Skip to main content Skip to main content
 

Evaluation of Modular Cooling Systems in Data Centers

Michael Ryan, Sun Microsystems
Tenfang Xu, Lawrence Berkeley National Laboratory

This abstract is for a study that evaluated four modular cooling technologies available on the market for data center applications, as part of a research project sponsored by California Energy Commission PIER program. Each of the field evaluations included specific characteristics of the selected modular cooling systems, characteristics of server loads, and measurements for use of evaluating their operation performance under various operating conditions. In general, the actual operating conditions were achieved by varying servers loads (e.g., 100 percent and 50 percent load), server inlet air temperatures (e.g., 68° to 80°F), and supply cooling water temperatures (e.g., 45° up to 65°F). The final case study will include the results for the four individual modular cooling systems:

  • Technical objectives.
  • Technical information on the individual modular cooling systems and servers.
  • Measured parameters.
  • Test procedures and sequence.
  • Data acquisition and compiling process.
  • Performance metrics for modular cooling.

To the extent possible, the case study will include monitoring results from the facility supporting the data center that housed the studies. In combination with the evaluation of energy performance of modular cooling systems and that of the facility systems, estimates of potential energy savings will be discussed and may be drawn from the data analyses in the report.

The project evaluated modular localized cooling solutions provided by vendors for a data center located in Sun Microsystems. These evaluations were to establish efficiency metrics for each selected system. The scope included quantifying energy performance of each modular cooling unit as it corresponds to server load and inlet air temperature. The primary objective of the tests was to demonstrate the effectiveness of localized, modular cooling alternatives for data centers, through performing on-site measurements and evaluation of the system performance.

  • Quantify the heat load that is removed by the modular cooling systems.
  • Quantify the energy required to remove that heat from the data center.
  • Estimate potential energy savings and densification potential of modular systems.
  • With certain assumptions to be made, scale the findings to a large data center.

The following parameters were monitored or measured during the evaluation:

  • Power demand of servers and cooling modules.
  • Actual power demand for servers used in this study.
  • Actual power demand for the cooling modules.
  • Electric power demand for the air and/or liquid cooling modules.
  • Air temperature.
  • Cold inlet air temperature to the server racks.
  • Hot outlet air temperature from the server racks.
  • Server cabinet inlet air temperature and air humidity.
  • Server air cooling module inlet and outlet air temperatures.
  • Outdoor air temperatures (dry-bulb and relative humidity).
  • Data center air temperatures.
  • Pump cooling module entering and leaving chilled water temperatures.
  • Pump cooling module refrigerant line surface temperature.
  • Pump cooling module chilled water flow rate.
  • Pump cooling module entering and leaving chilled water pressure differential.
  • Total cooling tonnage provided by chiller plant.

The following parameters were recorded manually to quantify the magnitude of power demand in the data center and chilled water plant.

  • Total power demand going into all IT equipment.
  • Total power demand to chiller plant.

The following parameters were monitored or measured during the evaluation

  • Power demand of servers and cooling modules.
  • Actual power demand for servers used in this study.
  • Actual power demand for the cooling modules.
  • Electric power demand for the pump cooling module and air cooling module.
  • Air temperature.
  • Cold inlet air temperature to the server racks.

Performance metrics for modular cooling systems:

Energy effectiveness: power demand per cooling or computer load under applicable operating conditions (a range determined by inlet air temperature, server load).

1. Measured total electrical power entering server and rack.
2. Measured total electrical power entering the cooling system.
3. Calculated heat removal through the modular cooling system.
4. Calculated heat removal through water side system where applicable.
5. Total electrical power required to remove thermal load from rack.

Back to Agenda

EPA Home | OARM Home | DOE Home | FEMP Home


This page is no longer updated.
EPA gave I2SL permission to house this page as a historic record of the Labs21 Annual Conference.