's behavior. Interval data which is a core deliverable of advanced metering infrastructure, is essential for customers to participate in demand response events. Hence, a systematic method based upon frequency and duration approaches is utilized to present the multi-state model of multiple demand response resources considering repairable advanced metering infrastructures, the so-called demand response firm. Moreover, a new two-step congestion management structure with proposed demand response firm model is introduced for relieving congestion besides diminishing the risk of supplying loads. Firstly, system operator clears the electricity market based on economic and/or reliability-driven issues without considering transmission network limits. Afterwards, he/she will alleviate the existing transmission congestion considering a trade-off between demand response firms and load shedding besides generation rescheduling. In this regard, the impact of several important factors such as demand response firms' maximum achievable potential and forced outage rate of advanced metering infrastructures on the proposed framework are assessed. The simulation results discuss economic- and reliability-driven measures of the proposed demand response model in congestion management. (C) 2016 Elsevier Ltd. All rights reserved.