When N+2 Meets the Real Building

Most of you will know this already, but in simple terms:

N = what a particular system needs to do its job.
+1 / +2 = the additional capacity or equipment provided for redundancy.

In a data centre, that could apply to power, cooling, pumps, generators or other critical plant.

The notation is straightforward. Delivering that redundancy through the real building - with equipment dimensions, access, maintenance requirements and future replacement all taken into account - is where the engineering becomes more interesting.

On 29 September, an Oireachtas committee heard that conventional data centres typically require between 10 and 25 MW of power, while large AI-focused facilities can exceed 100 MW. Data centres and other new technology loads could account for approximately 35% of Ireland’s electricity requirement by 2035.

That comes as Ireland continues to plan for future electricity capacity and security of supply.

For developers and design teams, that wider conversation reinforces something equally important at project level:

How do you design resilience into a facility from the outset?

For Metec, that question extends well beyond backup generation. It encompasses redundancy, maintainability, accessibility, system isolation and the ability to replace major plant throughout the working life of the building.

Turning redundancy into a physical reality

For Aeron Jordan de Leon, Senior Electrical Engineer and Technical Operations Manager with our Philippines team, reliability, redundancy and continuity of power are central to power-system analysis.

The objective is to establish the redundancy required to minimise downtime and maintain continuity of power.

On a data centre project, that might be expressed as N+1 or N+2.

The engineering challenge is ensuring that the building, plant layout and spatial coordination can actually deliver what has been designed on paper.

Metec has worked on Stage 4 data centre designs where detailed review of equipment dimensions and clearances identified that the physical arrangement could affect the redundancy available within the completed facility. In one example, accommodating the required CRAH units correctly was directly connected to maintaining an N+2 arrangement.

It is a useful reminder that redundancy is as much about physical coordination as calculation.

Space, access and equipment clearances all contribute to the resilience ultimately available to the operator.

Maintainability starts at design stage

The same principle applies to maintenance.

Richard Denver, Metec’s BIM Engineering Manager, puts it simply:

“Just because it fits doesn’t mean it’s buildable.”

A coordinated model needs to consider more than whether services clash. It also needs to establish whether equipment can be accessed, maintained and ultimately replaced.

Gary Quinn, Executive Director and Senior Mechanical Engineer at Metec, describes access and maintenance as key considerations when positioning plant.

That includes allowing the manufacturer-required maintenance zones around equipment while coordinating those requirements intelligently with the available building footprint. In some cases, maintenance spaces can even be shared between adjacent pieces of plant, allowing both operational access and efficient use of space.

For a data centre operator, this becomes part of long-term continuity.

Equipment that can be accessed readily can be maintained efficiently. Components can be isolated and serviced while the wider system continues to operate. Plant can perform the role the original resilience strategy intended for it.

Designing for the full life of the facility

Resilience also extends considerably further than practical completion.

Major plant has a finite working life. Over 10, 15 or 20 years, chillers, pumps, UPS systems and other equipment will require significant maintenance or replacement.

Daniel Lynch, Executive Director and Senior Mechanical Engineer at Metec, recalls an existing data centre where the replacement of ageing chillers became a substantial engineering exercise.

The facility had developed over the intervening years. Buildings had been extended and the original crane access was no longer available.

“It wasn’t designed with this in mind 15 years ago,” Daniel says.

Maintaining continuous operation required a new plant room to be constructed beside the existing building. The new equipment could then be commissioned before the facility transferred from the existing system.

There is a useful lesson in that example for projects being designed today.

Replacement routes can be considered alongside installation routes. Access for cranes and major equipment can be protected. Plant layouts can allow systems to be isolated and upgraded progressively.

That thinking gives future operators more options as the facility evolves.

Resilience is designed in

Ireland’s growing demand for digital infrastructure means electricity capacity will remain an important part of the data centre conversation.

At project level, resilience comes down to a series of much more detailed engineering decisions.

The redundancy shown in the power strategy needs the physical space to support it.

Plant needs sufficient access for maintenance.

Systems need to be capable of isolation and replacement.

And facilities need to accommodate equipment that may still be operating, maintained and replaced many years after the original project team has moved on.

For developers, operators and design teams, considering those requirements early creates a facility that is better equipped to maintain continuity throughout its working life.

That is ultimately what resilient design is intended to achieve.

Reshaping the way you think about building performance.