Future-Proofing Legacy Data Centres: Designing for Continuous Upgrade

As computing technology evolves, legacy data centres must evolve with it. The challenge is not simply upgrading infrastructure, but designing facilities that can accommodate continuous change while maintaining operational resilience.

Introduction

The rapid growth of artificial intelligence, cloud computing and high-density processing is placing unprecedented demands on data centre infrastructure. While much of the industry's attention is focused on new-build campuses, thousands of existing facilities continue to underpin the digital economy.

Many of these legacy data centres were designed for technologies, power densities and cooling requirements that differ significantly from today's operational demands. Yet replacing them entirely is often neither practical nor sustainable.

Instead, operators are increasingly investing in phased upgrades, infrastructure renewal and targeted refurbishment to extend the life of existing assets.

The challenge is ensuring these improvements can be delivered while maintaining continuous operation—a consideration that demands careful planning, intelligent sequencing and a design approach focused on adaptability rather than permanence.

Legacy Buildings in a Rapidly Changing Industry

Unlike many commercial buildings, data centres rarely remain static throughout their operational life.

Over the course of several decades, almost every major system is likely to be upgraded, expanded or replaced. Server technology evolves, power demands increase, cooling strategies change and security requirements continue to develop.

As a result, the building itself becomes a framework for continuous adaptation rather than a finished product.

Designing for change is no longer a desirable feature—it is a fundamental requirement.

Designing for Progressive Infrastructure Upgrades

Many legacy facilities were designed around significantly lower power densities than those required by modern workloads.

Infrastructure that once comfortably supported conventional IT environments may no longer provide sufficient capacity for AI-enabled computing.

Examples include:

• Air conditioning and cooling systems reaching operational limits.

• Chillers requiring increased capacity or improved efficiency.

• Electrical distribution requiring expansion.

• UPS systems reaching the end of their service life.

• Additional generators and battery energy storage.

• New liquid cooling infrastructure for high-density racks.

Rather than replacing every system simultaneously, successful operators often adopt a programme of progressive upgrades, allowing infrastructure to evolve incrementally while maintaining operational resilience.

Phasing Without Interrupting Operations

Perhaps the greatest challenge is that most legacy data centres cannot simply be switched off during refurbishment.

Critical infrastructure must remain operational while upgrades take place around it.

This requires carefully planned phasing strategies, including:

• Temporary cooling installations.

• Temporary electrical supplies.

• Sequential replacement of plant.

• Isolated work zones.

• Commissioning before changeover.

• Planned migration of critical loads.

• Maintaining redundancy throughout construction.

Every stage must be coordinated to minimise operational risk while ensuring continuity of service.

Removal Before Replacement

Replacing critical infrastructure is often more complex than installing it.

Large chillers, generators, transformers and air handling units may have been installed during the original construction sequence, with little consideration for future removal.

Successful refurbishment therefore considers:

• Replacement routes for major plant.

• Crane access and lifting strategies.

• Roof loading and temporary works.

• Structural alterations where required.

• Safe segregation between construction and operational areas.

Good design anticipates not only installation but eventual replacement.

Designing for Maintenance and Future Expansion

A resilient data centre is one that can continue adapting throughout its lifecycle.

Architectural design should therefore safeguard opportunities for future development through:

• Dedicated replacement routes.

• Spare plant areas.

• Expandable utility corridors.

• Structural capacity for future rooftop equipment.

• Additional risers and service zones.

• Flexible layouts that accommodate changing cooling technologies.

These measures allow facilities to respond to evolving operational requirements without major disruption or wholesale reconstruction.

The Architect's Role

Future-proofing a legacy data centre extends well beyond the specification of mechanical and electrical systems.

Architects play a central role in coordinating the relationship between building fabric, structural constraints, operational requirements and future infrastructure upgrades.

By considering logistics, maintenance access, phased construction, spatial flexibility and long-term adaptability from the earliest stages of design, architects help create facilities capable of evolving alongside rapidly changing technologies.

The objective is not simply to deliver a successful refurbishment, but to establish a resilient framework that supports decades of continuous improvement.

Conclusion

The most successful legacy data centres are not those that remain unchanged—they are those designed to evolve.

As computing technologies continue to advance, operators will need to increase cooling capacity, upgrade electrical infrastructure, replace ageing plant and introduce new systems, all while maintaining the resilience that their customers depend upon.

Designing for continuous upgrade requires more than technical expertise. It demands a holistic understanding of phasing, logistics, maintenance, operational continuity and long-term adaptability.

By embedding these principles into refurbishment and upgrade programmes, legacy data centres can continue to meet the demands of an increasingly digital world without compromising performance, resilience or sustainability.

---

GTA Perspective

At Gebler Tooth Architects, we understand that complex infrastructure is never static. Through decades of delivering phased projects within live airport environments, we have developed expertise in coordinating upgrades where operational continuity is essential. The same principles apply to legacy data centres: successful projects are defined not only by the quality of the finished facility, but by how effectively it can evolve throughout its operational life. Designing for continuous upgrade is ultimately about creating infrastructure that remains resilient, adaptable and ready for the next generation of technological change.