During the past two decades, the increasing demand for digital telecommunications, data storage and data processing coupled with simultaneous advances in computer and electronic technology have resulted in a dramatic growth rate in the data center (DC) industry. It has been estimated that almost 2% of US total energy consumption and 1.5% of worlds total power consumption is by DCs. With the fossil fuels and earth’s natural energy sources depleting every day, greater efforts have to be made to save energy and improve efficiencies. As yet, most of the DCs are highly inefficient in energy usage. A significant part of this inherent inefficiency comes from poor design and rudimentary operation of current DCs. Thus, there is an urgent need to optimize the power consumption of DCs. This has led to the advent of modular DCs, newer scalable DC architectures, that reduces cost and increases efficiency by eliminating overdesign and allowing for scalable growth. This concept has been particularly appealing for small businesses who find it difficult to commit to setting up a traditional DC with huge upfront capital investment. However, their adoption and implementation is still limited because of a systematic approach of quickly identifying a module DC design. Considering many different choices for subcomponents, such as cooling systems, enclosures and power systems, this is a non-trivial exercise, especially, considering the complex multiphysics interactions among components that drive system efficiency. For designing such DCs, there is no research available. Therefore, most of the time, the engineers and designers rely on experience, to avoid lengthy elaborate engineering analysis, particularly during the conception stages of a DC deployment project. Here, we are developing a design tool that will not only optimize the design of modular DCs but also make the design process much faster than manually done by engineers. One of the major problem in designing modular DCs is finding optimum placement of the cooling unit to keep the temperature under ASHRAE guidelines (recommended safe temperature threshold). In addition to finding the optimum selection and placement of the cooling units and its auxiliary components, the tool also gives an optimum design for the power connection to the cooling units and IT racks with redundancy. Also, a bill of materials and key performance index (KPI) for those designs are generated by the tool. Overall, this tool in the hands of the bidders or sales representatives can significantly increase their chance of winning the project. / Thesis / Master of Applied Science (MASc)
| Identifer | oai:union.ndltd.org:mcmaster.ca/oai:macsphere.mcmaster.ca:11375/23026 |
| Date | January 2018 |
| Creators | Nayak, Suchitra |
| Contributors | Puri, Ishwar K., Mechanical Engineering |
| Source Sets | McMaster University |
| Language | English |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.002 seconds