Story image

Schneider Electric: Whatcha got under the (data center cooling) hood?

07 Sep 2017

Not a typical question you ask in a data center but maybe you should ask that question the next time you’re in the market for a chiller or computer room air conditioner (CRAC). 

Why? 

Because the compressor is like the engine of the cooling system. It’s kind of like going into a car dealership to buy a car, and not asking what kind of engine it has. 

You may be hauling a heavy trailer, or driving lots of city miles, or not driving much at all. All of these situations call for different engines. 

Like cars, cooling systems use different compressors for different situations. Paul Lin and I recently released a new white paper  titled, The Different Types of Cooling Compressors that discusses these different compressor types.

The paper doesn’t get into the theory of refrigeration or the fine details of how each compressor works. 

But it does provide an overview of each of the main types of compressors you typically see in a data center and a brief explanation of how the compressor works. 

Then for every compressor type, we list their main benefits and limitations, along with the typical applications they best suited for.

If you don’t have time to read the paper, I’ll leave you with some high-level facts about compressor types:

  • There are generally 10 types of compressors, 5 of which are commonly found in data centers.
  • Centrifugal compressors have very high efficiencies, but are best suited for applications over 700 kW.
  • Reciprocating single-acting compressors have the lowest cost / kW but also have the lowest efficiency and most vibration.
  • Variable load compressors, like some digital scroll compressors, don’t vary the speed of the motor, they vary the load while the motor spins at constant speed. Variable speed compressors vary the speed of the motor and are more efficient.
  • Centrifugal compressors can use different energy sources for driving the compression, i.e. electrical motor, steam turbine, or gas turbine.
  • Rotary-scroll, Rotary-vane, and Reciprocating compressors (welded hermetic) provide the smallest capacities. But, some lower-capacity compressors like rotary-scroll can also be used in larger equipment when paralleled to achieve larger capacities, and may only be cost effective up to certain capacity limitations.

I’m in the middle of writing a paper that focuses on the fourth bullet.  This issue, in particular, causes much confusion in the marketplace. 

Simply put, variable loading does not equal variable speed.

Article by Victor Avelar, Schneider Electric Data Center Blog

Dropbox invests in hosting data inside Australia
Global collaboration platform Dropbox has announced it will now host Australian customer files onshore to support its growing base in the country.
Opinion: Meeting the edge computing challenge
Scale Computing's Alan Conboy discusses the importance of edge computing and the imminent challenges that lie ahead.
Alibaba Cloud discusses past and unveils ‘strategic upgrade’
Alibaba Group's Jeff Zhang spoke about the company’s aim to develop into a more technologically inclusive platform.
Protecting data centres from fire – your options
Chubb's Pierre Thorne discusses the countless potential implications of a data centre outage, and how to avoid them.
Opinion: How SD-WAN changes the game for 5G networks
5G/SD-WAN mobile edge computing and network slicing will enable and drive innovative NFV services, according to Kelly Ahuja, CEO, Versa Networks
TYAN unveils new inference-optimised GPU platforms with NVIDIA T4 accelerators
“TYAN servers with NVIDIA T4 GPUs are designed to excel at all accelerated workloads, including machine learning, deep learning, and virtual desktops.”
AMD delivers data center grunt for Google's new game streaming platform
'By combining our gaming DNA and data center technology leadership with a long-standing commitment to open platforms, AMD provides unique technologies and expertise to enable world-class cloud gaming experiences."
Inspur announces AI edge computing server with NVIDIA GPUs
“The dynamic nature and rapid expansion of AI workloads require an adaptive and optimised set of hardware, software and services for developers to utilise as they build their own solutions."