Energy Efficiency: Seven Steps On How To Cut Costs And Meet The ‘Copenhagen Agenda’
By Chris Smith, Marketing Director, on365
Unsurprisingly, the Copenhagen summit did not deliver binding targets for reducing carbon emissions.However, the summit has focused governments' and business' attention on energy efficiency issues as never before: so how will economies manage and reduce their energy usage? How can they make a practical difference?
The good news is that this issue is closely related to the continual need for businesses to better manage the overall energy demands of the IT infrastructures that underpin their operations - and Britain's wider service-based economy. The relentless build-up of computing resources needs to be planned and risks managed; technology industry analysts Gartner have already warned that computing rooms or 'data centres' handling business operations will start to suffer power supply difficulties in the next few years.
Many companies including TV stations and high street banks, have suffered embarrassing service outages in the last couple of years because of failure to check and maintain their IT and its less well-understood but vital back-up power systems.
The real issue is the continual growth in high density, high power IT systems based on computer servers that handle an increasing proportion of business and social transactions. Many IT managers are still not 100 per cent certain what they should do to make their data centre and its associated utilities to help achieve optimum system performance. Many companies are making mistakes such as maintaining the incorrect operating temperatures for their business systems or simply failing to maintain effective air flow for cooling around critical computing equipment.
Here are seven practical steps that companies' IT professionals and office managers should take to help manage down their IT energy bills, reduce carbon emissions and enhance IT equipment performance:
Step one: ensure good cable management around servers and other computing equipment.Many IT systems' server rack systems have poor containment of both their power and data cabling which can restrict airflow through the equipment and cause dangerous hot spots. Servers should be regularly inspected, danger areas identified and the use of racks with glass or solid doors - which can also have the same effect - carefully monitored.
Step two: Control airflow around servers. Install appropriate 'blanking plates' in spare server rack space to prevent air mixing at rack-level, which can "short circuit" hot air from back and sides of equipment back into itself or into other systems, destroying the efficiency of the original cooling system.
Step three: Ensure that adding new equipment does not lead to 'hot spots'. Despite the pressure of daily workloads, IT teams should avoid adding more computing equipment on an 'if it fits' basis without reviewing the cooling and power needs. This habit often means new, higher performance equipment, replacing or being added to racks, causing higher than previously experienced power and heat densities and increases in power demand.
Step four: Avoid over-cooling the company's computer room or data centre. IT managers often seek a "margin of safety" overcooling computer spaces to 16 or 17?C, when a maximum temperature of 20-21?C is often acceptable.
This 'overcooling' habit works the room's air conditioning units harder, and can end up in undermining the room's overall cooling infrastructure when it has to cope with peaks in system demand. IT managers must check that their room temperatures are set up for their specific servers' needs, rather than their own perceptions.
Step five: avoid installing more cassette-type air conditioning (ceiling mounted units) because the room is getting hot in areas furthest away from existing cooling. Cassette-type air handlers are designed for comfort cooling of people more than cooling of specific equipment in part of a data centre and are very efficient at mixing cold air with the returning hot air. This makes the unit very inefficient in terms of operational costs - and the addition of more units, making the situation worse. It is better practice to try and separate the hot and cold air to control air flows in the data centre.
Step six: Regularly check that power and cooling systems are in balance across the computing set-up. Seemingly insignificant changes in the data centre can combine to radically reduce airflows or increase the risk of particular server racks drawing excess power at peak load.
Step seven: Ensure your company maintains a budget for equipment cooling when expanding IT systems. Installing urgently needed servers or data storage equipment without planning for improved power or cooling budget is a high risk strategy, especially with today's energy costs being so volatile. We have seen organisations install new servers due to immediate business need, without a budget for cooling, leading to equipment overheating and failure.