Modular Data Center Redundancy: Power, Cooling, and Capacity Paths
Modular data center redundancy should be designed around the availability objective, not the number of modules in the layout. The power, cooling, control, and utility paths that feed each module matter as much as the enclosure itself.
Redundancy Applies to the Whole System
A modular data center may contain pre-integrated UPS and cooling equipment, or it may connect to external power and chilled-water systems. In either case, the redundancy claim must describe what happens when a module, feeder, pump, condenser, or control system is taken out of service.
Check the full path: incoming utility, transformer, switchgear, UPS, battery, distribution, cooling unit, water or refrigerant loop, controls, and monitoring. A shared path can become a single point of failure even when the module count appears redundant.
Power Redundancy Options
- N means the exact power capacity required with no spare capacity.
- N+1 adds one UPS module, rectifier, battery string, or distribution path so a single unit can be serviced or fail without shedding the critical load.
- 2N provides two independent power paths, each sized for the full load.
Modular UPS platforms can support several of these configurations by adding modules. Confirm the maintenance strategy, battery runtime, generator handoff, and overload behavior before selecting the number of modules.
Cooling Redundancy Options
For modular cooling, redundancy can be provided by extra cooling units, additional fans or compressors, or independent refrigerant and water loops. N+1 is common for a room or module where one unit can be taken out of service without exceeding the equipment inlet temperature limit.
With chilled water, include the pumps, valves, headers, cooling towers or dry coolers, and their power supplies. If any of these are shared, a failure may affect more than one cooling unit.
Confirm Controls and Monitoring
Redundancy is difficult to manage without visibility into each power and cooling path. The monitoring system should show UPS status, battery condition, cooling alarms, inlet and return temperatures, breaker status, and communication health.
The control system should also define what happens during a failure: which path remains online, how loads are shed, when alarms notify the operations team, and what maintenance actions are required to restore full redundancy.
What to Put in the Specification
- Design IT load in kW and required runtime.
- Power topology and any generator or utility constraints.
- Cooling capacity, airflow, and entering air conditions.
- Required behavior with one module or path out of service.
- Factory test scope and site commissioning acceptance.
- Maintenance and monitoring responsibilities.
常见问题解答
Is a modular data center automatically redundant?
No. Redundancy depends on the selected UPS, cooling, distribution, control, and utility paths. A supplier should explain the configuration and what happens during failure or maintenance.
What is N+1 in a modular data center?
N+1 means the modular power or cooling system has one extra unit or module beyond the capacity required for the design load.
Does 2N require duplicate modules?
A 2N design needs two independent paths, each able to carry the full load. The physical configuration may use separate modules, switchgear, feeders, or site systems.
Need help defining redundancy requirements for a modular data center project? Describe the load, availability target, and site constraints to VERHI.
