Reliability and availability analysis of a hybrid cooling system with water-side economizer in data center

Abstract The cost of data center downtime has increased significantly recently and it is critical for any data center to minimize the unplanned downtime. A reliable cooling system is of great importance to ensure the thermal environment and achieve this goal. The reliability and availability analysis (RAA) is a scientific way to quantify the risk and determine the potential solutions for a system design and operation. Hybrid cooling system with water-side economizer as a promising energy saving method in data center has complicated configuration and multi operation modes. Thus, it is very necessary to conduct the RAA on this novel system and understand it. In this study, the Markov model combined with reliability block diagram method, which can drastically reduce the computation dimension, was developed to evaluate the reliability and operational availability (OA). Functional Availability (FA) was presented as a new measure by integrating OA with specific function criteria such as room temperature. Additionally, we proposed a maximum allowable downtime (MADT) to quantify how long a repair time is acceptable in different scenarios. The detailed method, evaluation procedure and results of a hybrid cooling system were presented.