We have developed two case studies to review and analyze energy performance of thermal energy storage (TES) systems in commercial buildings. Our case studies considered two partial ice storage systems in Northern California. For each case, we compiled historical data on design, installation, and operation of the TES system; and also metered data. This information was further enhanced by data obtained through interviews with the building owners and operators. The performance and historical data of the TES systems and their components were grouped into issues related to design, installation, operation, and maintenance of the systems. Our analysis indicated that (1) almost all problems related to the operation of TES and non-TES systems could be traced back to the design of the system, and (2) the identified problems were not unique to the TES systems. Judging from the problems related to non-TES components identified in these two case studies, it is reasonable to conclude that conventional systems have as many problems as TES systems. After the initial trouble-shooting period, in a typical summer week for Case 1, the TES system saved 640, 780, and 660 on-peak kW compared to the single chiller, split chiller, and two-chiller systems, respectively, at a cost of additional energy use of 20.5, 37.6, 47.2 MWh per week. For Case 2, the diurnal energy performance of the conventional single chiller system and the TES system were comparable; however, the TES system used less power than the conventional system. Once the performances of the TES with split and two chiller conventional systems were compared, the energy penalties increased substantially and the peak power savings diminished.