Data Centre Chiller & Cooling Plant Automation: Redundant PLCs, Free Cooling & ASHRAE TC 9.9 Compliance
Mission-critical cooling plant automation engineered for UK hyperscale, enterprise, and colocation data centres. MASK designs and commissions dual-redundant hot-standby PLC systems (bumpless switchover under 50 ms), variable primary chiller sequencing, water-side free cooling economizers, CRAH fan differential pressure synchronization, and BACnet/IP DCIM telemetry integration.

High Heat Densities, PUE Scrutiny & Zero-Tolerance Downtime
Modern high-density data centres running AI inference, high-performance computing (HPC), and cloud workloads consume megawatts of power per hall, with server rack densities pushing beyond 30 kW to 80 kW per cabinet. At this thermal density, a total loss of chilled water circulation can result in white space thermal runaway in less than 90 seconds. Operators require fault-tolerant automation that balances continuous cooling reliability with aggressive PUE reduction.
Thermal Runaway & Server Throttling
If chiller staging lags or a single control processor hangs during an electrical bus transfer, supply air temperatures spike immediately. Servers throttle CPU clock rates or initiate thermal emergency shutdowns, breaching tenant SLAs and causing millions in contractual penalties.
Escalating Energy Costs & PUE Penalties
Cooling infrastructure accounts for up to 40% of total data centre facility electrical demand. Static chilled water setpoints, uncontrolled bypass valves, and unoptimized CRAH fan speeds inflate Power Usage Effectiveness (PUE) to 1.4 or higher, violating corporate ESG mandates.
Water Scarcity & Legionella Compliance
Evaporative cooling towers and adiabatic dry coolers face rigorous Water Usage Effectiveness (WUE) scrutiny and strict HSE ACoP L8 Legionella monitoring. Manual blowdown or faulty conductivity sensors risk severe bacterial growth or excessive municipal water waste.
Conventional Commercial BMS vs MASK Industrial Critical Facilities Automation
Why commercial building management systems (BMS) fail in mission-critical Tier III/IV data centre environments.
| System Characteristic | Commercial Building BMS (Standard HVAC) | MASK Critical Facilities Automation Architecture |
|---|---|---|
| Controller Architecture | Single DDC controller per plant. Processor crash halts entire cooling plant. | Dual-Redundant Hot-Standby PLCs with bumpless fiber cross-talk switchover (<50ms). |
| Scan Time & Execution | 1,000 ms to 3,000 ms polling loops. Sluggish valve response during load steps. | Deterministic 10 ms to 50 ms PLC execution; rapid hydraulic balance stabilization. |
| Electrical Transient Survival | DDC controllers re-boot during ATS mains-to-generator transfer (3-5 min outage). | Dual 24V DC battery-buffered UPS supplies keep PLC, network switches, and I/O online 100%. |
| Chiller Plant Optimization | Fixed chilled water setpoints (e.g. 6°C constant) with static staging thresholds. | Dynamic CHWST reset to 12°C-18°C, lift minimization, and COP-based chiller sequencing. |
| Aisle Containment Pressure | Constant volume fan operation or manual technician speed balancing. High fan power. | Closed-loop differential pressure (dP) modulation to fan affinity laws; 50% fan energy cut. |
| DCIM Telemetry Integration | Proprietary protocols or slow serial bridges with frequent polling dropped packets. | Native BACnet/IP, Modbus TCP, SNMPv3, and OPC UA uplinks with sub-second DCIM updates. |
Five Engineering Pillars of Data Centre Cooling Automation
From high-capacity variable primary chiller staging to hot-aisle differential pressure synchronization, our automation architectures guarantee concurrent maintainability and thermal stability:
Fault-Tolerant Dual-Redundant Hot-Standby PLC Architectures
Under Uptime Institute Tier III and Tier IV guidelines, cooling systems must feature concurrent maintainability. MASK engineers dual-redundant hot-standby PLCs (Rockwell ControlLogix 5580 Redundancy or Siemens S7-1500R/H) connected via dedicated high-speed fiber-optic synchronization links. The secondary controller tracks the primary in real time, guaranteeing bumpless failover under 50 milliseconds with zero interruption to chiller compressors, primary pumps, or CRAH units.
Chiller Plant Sequencing, Lift Minimization & Variable Primary Pumping
Fixed chiller staging based purely on return water temperature wastes substantial energy during low-load intervals. MASK engineers advanced variable primary flow (VPF) and primary-secondary decoupler control. Our PLC algorithms calculate real-time thermal load (kW / tons of refrigeration) and compressor lift, staging centrifugal, screw, or magnetic-bearing chillers at their optimal Coefficient of Performance (COP) sweet spot while enforcing minimum evaporator flow limits.
Free Cooling Economizers & Adiabatic Cooler Modulation Controls
Taking full advantage of the UK climate allows data centres to operate in 100% free cooling mode for up to 6,000 hours per year. MASK integrates dual-redundant weather stations measuring ambient dry-bulb and wet-bulb temperatures. When ambient enthalpy permits, modulating control valves transition chilled water through plate-and-frame heat exchangers or adiabatic dry coolers, shutting down mechanical compressors and driving facility PUE below 1.15.
CRAH Fan Differential Pressure Synchronization & ASHRAE TC 9.9
Computer Room Air Handler (CRAH) and CRAC units must respond instantly to IT load variations without generating acoustic resonance or air starved zones. MASK modulates EC fan speeds based on hot aisle / cold aisle differential pressure (dP) sensors (typically maintaining 10 Pa to 25 Pa). In accordance with the fan affinity laws, dropping fan speed by 20% reduces fan electrical power by ~50%, while keeping white space supply air precisely within ASHRAE TC 9.9 recommended envelopes (18°C to 27°C).
DCIM, EPMS & Multi-Protocol Integration (BACnet, Modbus, SNMP)
Data centre operations require seamless visibility across IT, thermal, and electrical domains. MASK builds control panels with native multi-protocol gateways supporting BACnet/IP, Modbus TCP/RTU, SNMPv3, and OPC UA. Real-time cooling data (valve percentages, return temperatures, chiller kW, fan speeds, flow rates, and PUE/WUE metrics) streams continuously to DCIM and SCADA suites without polling bottlenecks.
Chilled Water (CHW) vs Direct Expansion (DX) vs Direct Liquid-to-Chip (DLC)
Depending on rack kilowatt density and enterprise scale, data centres deploy diverse cooling topologies. MASK configures bespoke automation architectures across all three core models:
| Specification Parameter | Central Chilled Water (CHW) Plant | Direct Expansion (DX) Air-Cooled | Direct Liquid-to-Chip (DLC) / CDU |
|---|---|---|---|
| Typical Rack Density Range | 5 kW to 30 kW per Rack (Medium/High) | 2 kW to 10 kW per Rack (Edge/Low) | 30 kW to 100 kW+ per Rack (AI/HPC) |
| Free Cooling Potential | High: Water-side economizers yield 4,000 to 6,000 hrs/yr in UK | Moderate: Air-side economizers limited by ambient filtration/humidity | Maximum: Warm water cooling (32°C to 45°C) enables 100% year-round dry cooling |
| Redundancy Mechanism | N+1 / 2N Chillers & Variable Pumps with dual hot-standby PLCs | Independent refrigerant circuits with standalone packaged controls | N+1 Coolant Distribution Units (CDUs) with secondary loop pressure regulation |
| Primary Control Variable | Chilled water loop differential pressure & supply air temperature | Refrigerant suction/discharge pressure and expansion valve superheat | Secondary loop supply temperature (dew point safety) & manifold dP |
| Standard MASK Architecture | Rockwell ControlLogix Redundant / Siemens S7-1500R + Danfoss VFDs | Siemens S7-1200 / Schneider Modicon with Carel/Dixell OEM gateways | Siemens S7-1500F Redundant + Bronkhorst Coriolis flow metering + leak detection |
Engineered with Industry-Standard Critical Facility Platforms
We engineer control systems complying with BS EN 61439 Form 4 Type 6/7, deployed on the hardware platforms specified by leading hyperscale and colocation operators:
Rockwell ControlLogix Redundancy
ControlLogix 5580 redundant processor pairs with dedicated 1756-RM2 fiber modules, Device Level Ring (DLR) I/O, and PowerFlex 755 VFD chiller pump integration.
Explore Allen-Bradley ServicesSiemens SIMATIC S7-1500R/H
S7-1500R/H fault-tolerant redundant controllers with PROFINET MRP ring architecture, ET 200SP I/O, and Sinamics G120P specialized HVAC/chiller drives.
Explore Siemens ServicesSchneider EcoStruxure & Modicon
Modicon M580 ePAC high-availability systems with native Ethernet backplanes, BACnet IP gateways, and integrated StruxureWare DCIM protocol uplinks.
Explore Schneider ServicesOur 5-Stage Cooling Automation & IST Methodology
Data centre critical facilities demand exhaustive Integrated System Testing (IST) before receiving live IT payloads. MASK executes every project through formal Level 1 to Level 5 commissioning:
Thermal Simulation & FDS
Computational fluid dynamics (CFD) review, hydraulic circuit analysis, functional design specification (FDS), and sequence of operations (SOO) sign-off.
Form 4 Panel Build & FAT
Bespoke Form 4 Type 6/7 segregated enclosures built to BS EN 61439, complete with simulated hot-standby PLC failover factory acceptance testing (FAT).
Loop Checks & SAT (Level 3)
On-site calibration of temperature wells, electromagnetic flow meters, differential pressure sensors, and modulating actuator stroke limits.
Functional Testing (Level 4)
Individual chiller ramp-up, variable primary pump VFD PID loop tuning, free cooling changeover verification, and CRAH aisle dP synchronization.
Integrated System Test (Level 5)
Full electrical power failure simulation (Black Building Test), ATS transfer survival, chiller ride-through validation, and final handover sign-off.
Data Centre Cooling Automation Questions Answered
Technical answers on redundant PLC failover, PUE optimization, and free cooling changeover dynamics.
How do hot-standby redundant PLCs guarantee 100% cooling availability during electrical transients?
How does dynamic chilled water reset lower data centre PUE without violating ASHRAE TC 9.9?
What control strategy is used to transition between free cooling economizers and mechanical chillers?
How do MASK control systems modulate CRAH fans in hot/cold aisle containment setups?
What industrial communication protocols do you integrate with DCIM and BMS suites?
How do you prevent Legionella and scaling in adiabatic cooling water loops?
Explore More Critical Utilities & Infrastructure
Discover our specialized control architectures across water utilities, energy storage, and industrial cooling:
Water & Utilities Automation
WIMES 3.01 Form 4 MCCs, telemetry RTUs, pumping stations, and critical process water automation architectures.
BESS Battery Storage SCADA
Turnkey BESS master control panels, G99 grid compliance, container liquid chiller loops, and DNP3/IEC 61850 substation gateways.
Clean Water Treatment
Rapid gravity filtration (RGF) backwashing, DWI Reg 31 chemical dosing skids, and MCERTS telemetry for water networks.
Thermal Management Guide
Comprehensive formulas for cabinet heat dissipation, CFM airflow sizing, air conditioner BTUs, and filter fan selection.
Protect Critical Uptime with Redundant Data Centre Cooling Automation
Whether designing a new hyperscale cooling plant, upgrading obsolete DDC controllers to hot-standby PLCs, or executing Level 5 Integrated System Testing (IST), MASK delivers mission-critical industrial engineering backed by over 20 years of UK control systems experience.
