Case Study: Project Borealis
Updated: 2 days ago
Diverse Paths for Data Centre Fiber Connectivity
A data centre developer came to BAAC in early 2026 with a clear requirement: two physically diverse fiber paths connecting its data centres to a carrier-neutral internet exchange. Before anything was designed or ordered, we ran a feasibility study to find out what the region could actually support, and how quickly. What we found changed the plan for the better.
Project Borealis at a glance | |
Client | Data centre developer (confidential) |
Scope | Dual-path diverse connectivity between data centres and a carrier-neutral internet exchange |
Feasibility started | Q1 2026 |
Design and material procurement | Q4 2026 |
Scheduled delivery | Q4 2027 |
BAAC role | Feasibility and route validation, network design, procurement, and delivery |
The Challenge
Two diverse paths, on a schedule that could not slip.
For a data centre, connectivity is not a nice-to-have. If the fiber is not there when the power and the servers are, the building sits idle. Our client needed dual-path diversity so that a single cut or outage could never take its sites offline, and it needed both paths landing at a carrier-neutral exchange so it would never be locked into one provider.
The harder question was timing. Long-haul fiber depends on permits, land access, and materials that can take many months to secure. The developer needed to know early which routes were realistic, what could delay them, and what had to be ordered before design was even finished.
What the Feasibility Study Found
Four findings that shaped the project before a single conduit went in the ground.
Feasibility is where we earn our keep. By validating routes, suppliers, and approvals up front, we turned unknowns into decisions the client could act on.
A regional shortage of lit capacity
We found that carriers in the region had limited lit wavelength services available. Relying on leased capacity alone would have left the developer exposed on both availability and timing. That finding made a strong case for dedicated dark fiber.
The fastest permitting routes
Not every route moves through approvals at the same speed. We identified the paths with the quickest permitting outlook so early connections could come online sooner, without waiting on the slowest approval in the chain.
Early fiber procurement
Fiber and materials lead times can quietly set a project's real schedule. We established procurement needs early so orders could be placed in step with design, keeping the Q4 2027 delivery date protected.
Existing infrastructure, put back to work
We located existing infrastructure along the corridor that could be repurposed, and secured it for the project. Using what is already in place means less new construction and a faster path to connectivity.
Where the Project Stands
From feasibility to delivery in under two years.
Project Borealis is now moving from study into build. Here is how the work lines up against BAAC's six-stage process.
Q1 2026, Feasibility and Route Validation. Routes, suppliers, permitting paths, and reusable infrastructure assessed.
Q4 2026, Network and Site Design, plus material procurement. Design underway, with fiber ordered early to hold the schedule.
Q4 2027, Construction and Commissioning. Scheduled delivery of both diverse paths.
Because the hard questions were answered first, design and procurement can run side by side rather than one after the other.
Why It Matters
Fiber comes before megawatts.
Data centre timelines are often set by power, but they are won or lost on connectivity. Project Borealis shows what happens when fiber is planned at the start instead of at the end: shortages are spotted early, the fastest approvals are chosen on purpose, materials are secured before they become a bottleneck, and existing assets shorten the build.
Planning a data centre or AI campus and need diverse, carrier-neutral connectivity you can count on? Start with a feasibility study.
Comments