EDP Europe is a leading provider of Airflow Management Solutions that are designed to improve the effectiveness and efficiency of data centre cooling systems. In this Guide to Airflow Management we look at what Airflow Management is, why it is important, and how it can be improved.
What is data centre airflow management?
Data centre airflow management is the process of managing the airflow paths of cold air and hot air within the data centre.
Cold air is delivered by CRAC units or air handlers; in raised floor environments via a plenum under the floor tiles. This cold air is delivered via perforated floor grilles to the racks housing the IT equipment and is used to cool the equipment. The running IT equipment generates hot exhaust air which needs to be returned back to the CRAC units so that it can be re-cooled. Efficient data centre airflow management is all about controlling these air paths and preventing, as much as possible, them from mixing, conditioned air being stranded or bypassing the active IT equipment.
Why is management of airflow an important part of data centre management?
Efficient management of airflow plays an important part in overall data centre management. Producing the cold air requires vast amounts of energy so managing airflow as efficiently as possible is critical to optimising power-hungry cooling systems to enable energy cost savings to be made. Many data centres are overcooled for a number of reasons, these include poor design, poor room layout and poor airflow management. Poor airflow management can lead to a number of issues these include bypass airflow, latent cooling, short cycling and hot spots.
- Bypass Airflow – Bypass airflow is deemed as any conditioned air that fails to pass through the IT equipment that it is supposed to be cooling. Common issues that result in bypass airflow include: misplaced floor grilles or unsealed gaps within the raised floor which allow cold air leakage, or gaps within or around the racks where cold air can mix with hot exhaust air. Reducing the amount of bypass airflow and preventing the mixing of hot and cold airflow patterns helps enable changes to be made to the CRAC units that can lead to reductions in energy costs.
- Latent Cooling – If hot exhaust air mixes with cold conditioned air on its return to the cooling units it can reduce the cooling capacity of the cooling units and can cause latent cooling. Latent cooling results in a reduction of humidity within the data centre, which in turn can affect the performance of your servers. Latent cooling can also result in cooling coil condensation, which consumes cooling capacity. The resulting moisture collects in drain pans and exits the data centre further drying out the room. If the room ends up requiring humidification it further increases costs to both power usage and water usage.
- Short Cycling – Short cycling is another form of bypass airflow that results in floor grilles failing to deliver cold air to the location intended. Short cycling, occurs where the speed of the cold air moving beneath the grille produces negative pressure which hinders the air flowing out of the tile, and in some cases, it can even lead to air being sucked back into the plenum, this is known as the Venturi effect.
- Hot Spots – Hot spots within server racks result from areas that fail to be cooled effectively. These can occur anywhere in a rack but are predominantly found in the top third of the rack. Hot spots if left uncorrected can result in poor equipment performance and in severe cases equipment failure.
Use of environmental monitoring equipment can help identify some of these issues, which can then be eliminated by following airflow management best practices. These best practices can help improve equipment cooling and performance, as well as enabling adjustments to cooling units that result in energy cost savings.