The Four Delta-Ts of the Data Center
Welcome to Keep Your Cool - a series tackling simple cooling optimization strategies for the busy data center operators by former busy data center operator, Gregg Haley.
Last week, I wrote about the Server Delta-T and why you should care about it. As Paul Harvey would say: “Now the rest of the story”. In a raised floor, air cooled data center there are four Delta-Ts that you should pay attention to.
Server Inlet to Server Exhaust
The first key delta-t that I wrote about last week is the delta-T through the server. Server delta-t is the temperature difference between the intake air going into the server and the exhaust air leaving the server. As cold air flows through the server, it will pick up the heat produced by IT equipment, causing the exiting exhaust air to become hotter. It is generally accepted that Server Delta-T should be in the 10 to 20 degree Fahrenheit range. The higher the Delta -T the more efficient the cooling.
2. CRAC Return Air to CRAC Supply Delta-T
Air Handling Units; this includes CRAC, CRAH, and AHU as performing the same task; and generally operate most efficiently when the Delta-T between the Supply side and Return side is about 20 degrees Fahrenheit. Too low a Delta-T indicates the AHU is not efficiently removing heat from the air. An excessively high Delta-T may be the result of overcooling and the waste of energy.
A low Delta-T (such as 10 deg or lower) can be caused by any number of factors that should be checked:
Insufficient heat load - not enough IT equipment to raise the return air temperature;
Overcooling - AHU set points may be too low.
Airflow Imbalance - uneven or improper distribution of the air.
Leakage or Mixing of Airstreams - a mixing of the supply air into the return airflow will decrease Delta-T.
Dirty or Blocked Filters - contaminants captured by the filter can reduce airflow impacting Delta-T as AHU struggles to pass the air through the filter.
Inadequate refrigeration capacity - where the heat load exceeds the cooling capacity of the AHU.
Malfunctioning Components - incorrect readings from temperature and humidity sensors or malfunctioning control systems can lead to improper adjustments of the AHU, impacting Delta T.
Incorrect Sensor Readings - Faulty or incorrectly calibrated temperature sensors can provide inaccurate readings, leading to incorrect adjustments in the cooling system and potentially causing a low Delta T.
Poor Airflow Management - misplaced perforated tiles, unblocked wiring openings. Everything needs to be investigated if low Delta-Ts are being measured.
A high Delta-T (Such as over 25 deg F) can be caused by a number of factors that should be checked.
Airflow Imbalance - Uneven airflow distribution across the coils can lead to temperature variations. Check for blockages, obstructions, or improper damper settings in the air supply and return paths.
Dirty or Blocked Coils - Accumulation of dust, dirt, or other contaminants on the AHU coils reduces heat exchange efficiency. Regular maintenance, including coil cleaning, is essential.
Faulty Dampers or Valves - Malfunctioning dampers or control valves can result in improper regulation of airflow, leading to temperature imbalances.
Incorrect Temperature and Humidity Setpoints - Inaccurate or inappropriate settings for temperature and humidity control can cause the AHU to operate outside the desired range, affecting Delta T.
Insufficient Cooling Capacity - If the data center's cooling system is undersized for the heat load, it may struggle to maintain the required temperature difference across the AHU coils.
Inadequate Air Mixing - Poor air mixing within the data center can lead to localized hot spots and affect the Delta T. Ensure proper design and layout of air distribution systems.
Faulty Sensors or Controls - Incorrect readings from temperature and humidity sensors or malfunctioning control systems can lead to improper adjustments of the AHU, impacting Delta T.
Leakage in Ductwork - Air leaks in the ductwork can result in the loss of conditioned air, affecting the overall performance of the AHU.
Inadequate Return Air - Ensure an adequate supply of return air to the AHU. Insufficient return air can disrupt the balance and contribute to higher Delta T.
Changes in IT Load - Variations in IT equipment load can impact the overall heat generated in the data center. Regularly assess and adjust cooling systems based on changing IT loads.
3. CRAC Supply to Perforated Tile Delta-T
This should be a low numbered temperature differential. Some real life examples that can impact this are:
A perforated tile placed too close to the AHU supply outlet can create the venturi effect where the velocity of air passing across the floor under the perforated tile creates a vacuum effect and draws warmer air from the room through the tile perforations and enters the supply plenum.
In another example, I once took over a data center that was utilizing a glycol loop connected to roof top dry coolers for the heat rejection. The temperature of the glycol running through the piping system was in excess of 100 degrees fahrenheit. The raised floor was only about one foot above the slab. The piping for the Glycol loop ran down the center of the data hall and was uninsulated, effectively putting a radiator into the cold air plenum. I never knew who designed it, but we did insulate the loop, resulting in better Delta-Ts at the aisles where the air had to pass over the piping.
4. Server Exhaust to CRAC Return Air Delta-T
Ideally this should be a very low number as well. The goal is to get the hot air back to the AHU return portal as hot as possible. Examples of how this is compromised is as follows:
Airflow management is critical to this Delta-T. Mixing of supply air to the return airflow will lower the Delta-T and represents wasted energy. In the data center I cited earlier with the uninsulated glycol loop in the supply plenum, the site was using the plenum below the dropped ceiling for the return air, placing grates in the hot aisle ceiling to allow the heat to escape into the plenum. Two different fixes were employed to resolve airflow issues. The first was that there was a structural beam in the plenum that restricted the plenum size to a few inches where the air was trying to pass to the return air of the AHU. The plenum was adequate in size, with the exception of this beam which created an air dam. We constructed a soffit that lowered the ceiling the length of the beam, thereby creating an adequate path for the airflow. The other fix was to put extensions on the top of the racks toward the ceiling so the cold air would not flow over the rack tops into the return grates in the hot aisle, the improvement was significant.
In Conclusion
Airflow management is the key element in efficient cooling in data centers. A holistic approach of a Thermal Survey that includes all the Delta -Ts can provide the necessary information with which to make strategic decisions with regard to improving airflow and cooling efficiency which can reduce expenses while maximizing design capacity. At the end of the day, a good Thermal Survey is a fundamental tool for measuring the four Delta-Ts. You can get the right data and the right insights to figure on what truly needs to be optimized and set the appropriate benchmarks for evaluating your changes. Of course, this can certainly be quite time consuming. You can always consider Purkay Labs to come in and conduct the thermal survey for you. You’ll get a comprehensive evaluation of your thermal environment. You can learn more here: https://www.purkaylabs.com/assessment-service
About the Author
Gregg Haley is a data center and telecommunications executive with more than 30 years of leadership experience. Most recently served as the Senior Director of Data Center Operations - Global for Limelight Networks. Gregg provides data center assessment and optimization reviews showing businesses how to reduce operating expenses by identifying energy conservation opportunities. Through infrastructure optimization energy expenses can be reduced by 10% to 30%.
In addition to Gregg's data center efforts, he has a certification from the Disaster Recovery Institute International (DRII) as Business Continuity Planner. In November of 2005, Gregg was a founding member and Treasurer of the Association of Contingency Planners - Greater Boston Chapter, a non-profit industry association dedicated to the promotion and education of Business Continuity Planning. Gregg had served on the chapter's Board of Directors for the first four years. Gregg is also a past member of the American Society of Industrial Security (ASIS).
Gregg currently serves as the Principal Consultant for Purkay Labs.