Inside a Decommission: The Sprint Norcross Data Center Teardown

by | Sep 3, 2026 | Blog, Data Center Decommissioning

Quantum Technology Sprint Norcross data center teardown case study banner titled "Inside a Decommission: The Sprint Norcross Data Center Teardown" set against a background of blue server racks.

The Sprint Norcross data center teardown was one of Quantum Technology’s largest and most complex decommissioning projects. We handled the full disassembly and removal of a decommissioned data center at 5295 Brook Hollow Parkway in Norcross, Georgia, removing generators, chillers, UPS systems, 240,000 lbs of lead-acid batteries, raised floors, and full electrical infrastructure, and restoring the space in just four weeks. If you have a facility to clear and assets worth recovering, this case study shows what large-scale data center removal actually looks like on the ground, from de-energization through final site restoration.

TL;DR: The Sprint Norcross Teardown at a Glance

 

Why Was the Sprint Norcross Data Center Decommissioned?

The Sprint Norcross data center was decommissioned because a private investor and owner needed to convert a dormant, fully equipped facility back into a marketable property. After Sprint vacated, the building was marketed as a fully equipped data center, including a brief lease to a bit mining company that could not make its lease payments. With the asset sitting idle, the owner contracted Quantum Technology to decommission the equipment, purchase it, raise cash, give the property a cosmetic boost, and re-market it as a corporate campus.

That storyline is more common than it looks. TechTarget notes that companies decommission for reasons including consolidation, expanding cloud presence, and upgrading aging infrastructure. When a tenant leaves or a portfolio shifts, the building still contains millions of dollars’ worth of mechanical and electrical equipment. Somebody has to remove it safely, capture its value, and hand back clean space. 

This is also why decommission-plus-restore work continues to grow. A municipal planning white paper from Wichita and Sedgwick County defines a proper decommissioning plan as including site restoration, proper handling of outdated hardware and materials such as batteries, coolants, and other hazardous substances, and a removal timeline. 

 

What Exactly Was Removed at Sprint Norcross?

Quantum Technology removed the complete mechanical and electrical backbone of the facility, not just a rack of servers. This is the difference between a full facility teardown and simply pulling IT gear. The published asset list from the Sprint Norcross data center included the following:

  • Three Caterpillar 2 MW generators with paralleling gear
  • Two 10,000-gallon diesel fuel tanks holding 6,000 gallons of diesel fuel
  • Four outdoor 400-ton chillers
  • Eight Liebert UPS systems
  • Twenty chilled water CRAC units
  • Twelve RPPs, twelve PDUs, and twelve air handlers with VFD controllers
  • Four battery rooms containing 240,000 lbs of lead-acid batteries, plus ferrous and non-ferrous scrap

In addition to the major MEP equipment, we removed walls, racking, and piping. This scope reflects what a real data center demolition looks like when the goal is a clean, remarketable building rather than a stripped shell.

 

Can You De-Energize and Dismantle Systems Without Taking Down the Whole Building?

Quantum Technology used its own electricians to de-energize the data center systems without disabling building power. That distinction is critical for any site that still needs lights, life safety, and base building services running while the data hall comes apart.

Keeping the building live while killing the data center load is not something you flip a switch for. This project required rewiring the building’s fire alarm and troubleshooting lighting problems in the data center. Controlling hazardous energy is a safety baseline, not an afterthought. OSHA’s lockout/tagout standard, 29 CFR 1910.147, sets out specific procedures for controlling hazardous energy during servicing and maintenance, and electrical isolation is where a decommissioning project either stays safe or goes wrong.

 

What Does “Restore to Handback Condition” Actually Mean?

Restoration means the space is left clean, safe, and ready for the next tenant, not just emptied. At Sprint Norcross, restoration was a defined part of the scope, and the details show what landlord handback really involves.

Removing the twelve air handlers left twelve 30-square-foot holes in the raised floor system. Quantum Technology reconstituted the floor by adding floor structure and filling in with raised floor panels, so the finished surface was continuous and load-bearing again. 

That last point is the tell of a professionally managed job. A crew focused only on hauling out copper does not stop to fix trip hazards. Restoration is where a decommissioning project walkthrough separates a true partner from a scrapper.

 

Can the Vendor Buy the Assets and Offset the Teardown Cost?

Yes. On the Sprint Norcross project, Quantum Technology decommissioned and purchased the assets so the owner could raise cash from equipment that was otherwise sitting idle. That is our core model. 

Generators, switchgear, UPS systems, chillers, and raised floors all carry resale or recovery value. When a vendor can recover that value, the teardown stops being a pure cost line and starts working in your favor. For property owners and REIT teams clearing a data center asset for repurposing, that changes the math on the whole project.

 

How Does a Data Center Teardown Compare to Other Approaches?

Not every vendor does the same job. Here is how an asset-recovery-plus-dismantle model compares to a demolition-only general contractor.

Dimension Asset Recovery + Dismantle  Demolition-Only General Contractor
Primary focus MEP infrastructure removal, rigging, resale, recycling, restoration Tear-out and haul-off
Value recovery Buys and resells generators, UPS systems, chillers, and switchgear Typically none
Electrical de-energization In-house electricians keep the building’s power live Usually subcontracted
Site restoration Raised floor, wall patch and paint, tile replacement Varies by contractor
Environmental handling Zero-landfill policy: ferrous and non-ferrous recycling Varies by contractor
Best fit Full facility teardown with asset value to recover Structural demolition

Data Center Knowledge notes that decommissioning involves removing IT assets and does not typically mean shutting down an entire data center, though it can be part of that process. TechTarget frames decommissioning as a cross-team effort spanning planning, inventory, compliance, and documentation, a broader approach than a demolition-only approach.

 

A Step-By-Step Guide to Dismantling a Data Center

Here is a checklist that applies to any large-scale data center removal, built on the industry framework and mapped to how the Sprint Norcross teardown actually ran. Decomm Systems publishes a seven-phase decommissioning checklist spanning planning through final review, and these steps expand it into field practice.

  1. Scope and plan. Define the asset list, roles, timeline, permits, and insurance before anyone touches a bolt. TechTarget notes that cross-functional teams typically include IT administrators, facilities personnel, security personnel, and project management personnel.
  2. Inventory and document everything. Build a full asset inventory and a disposition plan that flags items to be resold, redeployed, or recycled.
  3. Isolate power safely. De-energize the data center systems in accordance with lockout/tagout discipline. At Norcross, this was done without disabling building power, which meant rewiring the fire alarm and fixing the lighting first.
  4. Drain and secure fluids and hazardous streams. Recover refrigerant from chillers and CRAC units, drain diesel from fuel tanks, and stabilize battery rooms before any equipment moves.
  5. Remove major equipment. Rig and remove generators, UPS systems, chillers, air handlers, PDUs, and RPPs. At Norcross, this meant three 2 MW generators and four 400-ton chillers, among others.
  6. Recover assets and generate records. Redeploy, resell, or recycle each asset and produce the disposition documentation that the ESG and legal teams need.
  7. Strip structure. Remove racking, walls, and piping, and lift raised floor sections where required.
  8. Restore the space. Reconstitute the raised floor, patch penetrations, replace damaged tiles, and repaint, exactly as was done on the Sprint Norcross data center.
  9. Close out. Complete a final walkthrough, archive the documentation, and hand back a clean, marketable building.

 

What Heavy Equipment Is Needed for a Data Center Teardown?

A data center teardown requires rigging and material-handling equipment sized for the loads being moved, plus trained crews to operate it. At Sprint Norcross, the loads were substantial: three Caterpillar 2 MW generators, four 400-ton outdoor chillers, and 240,000 lbs of lead-acid batteries across four battery rooms. Moving that safely requires rigging gear, forklifts, and pallet-handling equipment for battery strings, cranes or hoists for rooftop chillers, and cutting and demolition tools for walls, racking, and piping. Just as important, it takes licensed electricians to de-energize the systems before any of it comes out.

 

How Are the Hazardous Streams Handled, and What Records Are Produced?

Regulated components are handled under specific federal rules, and each one generates a paper trail. The Sprint Norcross site carried three high-liability streams: refrigerant in the chillers and CRAC units, 240,000 lbs of lead-acid batteries, and diesel fuel.

 

For refrigerant, EPA Section 608 restricts the resale of used refrigerant to a new owner unless it has been reclaimed by an EPA-certified reclaimer. The distinction between recover, recycle, and reclaim is precise and matters in your documentation, as the EPA lays out in its definitions of refrigerant handling. Section 608 also carries recordkeeping and reporting requirements for stationary refrigeration activity.

For batteries, EPA’s universal waste rules require that any battery showing leakage, spillage, or damage be contained. EPA also notes that state universal waste programs may add requirements beyond the federal baseline, which matters for multi-state portfolios. Our battery removal services exist to handle exactly the kind of volume Norcross presented.

 

Legacy data centers can also carry clean-agent fire suppression. EPA phased out production and import of virgin halons in the United States on January 1, 1994, driven by the Montreal Protocol. If a site has Halon or FM-200, our Halon and FM-200 recovery service covers it.

 

Why Zero-Landfill Recycling Matters for Your Project

The scale of the problem is why zero-landfill recycling matters. The world generated 62 billion kg of e-waste in 2022, and only 22.3% was documented as formally collected and recycled. The World Health Organization cites the same 22.3% figure and frames e-waste as a public health issue.

 

FAQ

1. Where can we find reviews and case studies for data center decommissioning companies in the US?

Look for vendors that publish specific, verifiable projects rather than vague claims. The Sprint Norcross data center case study documents the exact asset list, the four-week duration, the private client type, and the restoration work performed. A real case study names the equipment, the site, and what was left behind, which is the standard to which any decommissioning partner should be held.

 

2. What does a typical timeline look like for a large data center teardown?

Duration scales with equipment volume, hazardous streams, and restoration scope, not just square footage. Quantum Technology completed the full Sprint Norcross teardown in 4 weeks, including three 2 MW generators, four 400-ton chillers, eight UPS systems, and full site restoration. Some projects also run a phased “pilot light” period; TechTarget notes this can last 30 to 60 days in some cases before final shutdown.

 

3. What reporting should we expect from the site supervisor during a teardown?

Expect asset inventory tracking, disposition records for what is resold, recycled, or reclaimed, and compliance documentation for regulated streams. For refrigerant specifically, EPA Section 608 carries recordkeeping and reporting requirements that should be reflected in the handover pack. A disposition process that generates certificates, as outlined in the Decomm Systems checklist, is what ESG and legal teams will want on file.

 

4. What is included in data center decommissioning versus just removing IT gear?

Full decommissioning covers the mechanical and electrical infrastructure, structural elements, and site restoration, not only the servers. The Quantum Technology data center decommissioning service covers disconnection, removal, rigging, demolition, recycling, and restoration. At Sprint Norcross, that meant generators, chillers, UPS systems, batteries, walls, racking, piping, and a reconstituted raised floor, all in one managed scope.

 

Works Cited

Refrigerant Recovery Machines: The Tool That Prevents Data Center Downtime

by | Jan 13, 2026 | Blog, Data Center Decommissioning

When data center disasters show up in movies, they all look the same.
Everything feels controlled, almost sterile. Then someone notices a warning on a screen. A moment later, smoke crept along the ceiling. Alarms go off. People start running. The servers are overheating, the facility is failing, and chaos takes over.
It makes for a great scene.
Real data centers do not fail like that.
In real data centers, nothing explodes or catches fire. There is no dramatic moment when everyone realizes something is wrong. Instead, problems develop quietly, which makes them even more dangerous.
Temperatures rise slowly. Cooling units run longer than usual. A warning appears, then clears. Everything seems fine, so everyone moves on.
Until the alarms finally sound, the problem has already been building for days or weeks.

 

The Problem Usually Starts Long Before the Outage

 

When a data center experiences downtime, the investigation almost always starts in the same place: Servers, power, and network equipment.
Cooling is often treated as a secondary system, something that either works or doesn’t. If the cooling is running, it is assumed to be fine.
In practice, many cooling failures do not happen suddenly; they start during maintenance.
An upgrade, a routine repair, component replacement, or scheduled service window can be eventful. Once the system is back online and dashboards turn green, it’s easy to assume that everything went smoothly.
Even a small issue during refrigerant handling might have occurred, one that remains hidden until the right conditions reveal the problem.

 

Why Refrigerant Handling Matters in Data Centers

 

Every time a cooling system is serviced, repaired, upgraded, or decommissioned, refrigerant must be removed from the system. This step is unavoidable.
A refrigerant recovery machine enables the safe extraction of refrigerant, stores it in certified recovery cylinders, and prevents its release into the environment or contamination of the system.
When the process is handled correctly, the cooling system can be returned to service as designed. If rushed or done improperly, long-term risks are introduced.

 

What Goes Wrong Without Proper Refrigerant Recovery

 

Improper refrigerant handling does not always cause immediate failure. That is what makes it dangerous.
Common issues include:
  • Loss of refrigerant during maintenance.
  • Moisture is entering closed refrigerant loops.
  • Air trapped inside system lines.
  • Incorrect refrigerant charge during restart.
Each of these issues places stress on compressors and reduces cooling efficiency. The system may appear to function normally at first, especially under light loads.
As demand increases, the margin disappears.

 

The False Confidence of a “Clean” Restart

 

After maintenance, cooling systems often restart without any apparent issues: dashboards return to green, temperatures stabilize, and teams move on.
All of this creates a false sense of confidence.
Cooling systems typically fail under load, not during idle conditions. A traffic spike, a backup, a software deployment, or a seasonal temperature increase can push a compromised system beyond its limits.
When that happens, thermal alarms escalate to that point fast enough that the opportunity for prevention has passed.

 

Why the Server Room Is Rarely the Real Starting Point

 

When servers overheat, everything downstream feels like an IT problem. Systems throttle back. Applications slow down or drop. Virtual machines scramble to move. Customers notice almost immediately.
From the outside, it appears to be a server failure. Inside the facility, the story is often very different.
In many cases, the real issue stems from the cooling work done days or even weeks earlier. A small mistake during maintenance. Refrigerant handling seemed routine at the time. Nothing dramatic enough to raise concerns right away.
Cooling failures tend to be delayed failures. This means it is not immediately apparent, and that delay makes them harder to diagnose, harder to explain, and far more expensive to fix once they surface.
By the time the servers react, the real damage has already been done.

 

Refrigerant Recovery Machines as a Reliability Tool

 

In data center operations, reliability depends on consistency.
Processes must be repeatable, verifiable, and documented. Refrigerant recovery machines support these processes by standardizing how refrigerant is removed and stored during maintenance, ensuring:
  • Clean refrigerant extraction without contamination.
  • Accurate refrigerant preservation for recharging.
  • Stable system pressure during restart.
  • Faster and more predictable recovery times.
Data center reliability is defined by execution.
Every maintenance task is an opportunity to protect system stability or introduce risk. Refrigerant recovery is often treated as a background step, but it directly affects cooling performance, equipment lifespan, and resilience. Ignoring it does not remove the risk. It simply delays the appearance of the consequences.

 

Why This Matters More in Modern Data Centers

 

Modern data centers operate with higher rack densities and tighter thermal thresholds. There is less tolerance for inefficiency.
As cooling architectures become more complex, even small refrigerant-related issues can have amplified effects. What was manageable in older facilities becomes unacceptable in high-density environments.
This makes refrigerant recovery machines a critical part of modern data center infrastructure management.

 

Downtime Prevention Starts Before the Alarms

 

In movies, data center failures are dramatic and immediate. In reality, they are quiet and gradual.
The most damaging outages often begin long before anyone notices a problem.
They begin during routine maintenance. During system restarts. During refrigerant handling, which is often overlooked.
Downtime prevention does not start in the server room.
It starts in the cooling system.
At Quantum Technology, cooling infrastructure is treated as mission-critical, not background equipment.
Our data center services focus on:
  • Reliable cooling system maintenance and modernization.
  • Proper refrigerant recovery and handling practices.
  • Risk reduction during upgrades and decommissioning.
  • Protecting uptime through disciplined infrastructure processes.
Whether you are maintaining existing systems, upgrading cooling infrastructure, or planning a data center decommissioning, the details matter.
Quiet failures are still failures.