🖥️ Data Center Pre-Commissioning Checklist: What to Verify Before Go-Live

Why Pre-Commissioning Is Critical for Data Centers

A data center failure after go-live can cost $5,000 to $9,000 per minute in downtime according to Uptime Institute research. The overwhelming majority of data center outages are caused by human error and inadequate commissioning — not equipment failure. A rigorous pre-commissioning program finds problems when fixing them is cheap and the consequence of failure is zero. Skipping or shortcutting commissioning transfers all of that risk to operations.

This checklist covers the complete sequence of pre-commissioning activities from installation verification through integrated systems testing, organized by the phase and system type that engineering teams typically work through.

Data Center Tier Definitions (Uptime Institute)

Tier classification establishes the performance expectations your commissioning program must verify:

• Tier I (Basic Capacity): Single path for power and cooling, no redundancy. 99.671% uptime (28.8 hours downtime/year maximum). Suitable for small businesses with no uptime requirements.
• Tier II (Redundant Capacity Components): Single distribution path with redundant N+1 components. 99.741% uptime (22 hours/year). Adds redundant UPS modules, chillers, and generators.
• Tier III (Concurrently Maintainable): Multiple distribution paths, though only one active at a time. Any component can be taken offline for maintenance without disrupting operations. 99.982% uptime (1.6 hours/year). The standard for enterprise colocation.
• Tier IV (Fault Tolerant): Multiple active distribution paths. A single failure (equipment or distribution) does not affect operations. 99.995% uptime (0.4 hours/year). Required for mission-critical financial, government, and infrastructure operations.

Pre-Commissioning Phases

A well-structured commissioning program proceeds through four phases:

Phase 1: Installation Verification (Cx Phase I)

Verify that all equipment is installed in accordance with contract documents, manufacturer specifications, and applicable codes before energization. This is a "pencils-down" inspection before any power is applied.

• Verify equipment is installed per approved shop drawings
• Check all cable and conduit routing against single-line diagrams and riser diagrams
• Verify all equipment labels, nameplates, and circuit directories are in place
• Confirm all access panels, safety clearances, and code-required working space (NEC 110.26) are met
• Verify all mechanical connections (piping, ductwork) are complete and supported
• Confirm all equipment foundations and seismic restraints are per specifications

Phase 2: Factory Acceptance Testing (FAT)

For major equipment (UPS systems, switchgear, generators, chillers), FAT is performed at the manufacturer's facility before shipment. Key FAT activities:

• Witness UPS functional testing at rated capacity
• Verify generator load acceptance, frequency recovery, and fuel consumption rates
• Test switchgear protective relay settings and trip coordination
• Witness chiller performance testing against design conditions

Phase 3: Site Acceptance Testing (SAT)

Individual system verification after installation, before integrated testing. Each system is tested independently to verify it meets its performance specification.

Phase 4: Integrated Systems Testing (IST)

The most complex and critical phase — verifying that all systems work together correctly, including failure scenarios. IST includes simulated power outages, cooling failures, fire alarm activations, and concurrent maintenance scenarios (for Tier III and IV).

Power Systems Pre-Commissioning Checklist

Utility Service and Main Switchgear

• Verify utility service agreement and metering in place
• Confirm main switchgear short circuit withstand rating vs. available fault current
• Test all protective relay settings per coordination study
• Verify ground fault protection settings
• Perform megohm testing on all feeder cables
• Verify surge protective devices (SPDs) are installed and properly rated
• Test all metering and power monitoring instrumentation

Uninterruptible Power Supply (UPS)

• Verify UPS kVA/kW rating matches design
• Test battery string voltage and impedance testing for all battery modules
• Verify battery room ventilation (hydrogen dissipation per NEC 480)
• Test UPS transfer time (typically <4ms to battery)
• Test static bypass operation and return
• Test maintenance bypass operation and isolation
• Verify UPS output voltage, frequency, and waveform quality
• Test UPS alarms and SNMP/BMS integration
• Perform runtime test at rated load to verify battery autonomy matches specification

Automatic Transfer Switch (ATS)

• Verify source sensing thresholds (voltage drop, frequency deviation)
• Test transfer time from normal to emergency source
• Test retransfer logic and timing
• Verify closed transition or open transition operation matches design intent
• Test bypass-isolation capability (if specified)
• Verify ATS SCADA/BMS monitoring points

Generators

• Verify generator kW/kVA rating, voltage, and frequency
• Test starting sequence and time to rated voltage/frequency (typically <10 seconds)
• Perform load bank test at 100% rated load for minimum 4 hours (some specifications require 8 hours)
• Verify fuel system: tank capacity, day tank operation, fuel transfer pump operation
• Test low fuel alarms and automatic fuel transfer
• Verify wet stacking prevention provisions (load banks during commissioning)
• Test generator paralleling controls (if multi-generator system)
• Verify exhaust system, including CO monitoring if enclosed
• Test remote annunciation and BMS integration

Power Distribution Units (PDUs) and Rack Power

• Verify PDU ratings (kVA, input voltage, output circuit configuration)
• Test all circuit breakers — verify ratings match panel schedule
• Test branch circuit metering and remote monitoring
• Verify dual-corded rack PDU operation (A and B paths)
• Confirm all outlet voltages under load conditions

Cooling Systems Pre-Commissioning Checklist

Chillers and Chilled Water Plant

• Verify chiller kW/ton rating and design chilled water supply/return temperatures
• Test chiller controls, safeties, and alarms
• Flush and clean chilled water system before chiller startup
• Verify chemical treatment program is established
• Test chilled water pump operation, including standby pump auto-start on failure
• Balance chilled water flow to all terminal units (CRAC/CRAH units)
• Test condenser water system, cooling tower, and condenser water pumps
• Verify economizer mode operation (free cooling) if specified

CRAC/CRAH Units

• Verify CRAC (Computer Room Air Conditioning) or CRAH (Computer Room Air Handler) unit kW cooling capacity matches design
• Verify supply air temperature setpoint and control response
• Test fan speed variation (EC fans or VFD-driven fans)
• Verify DX refrigerant charge and superheat settings (CRAC units)
• Test chilled water valve operation and control loop (CRAH units)
• Verify unit alarms, including high return air temperature, low supply air temperature, high humidity, and unit failure
• Test N+1 or N+N redundancy: verify automatic standby unit start on lead unit failure
• Measure and verify supply air volumes (CFM) against design

Airflow Management

• Verify hot aisle/cold aisle arrangement is as designed
• Confirm all blanking panels are installed in racks without equipment
• Verify raised floor plenum is clean and grommets/cutouts are sealed
• Test perforated tile placement and confirm tile airflow volumes match design
• Consider smoke visualization or CFD validation for critical airflow paths

Network and Cabling Pre-Commissioning Checklist

• Verify all cable plant is installed per TIA-568 (copper) and TIA-492 (fiber) standards
• Perform Category 6A or Cat 8 channel testing per TIA-1152-A — verify insertion loss, NEXT, FEXT, return loss, delay skew
• Test all fiber strands: insertion loss, ORL (optical return loss); perform OTDR traces for multimode and single-mode fiber
• Verify all patch panels, ports, and cable labels match as-built drawings
• Test all network switches: port connectivity, VLAN configuration, spanning tree operation
• Verify out-of-band management network (IPMI/iDRAC/iLO paths) is functional
• Test all console server connections to critical infrastructure equipment

Fire Suppression Pre-Commissioning Checklist

• Verify clean agent (FM-200, Novec 1230, CO2, or inergen) system design concentration meets NFPA 2001 or NFPA 12 requirements
• Test smoke detection in pre-action zone: verify proper response and alarm sequencing
• Test abort switches and manual discharge stations
• Verify HVAC shutdown on pre-discharge signal (to prevent dilution of agent)
• Verify door hold-opens release and room sealing under discharge
• Perform room integrity (door fan) test per NFPA 2001 Annex B to verify the enclosure can hold agent concentration for required hold time
• Test monitoring panel integration with BMS and remote annunciation

BMS and DCIM Integration

• Verify all BMS (Building Management System) points for power, cooling, fire, and security systems are mapped and communicating
• Test all alarm points in BMS: verify correct point identification, priority levels, and notification routing
• If DCIM (Data Center Infrastructure Management) software is deployed: verify real-time power, temperature, humidity, and airflow data feeds are populating correctly
• Test remote notification (email, SMS, SNMP traps) for critical alarms
• Verify BMS historian is recording all required data points for trending and reporting

Load Bank Testing

Load bank testing is the process of applying a resistive electrical load (via a load bank device) to verify that power and cooling systems can sustain rated capacity under actual operating conditions. Standard data center commissioning includes:

• Step loading (25% → 50% → 75% → 100%) to verify transient stability of UPS and generator systems
• Sustained 100% load testing (minimum 4 hours) to identify thermal issues in cables, switchgear, and UPS batteries
• Simulated utility outage under full load to verify ATS transfer, generator pickup, and UPS ride-through
• Simulated UPS fault under load to verify bypass operation

ASHRAE Thermal Guidelines

ASHRAE TC 9.9 publishes thermal guidelines for data center equipment, defining recommended and allowable temperature/humidity envelopes:

• Class A1: Servers for controlled data center environments. Recommended inlet: 15–32°C (59–89.6°F). Maximum allowable: 5–35°C (41–95°F).
• Class A2–A4: Progressively wider temperature ranges for equipment deployed in less controlled environments, up to 45°C (113°F) inlet for A4.

Pre-commissioning should verify that supply air temperatures to all server inlets are within the ASHRAE A1 recommended range under design load conditions.

TIA-942 Standard

TIA-942-B (Telecommunications Infrastructure Standard for Data Centers) provides design and construction standards aligned with Uptime Institute Tier classifications. It covers spatial layout, cabling infrastructure, power distribution, and environmental control. Third-party TIA-942 certification is available from accredited certification bodies and is increasingly required by enterprise customers evaluating colocation providers.

Conclusion

Pre-commissioning a data center is a complex, multi-week process that requires coordination between electrical engineers, mechanical engineers, network engineers, the commissioning authority, and equipment manufacturers. The investment in thorough commissioning pays for itself many times over in avoided downtime and reduced troubleshooting burden during operations. Document everything — every test result, every corrective action, every as-found and as-left condition — so that the operations team has a complete baseline from day one.