Skip to main content Skip to main content
 

Recapturing Stranded Assets: Case Study of a Resource-Efficient Data Center Retrofit and Expansion

Mark Hydeman, Taylor Engineering, LLC
David Edgar, Lawrence Berkeley National Laboratory

This case study, presented from the perspective of both the designer and owner, describes the evaluation and retrofit of a data center at Lawrence Berkeley National Laboratory (LBNL) in Berkeley, California. The original data center was over 20 years old. Its original design had liquid-cooled servers with a combined overhead and underfloor distribution of central air for the non-liquid-cooled loads. At the time of the retrofit, the entire data center was cooled with air using the original central air system distributed overhead and the bulk of the cooling was from water-cooled DX CRAC units that supplied the underfloor.

At the time of the retrofit, the data center had approximately 240 kilowatts(kW) of high-density scientific computing load and 90kW of back-of-house, low density information technology(IT) load. The facility experienced significant hot spots and needed expansion capacity for an additional 100kW of high density load. Using a draft measurement and assessment protocol developed in conjunction with LBNL, the presenters uncovered a number of measures to improve the efficiency of the mechanical systems and provide capacity for the additional 100kW loads using the existing cooling equipment.

The retrofit of this project started in February 2008. It includes a reconfiguration of the overhead system to underfloor distribution; conversion of the constant flow condenser water system to variable flow; hot aisle containment; a number of control enhancements; and conversion of the overhead plenum to a return plenum for the water-cooled CRAC units. To the extent possible, the speakers will present the post-retrofit results in this session. This session will also present benchmarks (PUE, DCiE and airflow management ratios) of the pre- and post- retrofit performance.

Biographies:

Mark Hydeman, P.E., is a principal at Taylor Engineering, LLC in Alameda, California. Mark has over 20 years of experience in the design and commissioning of high-performance commercial buildings, with an emphasis on data centers, central plants, laboratory buildings, commissioning, and controls. Mark's contributions to building science research and the industry earned him advancement to a fellow in the American Society of Heating Refrigeration and Air-Conditioning Engineers (ASHRAE). Mark is an active Member of ASHRAE's TC 9.9, Mission Critical Facilities. Lawrence Berkeley National Laboratory has enlisted his services to assist in the development and deployment of a toolkit for assessments of existing data centers and evaluation of energy conserving retrofits. His clients include Oracle, Kaiser, Symantec, Stanford University, University of California at Davis, U.S. Department of the Treasury, California Department of General Services, Southland Industries, Dicon Fiberoptics, and Pacific Gas and Electric Company. In addition to his hands on experience, Mark has an extensive list of publications (seven books, 12 technical papers and four articles) and is widely sought as an instructor. For the past 10 years he has been an instructor at for the University of California at Berkeley Extension Service in their HVAC/R certificate program.

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.