The team at Integrated Design Group was discussing the pros and cons of thermal storage after reading about Google’s Taiwan data center. The technology, when used appropriately, can reduce costs by running air conditioning systems at night when power is cheaper. We have seen cases where thermal storage both works and does not.
A team headed up by Wayne Drooks, senior associate and head of mechanical engineering, reviewed a thermal storage option for a financial services client in the Northeast. In this case, we found little payback for the client. Too much space was required for storage systems (ice or water) and a VERY large separate chilled water plant was required to produce ice at night. This chilled water plant and/or another plant has to support the data center at night since load at night can be as high as during the day. Therefore, significantly higher cooling capacity must be installed – space, equipment and installation costs must be considered.
Ken Gill, senior mechanical engineer in Dallas says, “Installations used in the 90’s under time of day utility rates were mothballed when the rate structure for demand charges changed. This happened to schools and commercial thermal storage installations in North Texas.”
Despite these findings, we understand that chilled water thermal storage can be used to keep high density spaces from overheating during the restart time required for chillers after a power failure and transition to backup power. This type of system is typically part of the chilled water system and requires no additional installed capacity.
Integrated Design Group Chief Technology Advisor, Mark Monroe, sums it up by saying, “It depends on the difference between peak and off-peak power cost, and on demand charges. If you can shift load away from $0.15/kWh electricity and use $0.04/kWh instead, you might be able to make a system pay off.”
Are you considering thermal storage in your data center? Tell us about it! firstname.lastname@example.org.