Commissioning (Cx) is a systematic quality-assurance lifecycle. It ensures that newly installed or upgraded electrical, mechanical, and control systems perform as a cohesive integrated system. It also aligns systems with the Owner’s Project Requirements (OPR) and Basis of Design (BOD).
MCE specializes in independent Data Centre Commissioning services. This critical process verifies the performance, capacity, and failure-recovery paths of mission-critical facilities before they become operational. Commissioning is essential to identify and address infant equipment failures, safeguard capital investments, and ensure the facility operates exactly as intended — facilitating a verified, documented transition to the facility operations team.
Three factors differentiate MCE’s commissioning practice from standard industry delivery:
We operate with zero design or construction conflicts of interest, acting strictly as the owner’s representative to ensure the physical infrastructure aligns with contract documentation and engineering specifications.
We systematically assess and stress-test the facility’s configurations to verify that no single failure or maintenance activity can interrupt operations — and that single points of failure (SPOFs) have been identified and eliminated from the topology. This begins at design review, where MCE specifically targets SPOF identification before construction, and continues through purpose-designed test scripts that demonstrate operational continuity under each failure and maintenance scenario.
By enforcing a strict, gated validation framework early in the project timeline, we identify and remediate multi-vendor control anomalies before they stall late-stage deployment or threaten substantial completion deadlines.
MCE mobilizes at project inception — before design begins. This phase establishes the programmatic foundation for the entire commissioning lifecycle: developing the Owner’s Project Requirements (OPR), reviewing the Basis of Design (BOD), authoring the Commissioning Plan, and confirming load bank sizing against target rack density. Early engagement is not optional — true risk mitigation begins before breaking ground, not after.
MCE conducts an exhaustive independent design review from two perspectives — topology and reliability, and commissionability — assessing the project for:
Before equipment arrives on site, MCE reviews vendor equipment procurement submittals and coordinates Factory Acceptance Testing (FAT) at manufacturers’ facilities to identify and address infant equipment failures before long-lead assets are shipped. During active site assembly, MCE conducts regular field inspections to verify installation quality, structural torque compliance, and adherence to seismic anchoring requirements. MCE tracks comprehensive pre-functional checklists and oversees static, component-level equipment start-ups — ensuring a safe, verified transition from raw hardware to energized systems. All deficiencies are logged in MCE’s programme-level deficiency register and tracked to closure before energization.
Each system is subjected to dynamic loads and performance scripts under precision load-bank strings. MCE witnesses and signs off every functional performance test against the OPR and BOD — normal operation, failure modes, and failover sequences are tested system by system. No system proceeds to IST until all Phase 4 deficiencies are resolved and signed off.
MCE pushes the full facility to its physical limits through Integrated Systems Testing (IST), simulating unmitigated failure scenarios under full dynamic load. Load banks are ramped from 25% through 100% to 110% surge load. Black building scenarios are simulated to prove total system autonomy and resilience without human intervention. Post-testing, MCE provides cross-system transient analysis, a comprehensive final commissioning report, and an operational handover package — including seasonal thermal testing scripts to validate cooling infrastructure under extreme peak winter and summer ambient conditions.
In parallel with the commissioning phases above, MCE tracks execution against the data centre industry standard L1–L5 colour-tagging framework. Each level maps to the project phases above and indicates where in the programme that testing activity begins.
Validating equipment metrics directly on the manufacturer’s factory floor before long-lead assets ship to site — MV switchgear, transformers, UPS systems, generators, and CDUs. Problems identified at L1 cost a fraction of what they cost once equipment is installed and energized.
De-risking site logistics by verifying physical receipt quality, nameplates, and structural anchor compliance. Site verification that equipment has been received undamaged, installed correctly, and that it matches the drawings and specifications.
Managing standalone point-to-point wiring verification, initial component energization, and vendor-led start-ups. MCE witnesses all L3 activities and signs off before any system proceeds to L4.
Subjecting complete standalone systems to dynamic loads and performance scripts under load banks. Every functional scenario — normal operation, failure modes, failover sequences — is tested system by system against the OPR and BOD.
Simulating unmitigated, site-wide failure modes, including black building scenarios to prove total system autonomy and resilience. Load banks ramp from 25% to 100%, then to a 110% surge load. PUE, CDU temperatures, delta-T, flow rates, and generator transfer sequences are recorded continuously throughout the IST run.
The shift toward high-density GPU clusters — where rack loads routinely exceed 60 -100 kW — requires commissioning expertise that goes beyond what traditional air-cooled verification covers. MCE brings specialized commissioning capability to modern liquid-cooling topologies:
Verifying supply and return temperatures, flow rates, and differential pressure parameters across secondary cooling loops to confirm CDU performance at target rack density. CDU commissioning is performed at each load step during IST — not only at partial load — to confirm thermal performance holds at 100% and 110% surge conditions.
Pressure testing and leak verification of all cold plate installations prior to rack energization. Includes contact pressure confirmation, thermal interface validation, and individual loop isolation testing. A single cold plate failure in a live GPU cluster is a total loss event — verification at this level is a prerequisite to energization, not an option.
Hydrostatic pressure testing of all liquid cooling loops — primary and secondary — before filling and commissioning. Includes quick-disconnect verification, manifold isolation testing, and confirmation that all containment systems and drip trays are functional and alarmed. No loop is filled until pressure integrity is confirmed and documented.
Overseeing chemical treatment, secondary loop flushing, and micro-channel filtration to protect cold plates and manifolds from fouling — including particle count verification to cleanliness standards before cold plate energization. Coolant chemistry is verified for pH, conductivity, inhibitor levels, and microbial contamination at commissioning and documented for the owner’s baseline operational record.
Functional testing of all liquid leak detection sensors and alarm sequences throughout the cooling infrastructure. Verifying sensor placement, alarm threshold calibration, and automatic shutoff sequences before any liquid is introduced to the rack environment. Leak detection is commissioned as a standalone system before it is relied upon as a safeguard.
Calibrating control sequences governing the handoff between the primary chilled water loop and direct-to-chip or rear-door heat exchanger (RDHX) networks — verifying setpoint logic, valve sequencing, and cooling capacity allocation under varying load conditions. Interaction between facility water supply, CDU control, and rack-level cooling is tested as an integrated sequence, not system by system in isolation.
Verifying CDU delta-T achieves design targets under actual rack load using load banks at progressive load steps — 25%, 50%, 75%, 100%, and 110% surge. Confirms the cooling infrastructure sustains target rack inlet temperatures at peak and surge conditions. Any thermal drift identified during progressive load stepping is documented and resolved before the IST 24-hour run begins.
We provide objective confirmation that installed systems meet specified performance criteria, completely validating both design intent and day-to-day operational requirements.
By identifying and resolving equipment tolerances and control glitches early, commissioning significantly reduces the likelihood of late-stage operational failures.
Systems are precisely fine-tuned during dynamic load testing to achieve optimal Power Usage Effectiveness (PUE) and energy management.
Detailed commissioning reports provide a complete record of raw procedures, step-transients, and compliance data to support long-term warranty and operational strategies.
MCE maintains a program-level deficiency register throughout construction and testing. Every deficiency is tracked, assigned to a responsible party, and closed before substantial completion.
MCE has commissioned over 50 major data centres globally and reviewed or assessed more than 500 MW of data centre infrastructure. We are PEng-licensed, owner-side, and carry no design or construction stake on any project we commission.
MCE acts as independent commissioning authority on behalf of the owner. Our role is verification of installation and performance against design intent — MCE does not assume design liability or act as engineer of record for the facility.
Download the Critical Incident Response Checklist
Download the Public Procurement Commissioning Toolkit
Download Reliability Assessment Capabilities
Download AI Data Centre Commissioning Qualifications
Download white paper: Mastering Phase Rotation: OESC Compliance, Safety, and Operational Integrity
Engineering Uptime: Mitigating Risk in High-Density GPU Deployments
Get the latest insights and updates — sign up for our newsletter!
