Research by Morgan Stanley compares conventional cooling to in-row, close-coupled and in-rack

Industry News and Events

Research by Morgan Stanley compares conventional cooling to in-row, close-coupled and in-rack fan solutions
About two years ago, Morgan Stanley commissioned the Syska Hennessy Group to investigate the performance of five different equipment racks it was looking at purchasing for a data center.

The group’s associate partner and PE, Vali Sorell, shared the results of the experiment at the Silicon Valley Leadership Group’s Data Center and Energy Efficiency Summit in October 2009.

The test included a fan-assisted cabinet, two cabinets with different water-chilled inrow coolers and two close-coupled cooling solutions. The testers compared these products to a conventional, open four-post cabinet, used in conjunction with a CRAC – all tested at 8.5kW.

To level the playing field, the researchers used the previous set of ASHRAE guidelines for temperature, humidity and maximum temperature change rate (9o Fahrenheit per hour). ASHRAE has since expanded the envelope.

Sorell focused on the products’ ability to maintain stable temperatures.

The traditional open cabinet, in conjunction with a CRAC, turned out to be most stable. Without airflow hindrances within the cabinet, temperatures fluctuated by less than two degrees. It was also the most energyefficient solution.

Second in stability came the in-row cooled cabinets. One of them, however, was more stable than the other, as it did not contain hot air in the rear of the cabinet.

The solution that contained hot air allowed overall temperature to jump up when chilled water was cut off and took a long time to stabilise.

"It puts the hot air right back to the front," Sorell says. "It had a real hard time trying to maintain some level of stability."

The fan-assisted cabinet and one of the closecoupled solutions were unable to maintain temperatures stable enough to stay within ASHRAE’s maximum change rate.

One of the close-coupled cabinets had fans along the side carriage, and in some cases allowed temperature to change by 10 degrees within two or three minutes. The other closecoupled solution showed variations, but not quite as wide.

The fan-assisted cabinet had a tray of fans at the top. "Despite the constant supply of temperature that we put into the space with the CRAC unit... the cabinet itself was not capable of maintaining uniform temperature from bottom to top. The temperatures were pretty much all over the chart," says Sorell.

In conclusion, Sorell recommended open cabinets in conjunction with large-scale air handlers for greenfield developments and inrow cooled solutions when there is a need to increase density in an existing facility.