Server rooms and data centre environments depend on stable, well-integrated infrastructure. Power, cooling, fire protection, monitoring and access arrangements all need to work together to protect critical IT and network systems.

This guide provides practical planning guidance for UPS power systems, A/B feeds, rack infrastructure, cooling, fire suppression, monitoring and lifecycle maintenance.

IT systems depend on stable power, cooling and connectivity. Downtime can affect business operations, communications, security systems, hosted platforms and operational continuity.

Critical infrastructure in a server room may include UPS systems, batteries, generators, cooling, racks, fire systems, monitoring, access control and the electrical distribution that connects these systems to the load.

Server room power architecture should define how utility supply, UPS systems, generators and critical load distribution boards support the IT load. The design should separate essential and non-essential circuits and include safe maintenance bypass arrangements where required.

Online double conversion and modular UPS systems are commonly used for server rooms because they provide stable conditioned power and support planned runtime during utility failure or generator start.

Runtime planning, battery strings, redundancy, maintenance bypass, load monitoring and generator compatibility should be reviewed before equipment selection.

A/B feeds provide separate power paths to rack equipment. Dual-corded servers can connect one supply to the A path and one to the B path, reducing single points of failure when the upstream design is properly implemented.

IT equipment converts most of its electrical energy into heat. Cooling design should consider heat load, precision cooling or CRAC units, redundancy, airflow management, hot aisle / cold aisle layouts, temperature and humidity monitoring and maintenance access.

Raised floors can support cable management and airflow strategy, but underfloor routes must be planned carefully. Cable trays, segregation of power and data, cable labelling, access panels and maintenance access should be considered early.

Poor cable management can restrict airflow, complicate maintenance and increase operational risk during changes or troubleshooting.

Server rooms often require early warning detection, fire detection interfaces and clean agent or gas suppression systems. Suppression design must coordinate with UPS, generator, electrical isolation and building fire systems.

Specialist fire contractor involvement is recommended because shutdown sequencing, room integrity, detection zones and safety interlocks are project-specific.

Generators support extended outages while the UPS provides ride-through during generator start and transfer. Generator sizing should account for UPS loads, rectifier behaviour, load steps, harmonics and fuel autonomy.

ATS or changeover integration must be coordinated with UPS input requirements, critical load distribution and maintenance procedures.

Monitoring helps identify issues before they become outages. Server room monitoring may include UPS status, battery health, temperature and humidity sensors, leak detection, generator alarms and remote notifications.

If a server room or data centre supports critical business systems, the infrastructure should be reviewed by an experienced Critical Power Specialist before upgrades, migrations or new installations.

This guide is provided for general technical guidance only. Final data centre or server room design, electrical infrastructure, cooling, fire protection, UPS sizing and installation requirements must be confirmed against manufacturer data, site conditions, applicable standards and project-specific requirements.

© 2026 Emerald Infrastructure Solutions (Pty) Ltd. All Rights Reserved.