Alternative to a flowswitch.... 1

[deltasprk]

...I have a unique situation in an apartment building that my company sprinkled. Every so often the flow switch paddle "trips" sending a false signal to the alarm bell. After discovering that this happened without a leak, a blowout, or a sprinkler head going off we elected to replace the flow switch. Several weeks later, the same thing happened-we then replaced the flow switch again and once again the same thing happened. We then came to the deduction that there were pockets of air bubbles in the system-most likely at the end of the sprinkler line near the last head on the line. We figured that once the domestic pump kicked on at night the surge of water pressure pushed the air bubble to such an extent that the flow switch "thought" that a head had gone off. We noticed that if we bled the last heads on each line just slightly that the air bubbles would escape and dissipate. The problem is that whenever this system is tested and serviced the same thing will continue happening. I suggested that perhaps we eliminate the traditional paddle flow switch for an alarm check but due to cost that may not be equitable. Another idea I possibly had would be to see if there were microscopic pin-hole leaks in the piping that were too small for water to escape as anything other than condensation and that perhaps air is seeping in. Another idea would be to replace the head-the disc or the cap could have a small pin hole. A small enough hole would not necessary make one notice since the water leakage would be akin to condensation. Are there any other solutions out there or would we just have to keep going back to this one particular building and continuing to bleed off the last heads on each line ? Any help would be most beneficial.

 

[ContractorDave]

The advent of flow switches with retard devices has done what the industry pretty much wanted; it eliminated the need for an alarm valve. I generally like to set them at about 20 seconds. If you have 20 seconds worth of flow, you can pretty much assume you have issues. I'm assuming you have a double check valve assembly or at least a check valve arrangement you are sure is holding? I have a test gauge with a stop dial on it that I can set for both highest and lowest pressures so I can monitor my supplies. This has come in handy a number of times indicating supply problems when it wasn't thought of as the likely problem.

 

[LCREP]

Have you tried setting the flow switch to its max of 90 seconds?? I would think this would solve your problem, 90 seconds is a long time.

****************************************
Fire Sprinklers Save Firefighters’ Lives Too!

 

[FFP1]

You most definitely have large air pockets in the system. I have ranted about this problem in previous posts (because NFPA 13 allows the ITC to be placed at the riser.....poor design). If you chart the times of the false alarms and compare the chart to the temperature changes on those nights/mornings, I would venture to guess the false alarms occur when the temperature change decreases. The air pockets become smaller when the temp drops; therefore, water flows into the system to equalize the pressure. Find the highpoints (not necessarily the end of the system) and bleed off as much air as possible......then set the delay on the flow switch so that the alarm activates within 90 seconds of water flow at the most remote point of the system. You might have more than one high point with trapped air. This approach should solve the problem.

Good luck!

 

[DRWeig]

Just another possibility -- our building is in an industrial park in a small municipality that did NOT spend enough money on its water supply infrastructure. Our periodic flow-switch jiggles weren't happening because of air in the line, but because of neighbors starting really big pumps and dropping the inlet pressure by 20-30% for a brief time...

The time delay did the trick.

Good on ya,

Goober Dave

 

[NightME]

could you provide a retard chamber and pressure switch instead of the paddle flow switch?

 

[cdafd]

would go for the 90 second setting to see what happens

also is the sprinkler underground main a seperate line/ tap from the domestic, or is it tapped off the domestic at the building???

what is the staic pressure also??

 

[deltasprk]

Thank you to everybody for their suggestions and ideas. We usually set the flow switch for 90 seconds for just a scenario. In this case however, the flow switch may be set for 60 seconds. We have also looked into seeing if we can put a gauge on each side of the backflow and then set a pressure switch and retard chamber in lieu of the WFS. All in all, a lot of good suggestions which we will look into implementing as part of our troubleshooting process.

 

[FFP1]

Response regarding DRWeig post: The pressure in your system should not decrease as long as the alarm check valve and/or backflow prevention unit is fully functional; regardless of the % of pressure drop in the upstream pressure.

Response to NightME: The provision of one retard chamber and a pressure switch in lieu of the flow switch would most definitely lead to more frequent false fire alarms. The time delay associated with a flow switch is much longer and also easier to control/adjust.

 

[Slowzuki]

As a note, in large systems that have been recently drained, the jockey pump can sometimes operate a long time, often longer than the flow switch delay setting. We have found an excess pressure pump on an alarm valve using an alarm switch with mechanical/pneumatic delay to be a fairly good method. It is impractical to high point vent the whole system.

 

[DavidCR]

I,ve never seen this but I´d would think that an air vent valve located on a high point would eliminate the air as the cause.

I´ve used air vent valves on private mains, but I´m not sure if there is a NFPA 13 issue about a venting device on the sprinkler downstream the sprinkler control valve.