Use Sensors and Controls to Match Cooling and Airflow to IT Loads

Environmental instrumentation in the data center

Sensors for environmental variables such as temperature, input power, utilization, air inlet temperatures, and airflow (collectively called “instrumentation”) help managers monitor the data center and spot problems that need to be addressed.  These sensors may be either wired or wireless and is included in your ENERGY STAR certified server, data center storage, or large network equipment product.

One of the benefits of environmental instrumentation on server racks is that it can alert managers when safe operating temperatures are likely to be exceeded.  This information allows the data center manager to adjust cooling, airflow, and/or IT loads in response. 

Even better, some DCIM systems also provide remote monitoring and management of cooling units, and can make cooling adjustments automatically in response to fluctuating heat loads.  This can not only extend the life of IT equipment; it can dramatically lower the cost of cooling the data center.

Matching cooling capacity to heat load

The most efficient cooling systems provide just enough cooling to match the heat load.  This can be a challenge in a data center because cooling units are sized for peak demand, which rarely occurs.  The challenge can be addressed, however, through the use of intelligent cooling controls.  Thanks to environmental instrumentation, these controls are capable of understanding, predicting, and adjusting cooling capacity and airflow based on actual conditions in the data center. 

Intelligent controls make it possible to shift from cooling control based on HVAC return air temperature, to control based on conditions at the actual servers, which is essential to optimizing efficiency.  This often allows temperatures in the cold aisle to be raised closer to the safe operating threshold now recommended by ASHRAE (maximum of 80.5 degrees F), and sometimes even higher, depending on the safe operating temperature range of the server equipment being cooled.   In a nutshell, intelligent sensors and HVAC controls prevent over-cooling to drive greater thermal efficiency, and undercooling which could result in equipment failure.

Cooling teamwork

Centralized controls can also enable individual cooling units to work together as a team, supporting more precise control of airflow based on observed server inlet and return air temperatures.  A centralized system can shift cooling loads to units operating at peak efficiency, and prevent cooling units from working at cross purposes.  Without such a system, a CRAC unit in one area of a data center may add humidity to the room while another unit is extracting humidity from the room at the same time.  A centralized control system offers visibility into conditions throughout the room and the intelligence to match airflow to the heat load – and to determine whether humidification, dehumidification, or no action is required to maintain conditions in the room at target levels.

For supplemental cooling modules that focus on cooling just one or two racks, a centralized control system performs a similar function by shutting down fans based on the supply and return air temperatures, further improving the efficiency of supplemental cooling modules.1

Savings and Costs

  • By allowing you to “right-size” the data center infrastructure, DCIM solutions can reduce overall energy costs by as much as 30 percent.2
  • Wired environmental sensors can cost around $100 each, but wireless sensors are generally less expensive.  Additional costs for the monitoring system software and for system installation vary greatly, depending on the size of the data center, the number and type of sensors deployed, and system features.
  • Lawrence Berkeley Laboratory documented costs and benefits associated with the installation of a wireless sensor network in a 10,000 square foot data center operated by the Franchise Tax Board.  The data center had 12 computer room air handling (CRAH) units and 135kW of load.  Fifty wireless temperature sensors and intelligent control software costs totaled $56,824, but delivered dramatic energy savings: $30,564 in the first year alone.  As a result, the investment payback period was less than 2 years.3
  • Intelligent sensors and HVAC controls often allow data center operators to increase cold aisle temperatures.  One Emerson Network Power study found that a 10 degree increase in cold aisle temperature can generate a 20 percent reduction in cooling system energy consumption.4  Other studies show that data centers can reduce energy expenditures by an average of 4 percent for every degree by which they raise the thermostat.6 

Tips and Considerations

Multiple factors influence the quantity and location of temperature and humidity sensors needed in a data center.  Experienced data center operators typically recommend monitoring temperatures at three points per rack if possible: at the bottom front of the rack to verify the temperature of the cold air arriving to the rack (ideally combined with airflow monitoring); at the top front of the rack to verify if cold air gets to the top of the rack; and finally one at the top back of the rack which is typically the hottest point of the rack.

For additional recommendations on the placement of temperature, humidity, and other sensors in the data center, see Best Practices for Wire-free Environmental Monitoring in the Data Center (PDF, 650 KB), a white paper by 42U.

1 Seven Best Practices for Increasing Efficiency, Availability and Capacity: The Enterprise Data Center Design Guide, a 2012 white paper by Emerson Network Power. (PDF, 1.44 MB)

2 DCIM Yields Return on Investment, by Michael Potts,, June 5, 2012.

3 Better Buildings, Better Data Centers: Applying Best Practices, by Dal Sartor, August 10, 2015. (PDF,  38.5 MB)

4 Seven Best Practices for Increasing Efficiency, Availability and Capacity: The Enterprise Data Center Design Guide, a 2012 white paper by Emerson Network Power. (PDF, 1.44 MB)

5 Why Environmental Monitoring Is Vital to Data Centers, by Tony Santos,, May 7, 2015.