Protection in a data centre has two layers: detect the fire very early and remove the electrical source, and stop a fire that grows anyway while managing the water damage risk. Neither layer substitutes for the other.
Detection: why sensitivity is critical
An electronic fire releases invisible particulate long before flaming. Point smoke detectors usually miss that stage because the high air change rate dilutes the smoke and the hot aisle / cold aisle arrangement disperses it quickly.
Aspirating (sampling pipe) detection is therefore preferred: air is drawn through a pipe network and analysed centrally at very low thresholds. What to check on site:
- Are the sampling holes positioned relative to the air handling return flow?
- Is the pipe network balanced, and has transport time from the most remote hole been measured?
- Is filter maintenance being done — a dirty filter reduces sensitivity?
- Are the alarm thresholds staged (pre-alarm to alarm to release)?
- Are the underfloor void and ceiling void monitored separately?
The relationship between gaseous suppression and pre-action
The correct arrangement triggers the two systems from separate detection logic. The gas operates early and at high sensitivity. The pre-action system delivers water only once the fire has grown and a sprinkler has opened. If both are fed from the same line, the pipe fills the moment the gas discharges and the water damage risk is taken for no reason.
Do not skip the door fan test. A gaseous system works only if the agent concentration is retained in the enclosure. Cable penetrations, underfloor connections and door undercuts all leak. The door fan test measures that tightness and verifies the hold time; it must be repeated after any refurbishment.
Pressure relief
Agent discharge produces a sudden pressure rise in the enclosure. If the pressure relief damper is not correctly sized, structural damage and door deformation can result. Check not only that the damper is present but that it discharges into a safe space.
Eight defects seen often
- Aspirating detection sampling holes positioned wrongly relative to the airflow.
- A door fan test never done, or not repeated after refurbishment.
- A pressure relief damper missing or locked shut.
- Gas and pre-action triggered from the same detection zone.
- Cable congestion under the raised floor with no detection there.
- Cable penetrations not fire-stopped.
- The battery (UPS) room without separate compartmentation and suitable ventilation.
- No defined post-discharge ventilation scenario.
Battery and UPS rooms
Lead-acid batteries accumulate hydrogen; lithium-ion batteries carry a thermal runaway risk. The two need different measures: ventilation and gas detection for the first, separation, cooling capacity and explosion control for the second. The battery room should be separated from the main data hall by fire-resisting construction and have its own detection and suppression solution.
Cable voids and vertical shafts
Cable is the most widespread combustible in a data centre. Check the cable fire performance class, fire-stopping of floor penetrations in vertical shafts, that cable density does not exceed the design assumption, and that unused cable is removed. Abandoned cable accumulating over years both raises the fire load and disrupts airflow.
Test regime
| Test | Purpose |
|---|---|
| Aspirating detection smoke test | Transport time from the most remote hole, and sensitivity |
| Door fan test | Enclosure tightness and agent hold time |
| Release chain test | Detection to warning to shutdown to discharge logic |
| Pre-action valve test | Valve opening and water transit time |
| Ventilation scenario | Shutdown during discharge, extraction afterwards |
Frequently Asked Questions
Why is aspirating detection preferred in a data centre?
The high air change rate dilutes smoke and the hot/cold aisle arrangement disperses it quickly. An aspirating system draws air through a pipe network and analyses it at very low thresholds, catching the pre-flaming stage.
Can gaseous suppression and pre-action share a detection line?
No. They need separate detection logic. Sharing a line fills the pipe the moment the gas discharges, taking the water damage risk for no reason.
Why is a door fan test needed?
A gaseous system works only if the concentration is retained. Cable penetrations and door undercuts leak; the door fan test measures tightness, verifies the hold time and must be repeated after refurbishment.
Can the battery room be protected with the main hall?
No. Lead-acid batteries accumulate hydrogen and lithium-ion batteries carry a runaway risk. The battery room needs separate compartmentation, suitable ventilation and its own detection and suppression.
SprinkCalc — Fire Sprinkler Design Across Three Standards
SprinkCalc covers hazard classification, design density and area, K-factor selection, water demand and hydraulic calculations for NFPA 13, FM Global and BS EN 12845 in a single iOS app, and exports a professional PDF report.
Download SprinkCalc on the App Store
MEP Calc — 86+ Engineering Calculators
MEP Calc bundles 86+ engineering modules in one iOS app: 21 fire calculations plus heating, cooling, HVAC, plumbing, steam and natural gas.
Download MEP Calc on the App StoreNFPA 13 (2025) · NFPA 25 · NFPA 75 · BS EN 12845:2015+A1:2019 · FM Global DS 8-1, DS 8-9, DS 5-32. The findings here are typical defect patterns defined in the standards and survey guidance, not an account of events at any particular site.