Sequenced Build-Out of Four-Building Campus Enhances Cost Efficiencies While Adding Logistical Complexity for Project Team

Imagine going into work everyday with the primary purpose of making systems fail.

For the team at Kaiser Permanente’s data center site in Walnut Creek, CA, analyzing each component to understand the multiple potential failures and testing how the system will react played a critical part in the 44,000-sq.-ft., two-story renovation project.

Featuring administrative offices above an operating data center that serves Kaiser’s Northern California locations, the nearly nine-month job included the installation of two Piller rotary systems to the existing Uninterrupted Power Supply (UPS) for increased stability and reliability. The project also consisted of upgrading the 30-year-old building to conform to current IT guidelines, voice/data needs, structural seismic standards and American Disabilities Act (ADA) codes.

“When working in a functioning data center that must remain online 24 hours a day, seven days a week, there is no room for error,” said Bill Helfrich of DPR. “We spend a lot of time planning for every possible operating occurrence—thinking of ways to throw a curve ball into the system that might cause a failure so we can come up with plans to mitigate potential interruptions.”

DPR also called upon its in-house mechanical and electrical experts to study the UPS system and gain a clear understanding of all the various components. According to Helfrich, every data center has its own unique single-line diagram with an electrical path from source to load. With the objective to never interrupt power to the load, the team takes meticulous care to plan each activity before executing work, always being cognizant of potential cause and effect.

In addition to running power through a battery string like the static UPS system, the Piller rotary UPS generates power through an internal generator powered by a spinning flywheel. During a power surge or outage, the inertia created from the flywheel keeps the system running until battery power is activated, allowing the system loads to be impervious to any outside power fluctuations.

“DPR’s knowledge of data centers was especially valuable during preconstruction and the bidding process,” said Bradley Noel, project manager for Kaiser Permanente. “All major subcontractors, especially electrical, were interviewed specifically for their expertise in mission critical work, and a plan of action was established to make sure that all essential medical facility functions remained operational throughout the project.

“DPR’s performance on this contract was excellent—among the top general contractors with whom I’ve worked” added Noel.

Along with the highly technical aspects of the project, another challenge was the structural upgrades, noted Ray Trebino of DPR. To bring the building up to current seismic codes, DPR bolted 80 three-ft. by five-ft. steel plates, each weighing more than 300 pounds, to the concrete walls on the second floor to tie the panels of the existing tilt-up building together.

“We went through an intense amount of pre-task planning to look at all the different options for installing the steel panels and determined that using personnel over machinery was the most efficient choice,” said Trebino. “This required us to lift each panel three feet into the air, stand it up and then utilize a unistrut bracket against the wall while someone drilled and bolted it. We did all 80 panels without a single injury.”

Working in a 30-year-old building created an additional challenge. DPR worked closely with the building engineer to coordinate the shutting off of valves for the water and circuit breakers for the electrical system to achieve critical milestone dates. The team also held a special meeting every Monday to discuss electrical work and coordinate all of the different scopes. “Any impact to the facility was unacceptable, and these meetings assured the project’s success,” said Trebino.

The amount of detailed coordination and communication helped tremendously, especially when a piece of equipment on the project malfunctioned and needed to be repaired, according to Trebino. The team was able to get the equipment installed and retested in less than two weeks, minimizing the impact to the schedule and delivering the interior improvement portion on time in February and the data center upgrades in early March.