Clean Agent System Utilizing FM-200® and Automatic Sprinkler System

Abstract

Data processing and telecommunication (telco) facilities are commonly protected against fire with a clean agent system, an automatic sprinkler system, or with both a clean agent and an automatic sprinkler system. The degree of protection provided by these systems is vastly different, as is the cost incurred in employing these systems.

Comparison testing of a clean agent system utilizing FM-200® (HFC-227ea) and an automatic sprinkler system was performed in order to illustrate the differences in the protection provided by gaseous clean agent systems and automatic sprinkler systems. The tests were performed in a simulated data processing/telecommunications facility, and examined the performance of the suppression systems on a plastics fire located inside a metal electronic equipment cabinet.

In the clean agent system test, the fire was detected while it was small in size and was extinguished 17 sec after system actuation. Damage to the simulated facility and its contents was limited solely to the dislodging of several ceiling tiles near the clean agent nozzle, and a slight scorching of the cabinet containing the fuel array. The facility could be re-occupied and all systems within the facility, other than those damaged inside the cabinet involved in the fire, would be functional immediately following the extinguishment of the fire. Cleanup of the facility would not be required prior to its use following the fire event, as no residues were produced during extinguishment.

Under the identical fire scenario, the automatic sprinkler system actuated approximately three minutes later than the clean agent system, and the fire was not extinguished by sprinkler system. The sprinkler system did achieve its primary objective of containing the fire to the room of origin and managing the temperatures at the ceiling to prevent structural damage and/or collapse, as the suspended ceiling remained intact, and the fire did not spread beyond the cabinet that contained the fuel array. Damage to the simulated facility was much more extensive than that resulting from the clean agent system test. All ceiling tiles were discolored and the tiles around the sprinkler heads that activated were warped and sagging. The walls were discolored throughout the entire simulated facility and a dark ring of varying intensity and width was observed just below the level of the suspended ceiling. Paper items located within the simulated facility suffered extensive water damage, and the simulated facility floor was covered in water containing large amounts of soot. Immediate reoccupation of the facility and resumption of business would be impossible under these circumstances, as the fire and smoke damage experienced by the facility and its contents would require the replacement of the facility walls, ceiling, and floor, and would likely require the replacement of water-damaged electronic equipment.

The test results are discussed in detail as are the implications of the results in the design of fire protection systems for data processing and telecommunication type facilities.