HKIE Professional Assessment - Fire Discipline study materials
HKIE Professional Assessment - Fire Discipline study resources

Clean-agent control philosophy and room integrity in a data hall

HKIE Professional Assessment - Fire Discipline · Put your knowledge to work in context.

All case studies

Clean-agent control philosophy and room integrity in a data hall

A new 400 m² data hall (20 m × 20 m × 3.2 m, 1,280 m³) is protected by an IG-541 total-flooding system designed to 40% concentration at 20 °C, with a 10-minute hold time. The room has a raised floor of 600 mm, a suspended ceiling, 14 CRAC units, cable penetrations, and a 1.2 m² pressure-relief damper. Fire detection uses two cross-zoned photoelectric smoke detectors per 20 m², but the building operator has suffered six false alarms in three months from construction dust. The operator proposes to (a) remove cross-zoning so a single detector discharges, (b) reduce the discharge delay from 30 s to 0 s, and (c) keep the room doors open for cable pulling during commissioning. The insurer requires a hold time of at least 10 minutes and an integrity test. The contractor's door-fan test reports 0.35 m² equivalent leakage area at 50 Pa and a predicted hold time of 4 minutes.

1. Evaluate the operator's three proposals against clean-agent system design and control philosophy. What would you accept, reject, or modify?

Read the worked answer

(a) Remove cross-zoning - reject. Cross-zoning is the main defence against a single false detector discharging agent into an occupied data hall. Six false alarms from construction dust are a detector-selection and housekeeping problem, not a reason to remove the safeguard. Use multi-criteria or addressable detectors, duct detectors in CRAC units, better filters, and a proper cause-and-effect matrix. Retain two-detector coincidence, preferably from different detector types or zones.

(b) Reduce the pre-discharge delay from 30 s to 0 s - reject. The delay is for occupant warning and evacuation. Removing it creates a serious life-safety risk. Keep at least 30 s, with abort stations, sounders and clear signage. If the room is normally unoccupied, a shorter delay could be considered only with a documented risk assessment and insurer agreement, not as a default.

(c) Keep doors open during commissioning - reject as a normal operating condition. Doors must be closed for the system to work. Fit door-position switches and interlock them so that agent release cannot occur with doors open, or isolate the agent under a formal permit-to-work during cable pulling. Post warning signs and train staff.

Also check that 40% IG-541 gives an oxygen concentration of about 12.6%, and confirm that this is acceptable for the occupied and egress scenario and for the insurer. The design must include pressure relief, discharge nozzles, cylinder location and re-entry procedures.

2. The door-fan test predicts a 4-minute hold time. Explain why this is a problem, how you would diagnose and remedy it, and what acceptance criteria you would apply.

Read the worked answer

The 4-minute predicted hold time is below the specified 10 minutes, so the enclosure as tested does not meet the design intent. A leakage area of 0.35 m² at 50 Pa is large for an IG-541 system; the agent will leak out through the raised floor, suspended ceiling, cable penetrations, CRAC ducts and the pressure-relief damper.

Diagnosis and remedy:

  • Repeat the door-fan test with all doors closed, the damper in its normal position and the CRAC units sealed or running as in the fire scenario.
  • Use a smoke pencil or thermal anemometer to find leakage paths at the raised-floor perimeter, ceiling tiles, cable trays, pipe penetrations and duct dampers.
  • Seal penetrations with approved firestop and grommets; seal the raised floor and ceiling voids; check that the pressure-relief damper opens at the correct pressure and closes again.
  • Consider a dedicated pressure-relief path sized for the agent discharge, and ensure it does not short-circuit the agent.

Acceptance criteria:

  • A documented room-integrity test at the design concentration, with a hold time of at least 10 minutes (or the insurer's or authority's requirement) above the minimum design concentration.
  • A discharge test or validated calculation showing concentration distribution and oxygen levels.
  • A cause-and-effect test, door-switch test and abort and restart test.

If sealing cannot achieve the hold time, do not simply accept 4 minutes. Re-design the enclosure, reduce the leakage, or change the agent or system with insurer and authority agreement.