On this blog site and elsewhere in data center circles you will routinely see coverage of cooling systems that typically assumes readers have a base level of knowledge about such systems. If you’re new to the data center world, you may find some of the terms and concepts confusing. But there is a solution.
Schneider Electric’s Energy University program has a free course, “Choosing Between Room, Row, & Rack Based Cooling For Data Centers I,” that provides a good primer about cooling systems. In less than an hour you’ll come away with a working knowledge of what room, row and rack-based cooling systems are all about and the pros and cons of each in practice.
The course starts with some background on the evolution of cooling requirements, and how the advent of high-density and variable-density IT equipment drove the need for new approaches to cooling. Whereas once cooling systems placed around the perimeter of a data center could do the job, they become less effective when high- and variable-density equipment is introduced, because such systems throw off far more heat.
You’ll learn about the two key functions of all cooling systems – to provide bulk cooling capacity, and to distribute cool air to the IT loads – and how the different approaches accomplish those goals.
The main difference among them is in how they distribute air to the loads. Indeed, controlling the airflow is the main objective in the different design approaches.
As the course explains, with room-based units, the computer room air handler (CRAH) units that supply the cool air are associated with the room as a whole. In some instances, the air is distributed to IT loads from underneath a raised floor.
You’ll learn about the shortcomings of the room-based approach, including the effect of moves, adds and changes on cooling performance, providing CRAH redundancy, and utilizing the full rated capacity of the CRAH units. The course also explains why data centers of greater than 200 kW should use either hot aisle containment systems (for new data centers) or cold aisle containment (for existing raised floor data centers).
With row-based cooling, the CRAH units are dedicated to specific rows of IT racks. That means the airflow paths are shorter as compared to room-based systems and, thus, more clearly defined. The course goes through the benefits of the row-based approach, such as increased efficiency and improved redundancy.
In a rack-based cooling system the CRAH units are dedicated to specific IT racks and are mounted directly to or within each rack. As a result, airflow paths are even shorter than the row or room approach, and exactly defined. Rack-based cooling is even more efficient than row-based and is ideal for high-density IT applications.
As you’ll learn, such benefits do come at a price, as rack-based systems require a large number of air conditioning devices and associated piping when compared to the other approaches, particularly at lower power density.
Finally, the course explains a hybrid cooling approach in which two or more of the three cooling approaches are used in the same data center. That can be a good option for a data center that has racks with widely varying power densities and for installing high-density racks in a previously low-density data center.
Learn more about the pros and cons of each of these approaches to data center cooling by checking out the free course, “Choosing Between Room, Row, & Rack Based Cooling For Data Centers I.” You’ll find it in the College of Data Centers at Schneider Electric’s Energy University.