|2008 WORLD SAFETY CONFERENCE & EXPOSITION PREVIEW
How cooling cows with high-volume, low-speed fans led one company to make new inroads in the world of sprinkler system activation.
NFPA Journal®, May/June 2008
By Peter Willse
In the mid 1990s, a dairy farm installed high-volume, low-speed (HV/LS) fans in its barns to cool off the cows, helping to reduce the animals' stress. The farm quickly discovered that the more comfortable, less stressed cows produced more milk. Other farms quickly took note, and fans were installed in other animal barns to keep those animals cool, too.
By the early 2000s, these fans were also being installed regularly in warehouses and distribution centers, manufacturing facilities, sports arenas, aircraft hangers, and large convention centers and auditoriums.
Not just for cooling cows
The HV/LS fans range in size from 10 feet (3 meters) to 24 feet (7.3 meters) in diameter and can have from 4 to 10 fan blades that rotate up to 50 rpm. Some of these blades are shaped like a helicopter blade while others are cupped.
Swiss RE Global Asset Protection Services, now XL Global Asset Protection Services, wanted to know if these fans had any effect on sprinkler activation in warehouses. In particular, the company
wanted to know whether the airflow created by the fan would overtax the sprinkler system during a fire, whether the blades would obstruct the sprinklers, and whether the fan operation would cause a fire to burn undetected and damage more stock.
Swiss Re also wanted to know whether these fans created a problem in manufacturing facilities.
In June 2007, Swiss Re contracted with Underwriters Laboratories to conduct a series of full-scale fire tests to determine if HV/LS fans had an adverse effect on sprinkler activation. The fire tests were conducted at the Underwriters Laboratories' large-scale fire test facility in Northbrook, Illinois. The facility has a 120- by 120- by 54-foot (36.6- by 36.6- by 16.5-meter) test room fitted with a 100- by 100-foot (30.5- by 30.5-meter) movable ceiling. The test room has an exhaust system set to draw at 60,000 cfm (1,600 m³/min) when the first sprinkler is activated, and make-up air is provided through four inlet ducts positioned along the room's walls.
The fan test
A commercially available fan 24 feet (7.3 meters) in diameter was installed 60 inches (1.5 meters) below the movable ceiling. The fan consisted of a motor and hub assembly with four airfoil blades, each approximately 2 feet (0.61 meters) wide near the hub and 6 inches (15.2 centimeters) wide near the tip. At full power, the blades, which were held at a fixed angle, rotated at 48 rpm.
UL conducted three tests. The first used K-11.2, standard response sprinklers on an 8- by 10-foot (2- by 3-meter) spacing and a fan that was operating at half speed. The test used 20 stacks of Group A plastic commodity up to 15 feet (4.6 meters) high, which were ignited directly below the external edge of the fan blade.
The second and third tests used K-5.6, standard response sprinklers on a 10- by 12-foot (3- by 3.7-meter) spacing. These tests were conducted on 20 stacks of a Class II commodity up to 12 feet (3.7 meters) high. The second test was a baseline test to see how many sprinklers operated without the fan in operation. In the third test, the fan was also operating at half speed, and the Class II commodity was ignited directly below the external edge of the fan blade. In all three tests, the clearance between the top of storage and the ceiling was 10 feet (3 meters).
In Test 1, the first sprinkler operated at 3 minutes, 26 seconds after ignition, and the fan was shut off and allowed to coast to a stop. The test was terminated 8 minutes after ignition since the fire grown out of control.
During Test 2, the control test, the first sprinkler operated at 1 minute, 14 seconds after ignition. A total of 21 sprinklers operated between 1 minute, 14 seconds and 3 minutes, 40 seconds. The test was terminated 30 minutes after ignition. The post-test observation of the test array indicated that the fire did not spread to the extremities of the test array.
In Test 3, the first sprinkler operated at 1 minute, 57 seconds after ignition, and the fan was shut off at that time. The test was terminated 30 minutes after ignition. The post-test observation of the test array indicated that the fire had spread to its extremities.
The tests showed that if a fire starts in a warehouse containing Group A plastics and there is a fan over the storage area, the fire could spread out of control if the fan is not interlocked to shut off or interlocked with the sprinkler system water flow. These tests indicate that additional information is needed about which type of detection systems should be used to shut a fan off early in a fire if lower-hazard commodities are an issue, if different sprinkler types are an issue, and if the speed of the fan is an issue. A new Fire Protection Research Foundation project is now underway to address these topics.
Peter Willse is vice-president, director of research of XL GAPS in Hartford, Connecticut.