Duct Detector Sensitivity Testing

[Defenseman7]

Hopefully I have posted this in the right place:

In the last few days we have had 2 separate incidents of an AHCA inspector citing facilities for no records of Duct Detector Sensitivity testing. He references NFPA 72 Table 14.4.2.2(g)(1). I see in (g)(6) that it says "in addition to the testing required in(g)(1)"...

Obviously we did not previously interpret this in the same manner he did, so my question is: How many of you out there are performing these tests and can you provide insight into this?

As many have, we tend to follow the herd when it comes to these types of things. I myself spent many years in the field and cannot recollect ever having to do this. So if nothing else comes of this inquiry, other than me looking like a newbie, perhaps it will be educational to someone in a similar situation.

Many thanks to anyone who can clarify this code and provide positive input.

 

[dhalleron]

I was testing sensitivity on them in the early 90's. The testor we used would not fit in the duct housing, so we set up a test jig and removed the duct head and installed in our jig to do the test.

Most modern addressable panels can do the test for you and you just need a way to print a report from the panel.

If you don't have a sensitivity tester, you will find there are only a few on the market that can test multiple brands of detectors and they are very expensive in the $5000 to $6000 range or more. Some detectors such as System Sensor i3 series can be sensitivity tested using about a $100 or so tester.

 

[nhfire77]

Hopefully I have posted this in the right place:

In the last few days we have had 2 separate incidents of an AHCA inspector citing facilities for no records of Duct Detector Sensitivity testing. He references NFPA 72 Table 14.4.2.2(g)(1). I see in (g)(6) that it says "in addition to the testing required in(g)(1)"...

Obviously we did not previously interpret this in the same manner he did, so my question is: How many of you out there are performing these tests and can you provide insight into this?

As many have, we tend to follow the herd when it comes to these types of things. I myself spent many years in the field and cannot recollect ever having to do this. So if nothing else comes of this inquiry, other than me looking like a newbie, perhaps it will be educational to someone in a similar situation.

Many thanks to anyone who can clarify this code and provide positive input.

If its an older system sensor there is a device that you connect to the smoke head and a meter to test it.

Now if you're talking abou just a few odd brand duct smokes you cant directly test, it would be easier to swap it out with a newer model. Interestingly enough the newer system sensor D4120 no longer opens its trouble relay when dirty or has drifted out of compliance. Just the amber LED blinks, they expect the issue to be caught during the biannual visual inspection. (as if that normally happens).


Also, you are required to perform a pressure differential test annually as well.



As previously stated, an addressable smoke doesn't need the sensitivity test, but still needs the pressure differential test. I understand this is almost never done, but it's supposed to be.

 

[MisterCMK]

As previously stated, an addressable smoke doesn't need the sensitivity test, but still needs the pressure differential test. I understand this is almost never done, but it's supposed to be.

The pressure differential test? Are you talking about testing the airflow when you install and place the detector?

 

[plumb bob]

What about the old fashioned way, put the system on stand-by, open the test port and break out a can of smoke? Maybe i'm just making this too simple!

 

[GoldDigger]

What about the old fashioned way, put the system on stand-by, open the test port and break out a can of smoke? Maybe i'm just making this too simple!

Well, maybe the sensitivity, as opposed to simple go-no-go, testing would require a calibrated airflow during the test combined with a calibrated squirt of calibrated smoke distributed uniformly across the duct?

 

[plumb bob]

Well, maybe the sensitivity, as opposed to simple go-no-go, testing would require a calibrated airflow during the test combined with a calibrated squirt of calibrated smoke distributed uniformly across the duct?

Just as I suspected, I was making it too simple. I'm curious, what is the uniformed across the duct calibrated amount of smoke during an actual fire event?

 

[GoldDigger]

Just as I suspected, I was making it too simple. I'm curious, what is the uniformed across the duct calibrated amount of smoke during an actual fire event?

I guess you would have to ask that question to the detector manufacturer who specifies the sensitivity test parameters :angel:

 

[plumb bob]

A common occurance out in the field, I'm sure. :lol: I really don't know how you can test a smoke device without actually smoking it. Maybe there truly is a way to test sensativity but it's over my pay grade. I agree that it would be easy to alarm a weak duct detector or smoke head by super saturating it with canned smoke.

 

[gar]

140425-2132 EDT

Somewhat over 50 years ago when I was involved in the development of a reflective type optical smoke detector we used a closed test chamber with circulation into which we introduced gray smoke from punk sticks. In the chamber was an obscuration sensor that provided the calibration reference measurement.

A reflective optical sensors produces a grossly different result for black vs gray smoke as compared to an obscuration sensor.

When ionization type detectors are compared, then one has to be concerned with dead vs active smoke.

In UL tests our sensor always beat ionization detectors on dead smoke and were about comparable on live gray smoke.

A reflective or obscuration type optical detector could be tested with an inserted probe into the optical path instead of using actual smoke.

I have no idea what the present standards are for field testing smoke detectors.

.

 

[nhfire77]

The pressure differential test? Are you talking about testing the airflow when you install and place the detector?

That test (pressure differential) ensures the sampling tubes " will properly sample the airstream in the duct using a method acceptable to manufacture or in accordance with their published instructions" -table 14.4.3.2 of 2013 72.

This is an annually required test. System sensor (and all Honeywell brands) require a manometer to do this.


This is in addition to sensitivity testing, which after two consecutive acceptable tests, may be extended to a 5 year cycle or every other year.


Gar-

the standard is still light gray smoke 0.5-4.0% obscuration, or manufacturers listed and marked range.

 

[nhfire77]

Just as I suspected, I was making it too simple. I'm curious, what is the uniformed across the duct calibrated amount of smoke during an actual fire event?

NFPA 72 has sensitivity at 0.5-4.0% obscuration with light gray smoke. The manufacturer has to mark the device with it's listed range, that is determined when it's listed.


To properly test, you need to follow the manufacturer instructions and NFPA72. FYI-remote test switches are there to perform an alarm test. A functional test is also required. 72 states magnet tests are not acceptable smoke entry tests either.

Properly using smoke can satisfy both alarm and smoke entry/response. This functional test ensures smoke is capable of entering the sensing chamber. A test switch does not satisfy this, although it does need to be tested as well.


When I test a duct smoke I do the pressure differential at the unit, then smoke entry, clean it out, reset it at the remote test switch, then alarm it at the test switch, reset it and confirm the alarm at the panel. I technically you should reset it, then test the alarm function at the test switch separately. Any auxiliary functions, such as fan shutdown or damper closure must also be confirmed.

I recognize this is rarely done. There are certain companies that don't want to pay for it, so we remote test switch test only and note the lack of a smoke entry test as a deficiency. Some AHJ's make them do it, others don't care.