Data center design practices have improved over the years due to industry lessons, equipment design improvements, increased operational reliability, innovations in mechanical system control technologies, and enhanced programming capabilities. However, electrical equipment’s reliability rapidly decreases after 20 to 25 years; parts availability dwindles, compatibility between new and older technology becomes problematic, and failures become more frequent and difficult to repair.
Built in 1978 as an office building for an aerospace company and later converted to a data center, most of the electrical equipment in this facility exceeded its usable life expectancy. As a result, the owner of this colocation facility asked Hood Patterson & Dewar (HP&D) to assess the critical infrastructure systems, including mechanical, electrical, fire alarm, and pre-action fire protection systems.
During the assessment, in addition to the expected aging issues, HP&D discovered several other problems that required attention, including:
- Location of external cooling towers
- Utility equipment conspicuously located above-ground
- Lack of physical equipment barriers
- Data rooms located on upper floors with no buffer to exterior walls
- Improper management of raised floor spaces
- Lack of as-built drawings for equipment power feeds and floor loading
- Leakage of the building envelope
As the building systems were modified during the data center’s life, some components were abandoned or repurposed, resulting in inconsistent equipment labels, inaccurate panel schedules, and confusing one-line diagrams. Critical facilities need proper labeling and documentation to ensure operators can safely maintain systems and recover from a failure or outage as quickly as possible.
Although the mechanical systems provided years of reliable service, many components were approaching or exceeding their useful service life. As mechanical equipment ages, coupled with an increase in data center cooling requirements, environmental conditions in the space gradually decline. Mechanical concerns discovered during the assessment included:
- Corrosion of the piping systems
- Lack of accurate as-built documentation of mechanical and control systems
- Absence of control system sequences of operation details
- Operating procedures not prepared for routine and emergency maintenance
- Missing performance testing records to verify reliable performance for failure scenarios
Maintenance should be regularly scheduled and carried out on time. However, the facility failed to provide records, making it unclear whether the systems received proper maintenance.
When inspecting the building envelope, HP&D examined the front and rear of the building, roof drains, and doorways. The team determined the storefront was watertight with no signs of aging or cracking but noted minor cracking above and around two exterior doors, clogged roof drains, and rusted anchor bolts.
HP&D’s Recommendations
After thoroughly reviewing the electrical and mechanical systems and the building envelope, HP&D submitted a detailed report with recommendations. The assessment recommended replacing the aging electrical and mechanical systems with fully redundant equipment (2N, 2N+1, or 2(N+1)) with diverse power paths, physical separation, and external components protected from the elements. The team also suggested a clear separation between data center critical systems and the non-critical building support systems wherever possible, including distinguishing equipment with color-coded labels.
In addition, HP&D recommended developing and testing detailed operating procedures for routine maintenance and emergency conditions. The report noted multiple points of failure and areas of improvement for the mechanical system, specifically the original CHW plant piping, which was 40 years old. Finally, the assessment recommended developing a strategic plan for the cooling system plant and data halls.
Further recommendations included fortifying the building envelope by sealing any pipe fenestrations with weathertight sealant, monitoring cracking found on the exterior doorways, replacing weather stripping, clearing the roof of vegetation to prevent leakage, and replacing rusted bolts to ensure the pipe system remains secure.
Data centers need a detailed plan to address aging buildings and infrastructure, maintain reliability, and take advantage of new technology. A knowledgeable consultant can assist you in developing a plan that addresses critical areas and identifies procedures and timelines to enhance the life and reliability of your facility.
