Take away the intangibles involving in a healthcare facility’s building a new data center versus continuing a piecemeal or stopgap approach to data storage, and the whole scenario can be summed up in a financial model used to evaluate a business decision. A case in point is a large educational institution, with multiple buildings across an urban campus, including a hospital. In this case, the institution had a handful of data centers, with the major hub in the basement of one of its larger facilities. The goal of the institution in this case had been to find space wherever available, potentially based on the least valuable real estate — the basement. Based on current models, this choice would raise a red flag, but at the time, it was the right business decision for the institution.
Later, this basement, which houses the data center’s major hub, flooded. The institution rapidly deployed a backup plan using its other storage locations. Because the data centers were not built as backups but more as parallel storage, the transition was not seamless. But, with a stopgap in place, the institution could now think forward. Should it rebuild in place? Or find other available space to just replace the flooded square footage? Should the institution consolidate its data centers in an existing building extensively renovated to serve as a data center — a “new but existing building” — or begin development of an entirely new facility?
The best answer for the institution in this case was to consolidate its data centers into a new but existing building. This meant finding the appropriate real estate and upgrading it architecturally, structurally, and from an MEP (mechanical, electrical and plumbing) perspective. The location had flaws from the perspective of each discipline, but none fatal.
Architecturally, the new hub was located four stories up to mitigate any flood concerns. However, this presented challenges with deliveries and the transport of equipment to rooms that were not at level with a loading dock. From a security perspective, an elevated data center is more secure, but the existing building had windows and soft spots in the façade that needed to be blocked up and enhanced. Also to be considered were the thermal characteristics of the system.
Structurally, the new location was not built for the capacities to support data center programs, and reinforcing was required. The existing floor system was a cast-in-place, ribbed concrete slab system that framed to steel beams and columns. To increase the capacity, additional steel beams were introduced to shorten the concrete spans and reduce the force demand. The floor-to-floor height was limited, so the depth of the beams had to be minimal. A uniform increase in floor capacity was provided to allow for some flexibility within the spaces. The corridor also had to be reinforced for equipment-rigging paths.
The MEP infrastructure was upgraded to meet the electrical and mechanical needs of a data center. However, the central plant for the institution, as well as the backup generators, was located buildings away. This led to costly conduit routing and infrastructure distribution.
All the necessary design parameters were incorporated, and the facility is a functioning data center. The decision to build this data center in the way it was and the location it is in was not a knee-jerk reaction. The institution took its time and evaluated its options. Clearly, the capital costs associated with the approach finally taken were less than what it would have cost to construct a new built-to-suit data center. However, building new at the very beginning — when the institution began creating data centers and the decision was made to put the hub data center in the basement — would likely have been the most cost-effective strategy, given the costs of each piecemeal data center constructed, the costs to run them, the loss associated with the flooded basement, the costs of the current solution, and the potential costs associated with new space needed in the future.
Stephen Szycher, PE, LEED AP BD+C, is a principal with Thornton Tomasetti. He leads the mission-critical market sector, overseeing project delivery for confidential data and telecommunications centers throughout North America.
Rethinking Healthcare Data Centers: Piecemeal Approach Versus New
Evaluating Data Center Rebuilding Decision with Financial Model
