In this paper, we discuss Economic Model Predictive Control (E-MPC) in the context of buildings with active energy storage. In particular, we propose a strategy for the optimal control of building Heating, Ventilation and Air Conditioning (HVAC) systems with chilled water thermal energy storage (TES). Owing to the multiple time scale dynamic behavior of buildings, coupled with the need to account for potentially extended forecasts of disturbances (e.g., weather, energy prices), the implementation of a centralized E-MPC must consider a relatively long prediction horizon. In turn, this results in computational difficulties that impede on real-time implementation. Computational complexity is further increased by the presence of integer decision variables, related to on/off states and operating modes in the HVAC and TES systems. In response to these challenges, we introduce a novel hierarchical E-MPC framework based on (i) establishing the optimal operation of the TES by solving a dynamic scheduling problem in the slow time scale, and (ii) using a control scheme with a shorter horizon in the fast time scale, which addresses objectives related to maintaining the indoor air temperature within comfort bounds at all times during the day. A simulation case study concerning the operation of a TES system at the University of Texas Thermal Facade Laboratory is presented, showing excellent computational and control performance. (C) 2014 Elsevier Ltd. All rights reserved.