Behind every cloud service, digital transaction, and AI application is a physical facility that needs dependable electricity. Servers attract most of the attention, but their operation depends on the electrical infrastructure supplying them.
In its April 2026 report, the International Energy Agency projects that global data center electricity consumption will rise from approximately 485 terawatt-hours in 2025 to 950 terawatt-hours by 2030. This is a projection, rather than a guaranteed outcome, but it highlights the growing importance of power infrastructure.
For owners, engineers, and equipment buyers, the challenge extends beyond securing sufficient power. They also need to consider how that power reaches critical equipment, how faults are contained, and how maintenance can take place.
The Electrical Infrastructure Behind the Servers
A data center’s power system brings together equipment with different responsibilities. Switchgear controls and protects distribution circuits. Uninterruptible power supply systems, or UPS systems, support critical loads during interruptions. Generators can provide longer-duration backup.
These components must operate as a coordinated system. Installing a high-capacity UPS does not, by itself, resolve weaknesses elsewhere in the electrical network.
ABB’s data center reference design illustrates this system approach through separate distribution paths incorporating generators, UPS units, distribution boards, and downstream equipment serving the IT load.
The exact arrangement varies by facility. A small server installation and a large data center will have different capacity requirements, operating priorities, and maintenance strategies.
What the UPS Does
The UPS bridges the period between an interruption in the normal supply and the availability of an alternative source.
Schneider Electric describes the UPS and generator as complementary parts of a backup arrangement: the UPS provides immediate short-duration support, while the generator supports longer interruptions after it becomes available.
Battery runtime therefore needs to match the facility’s operating strategy. It should account for the transfer sequence, the supported load, and the required contingency margin.
Selecting a UPS involves more than choosing a headline capacity. Buyers should establish the critical load, future expansion needs, battery requirements, bypass arrangements, and maintenance expectations before comparing equipment.
Why Distribution Design Matters
Backup power is useful only when it can reach the equipment that needs it.
Distribution boards, protective devices, cables, and downstream power equipment form the connection between the source and the load. Their arrangement influences how the facility responds when equipment fails or a circuit must be taken out of service.
ABB’s reference design uses independent active distribution paths as part of its system-plus-system arrangement. This demonstrates why resilience depends on the distribution architecture as well as the backup equipment.
For project teams, an essential question is: what remains available when a particular component or path is unavailable?
That question should be addressed in the design and operating plan before equipment is ordered.
Understanding N, N+1, and 2N
Redundancy describes the additional capacity or systems provided beyond the basic requirement.
| Configuration | Basic meaning |
|---|---|
| N | Capacity sufficient to support the defined critical load |
| N+1 | The required capacity plus one additional capacity component |
| 2N | Two systems, each capable of supporting the defined load |
Schneider Electric identifies N+1 as a parallel redundant arrangement and 2N as a system-plus-system arrangement. These labels describe an architecture; they do not replace a detailed assessment of its actual behavior.
For example, additional UPS capacity does not automatically establish an independent second distribution path. Project teams must examine the complete arrangement and identify shared dependencies.
Cooling Is Part of the Reliability Discussion
Keeping servers energized is only one part of keeping them operational. Cooling equipment also needs an appropriate continuity strategy.
Schneider Electric highlights that cooling systems may stop during a supply interruption and require time to restart after generator power becomes available. This recovery period becomes particularly relevant as computing density increases.
Electrical and mechanical teams should therefore coordinate their requirements. The design needs to establish which cooling loads require immediate support and which can tolerate a controlled interruption.
What Equipment Buyers Should Clarify
A useful purchasing specification should answer these questions:
- Load: What must be supported initially, and what expansion is planned?
- Architecture: Is the project based on N, N+1, 2N, or another arrangement?
- Interfaces: How will the UPS, generator, and distribution equipment connect and operate together?
- Maintenance: What equipment can be isolated while the required loads remain supplied?
- Documentation: Are drawings, operating sequences, and equipment schedules consistent?
- Verification: What factory tests and site commissioning activities are included, and who is responsible for them?
This checklist is an editorial planning framework. Its purpose is to help buyers identify missing information before evaluating quotations.
Planning for Growth
The IEA’s 2026 analysis identifies constraints involving grid connections and energy equipment supply chains alongside growing data center demand. It also highlights increasingly concentrated power requirements associated with advanced computing.
For project planning, this supports an early discussion about electrical capacity, procurement schedules, and expansion stages.
A practical approach is to define the ultimate requirements while documenting what will be installed in each phase. This helps suppliers understand both the immediate order and the interfaces needed for future development.
Reliability Comes From the Complete System
UPS systems, switchgear, distribution equipment, generators, and cooling infrastructure each contribute to data center operation.
The strongest purchasing decisions begin with a clear understanding of how these components work together. Capacity matters, but so do distribution paths, maintenance arrangements, operating sequences, and verification.
As computing demand grows, coordinated electrical infrastructure will remain central to dependable data center operation.
Sources
- IEA: Key Questions on Energy and AI, 2026
- Schneider Electric: Design Considerations of UPS Systems in Data Centers
- ABB: System Plus System (2N) Electrical Distribution, Data Center Design IEC.
Leave a Reply