RISK ASSESSMENT AND MITIGATION OF TELECOM EQUIPMENT UNDER FREE AIR COOLING CONDITIONS

Loading...
Thumbnail Image

Files

Publication or External Link

Date

2012

Authors

Citation

DRUM DOI

Abstract

In recent years, about 40% of the total energy is devoted to the cooling infrastructures in data centers. One way to save energy is free air cooling (FAC), which utilizes the outside air as the primary cooling medium, instead of air conditioning, to reduce the energy consumption to cool the data centers. Despite the energy saving, the implementation of free air cooling will change the operating environment, which may adversely affect the performance and reliability of telecom equipment.

This thesis reviews the challenges and risks posed by free air cooling. The increased temperature, uncontrolled humidity, and possible contamination may cause some failure mechanisms, e.g., Conductive anodic filament (CAF) and corrosion, to be more active. If the local temperatures of some hot spots go beyond their recommended operating conditions (RoC), the performances of the equipment may be affected.

In this thesis, a methodology is proposed to identify the impact of free air cooling on telecom equipment performance. It uses the performance variations under traditional air condition (A/C) to create a baseline, and compares the performance variation under variable temperature and humidity representing FAC with the baseline. This method can help data centers determine an appropriate operating environment based on the service requirements, when FAC is implemented.  In addition, a statics-based approach is also developed to identify the appropriate metric for the performance variations comparison. It is the first study focusing on the impact of FAC on the telecom equipment performance.

 This thesis also proposes a multi-stage (design, test, and operation) approach to mitigate the reliability risks of telecom equipment under free air cooling conditions. Specifically, a prognostics-based approach is proposed to mitigate the reliability risks at operation stage, and a case study is presented to show the implementation process. This approach needn't interrupt data center services and doesn't consume additional useful life of telecom equipment. It allows the implementation of FAC in data centers which were not originally designed for this cooling method.

Notes

Rights