With the amount of power used by data centres set to double by 2020, there are calls from the EU, regulatory bodies and UK government schemes to help limit this. With tough targets to cut carbon emissions by up to 95 per cent by 2050, there has never been a more critical time to take efficiency measures seriously, or risk an uncertain energy future.
One of the schemes driving efficiency is the European Commission’s ‘Code of Conduct for Energy Efficiency in a Data Centre’ – a voluntary scheme to help improve understanding of energy demand within the data centre. Meanwhile, the Department of Energy and Climate Change (DECC) has introduced the mandatory Energy Savings Opportunity Scheme (ESOS) which all large businesses in the UK must adhere to, with the first audits due by December this year.
DECC estimates that if participants in the ESOS scheme reduce energy consumption by an average of just 0.7 per cent, this would reduce their total energy bills by over £250m per year, and save around 3TWh per year of energy. Similarly, the aim of the European Commission’s code is to drive data centre infrastructure efficiency (DCiE) from its current levels of 50 per cent or less at most sites to the 80 per cent mark. But if data centres must now take a proactive approach to energy management, how are they measuring the results?
Measuring UP(S) in the data centre – what’s the CRAC?
Whilst it is easy to update to the latest UPS with the highest efficiency and change or upgrade the cooling system, this is not the key to an efficient data centre. Instead, getting a base line on where the energy is being used should be the starting point. This means monitoring the power used and the key temperature and humidity readings across the data centre. It’s important to note that not everything needs to be monitored – at the outset, managers should identify clear goals – for example, monitoring energy savings / coordinating data / reporting for audits.
The ideal is to install a Data Centre Infrastructure Management (DCIM) monitoring system to oversee the key power users supporting the data centre such as the UPS, computer room air conditioners (CRAC) and computer room air handlers (CRAH). This monitors the performance of the cooling system by looking at the temperature of the air entering the server racks and the return temperature to the CRAC’s. Simple monitoring of just these two parameters enables the data centre manager to firstly ensure that the equipment is actually performing efficiently and that the flow of cooling around the data centre is where it should be and not bypassing the server racks and returning straight to the CRAC’s.
At Riello UPS, we recently launched the Multi Sense power and environmental monitoring software which gives data centre managers an important tool to analyse their environmental performance. Multi Sense offers an events log and graphic display of the main parameters and displays the location and operating status for any given criteria. It also shows run time and maintenance information and detailed real time and historical data reports.
It’s the simple things…
Simple things within the data centre environment can make the biggest difference, such as ensuring that blanking plates remain in place within the server racks to stop bypass airflow. There should also be control on cabling within data centres to ensure there are no airflow pinch points due to legacy cables not being removed and new cabling simply put over the top. It’s the ‘low hanging fruit’ that is simple and cheap to do which can bring the greatest savings and ultimately contribute to our greener economy.