Data Hall Clean Agent System Fails the Door Fan Test
An HFC-227ea total flooding system protects a 14,400 ft³ data hall with a suspended ceiling and a raised floor. The stamped design drawings specify a 7% design concentration and a 10-minute hold time, and the flow calculation was accepted by the AHJ. During acceptance testing the contractor performs a door fan test and measures a hold time of 3.5 minutes. The test notes air moving freely through two cable tray penetrations in the wall, an undampered return air grille above the entry door, and a roughly 25 mm gap under the door. The air handler was running during the test. The contractor proposes adding 15% more agent to the dedicated cylinder bank and calling the job complete.
1. Is adding agent an acceptable way to resolve the failed hold time? Explain the reasoning.
Read the worked answer
No. Agent quantity and enclosure leakage solve different problems.
- The flow calculation sizes the discharge so the design concentration is reached within the required discharge time; the door fan test measures how long that concentration stays in the room. Adding agent raises the peak concentration, but it does not change the leak paths.
- Hold time falls roughly with the logarithm of the concentration ratio, so a 15% increase in agent buys very little extra retention against an enclosure this leaky - nowhere near the missing 6.5 minutes.
- Any concentration change must be checked against the agent's NOAEL/LOAEL and the manufacturer's listed limits, plus the cylinder and piping design envelope. Those are engineer-of-record decisions, not field decisions.
- The AHJ accepted a specific design. Silent field changes to that design are not permitted.
The governing failure is enclosure integrity, and it must be fixed at the enclosure, not at the cylinders.
2. List the corrective actions in the order you would take them, and note anything about the test conditions you would question.
Read the worked answer
- Question the test conditions first. Enclosure integrity testing must be done with the room in its protected condition: HVAC shut down, dampers closed, doors closed. With the air handler running, the measured hold time is not representative and may be artificially low. Verify the interlock shuts down the air handler and closes dampers on alarm - if it does not, that is a deficiency in its own right.
- Seal the identified leak paths: cable tray penetrations with an airtight listed sealant/firestop system, a listed damper or sealed closure on the return grille, and door sweeps/gaskets for the undercut. Inspect ceiling tile joints, raised-floor penetrations, and any other unsealed openings found during the test.
- Re-run the door fan test and record the hold time against the 10-minute criterion. Record equivalent leakage area so the improvement is quantified.
- If it still fails, keep reducing leakage (barrier the ceiling void, seal floor grilles, add vestibule). Only if the room cannot physically be sealed does the engineer of record evaluate another approach - for example a reduced protected volume with a physical barrier.
3. What documentation should close this item out?
Read the worked answer
- Door fan test report for both the failed and the passing test: enclosure volume, room temperature, barometric pressure, calculated equivalent leakage area, measured hold time at design concentration, and a description of test conditions (HVAC state, doors, dampers).
- Photographs and a list of every leak path sealed, tied to the locations identified in the test.
- Revised as-built drawings if any barrier, damper, or sealing solution changes the enclosure.
- Written concurrence from the engineer of record and AHJ acceptance of the retest.
- Verification record for the related interfaces - HVAC shutdown, damper closure, predischarge alarm, and time delay - because those functions directly affect retention and were the suspect condition in the first test.
- Updated cylinder tag/maintenance record if any agent quantity was adjusted with approval.