AccQ Series · End-to-end certified
Certified accuracy for datacentre cooling
Lab-grade temperature and dewpoint measurement at the rack, the CDU, and the chilled-water loop, connected to your BMS, not instead of it.
Hardware detail lives on the AccQ page .
Why accuracy is the bottleneck
30–40 %
Share of datacentre energy consumed by cooling. Overcooling is the default because operators cannot trust their sensors.
±0.5 °C
Typical accuracy class of BMS/DCIM sensors. On a 2–4 K liquid-loop ΔT that is 25 %+ error in measured heat rejection.
1–2 °C
Setpoint raise typically justified when rack-inlet compliance is measured with instrument-grade confidence. Each degree of raised setpoint can save several percent of cooling energy.
ASHRAE envelopes (18–27 °C; 18–22 °C for high-density AI clusters) are tight enough that sensor error eats the whole operating margin. You cannot optimise what you cannot measure.
The water path: chiller plant to rack cooling
A measurement story at each station along the water, from the plant to the racks. This is data centre cooling monitoring at the points that decide kW of heat rejection, rack inlet temperature, CDU ΔT, and dewpoint.
Station 1
Chiller plant / free cooling
Plant-side supply/return ΔT with matched pairs: true kW/ton, economiser changeover on measurements you can trust.
Station 2
Distribution / secondary loops
Low-ΔT syndrome diagnosis: identify which loops are dragging plant efficiency; balance with evidence instead of guesswork.
Station 3
The facility–TCS boundary (CDU / heat exchanger)
The headline measurement: matched-pair ΔT on both sides of the exchanger gives true kW of heat rejection per CDU, rear-door HX, or loop, and catches fouling and approach-temperature drift early. At the 2–4 K ΔTs these loops run, ±0.5 °C pairs are 25 %+ error. Matched pairs are what make this number real.
Station 4
Rack cooling connection
Direct-to-chip cold plates and rear-door exchangers: TCS supply/return per row or rack cluster. Verify each rack solution receives water at design temperature and returns the expected heat.
Station 5
Everywhere along the path: dewpoint
Certified dewpoint at the rack and on cold-water surfaces lets you run facility water warmer (better chiller COP, more free-cooling hours) while staying provably above condensation risk on pipes, manifolds, and cold plates.
Air side
- Rack inlet compliance with certified confidence: enforce ASHRAE A1/H1 without guard-banding for sensor error.
- Matched-pair ΔT across containment: leakage, short-circuiting, failed server fans.
- CRAC/CRAH supply/return ΔT for staging and degradation trending.
The separate loop
AccQ sits alongside the existing BMS/DCIM over Modbus TCP. MQTTS, API and webhooks feed the control plane, the second loop. BACnet is on the roadmap and is not a shipping protocol.
Step 1
BMS commands as normal
The BMS/DCIM sends a start/setpoint command to AccQ over Modbus TCP, exactly as it would to any field device.
Step 2
AccQ executes and confirms
The equipment responds; status returns to the BMS. Sequences of operations, alarms, and compliance logic stay untouched.
Step 3
The control plane optimises in parallel
A second loop trims fan speeds, valve positions, or setpoints against certified measurements. The BMS does not need reprogramming and never loses authority.
Works with, not instead of
Compatible with existing BMS/DCIM (Niagara, EcoStruxure, and others) and with third-party cooling optimisation platforms. See the AccQ unit for hardware, channel counts, and the protocol list.
Compliance & evidence
Colo SLA evidence
Traceable, calibrated temperature records for tenant SLA compliance and dispute resolution.
PUE / ESG reporting
Defensible cooling-energy accounting from certified thermal measurement. Thermal energy metering stays credible even at low ΔT.
Measure what your cooling is actually doing
Pilot deployments alongside your existing BMS: no rip-and-replace, no downtime.