[Discussion] Smoke Alarm Sensitivity Testing

[Scott Wytosick]

I am curious where this area will divert in the future. Testing smoke alarms with smoke guarantees that the device activates as desires. Although, residue is left behind. Alternative testing (manufactures seem to like) is activation of device via magnet. Problems have arose that devices wont activate with smoke but they will with a magnet. My curiosity is, isn't the sensitivity test being done every 5 years picking up that loss of range?

 

[cda]

First you are talking about fire alarm system smoke detectors

Have to see how nfpa 25 reads as far as type of testing required

Because there are some systems that can self test

Seems like the sensitivity testing should pick up problems, but possibly not the shameless fire alarm company doing it every-time, so no history

I prefer smoke testing, but do not demand it

 

[pyrguy]

Saw a system tested one time. There was so much smoke used trying to get the duct detectors to trip that the room smokes were going off.

 

[FM William Burns]

The jurisdictional party responsible for verification of fire alarm system alarm initiating device testing for occupancies there in Holland should be requiring reporting verifications from system service providers. In these reports, should be areas indicating sensitivity testing results and needs by detector if the service provider is a reputable one in western Michigan. Are you a fire inspector?

NFPA 72, 2010 as referenced requires system component testing and inspection in accordance with [Table 14.4.2.2 (g)] and [14.4.5] to my knowledge and understanding, the magnet test does not verify smoke entering the chamber as required by the code. We do not allow magnet tests and the alarm service provider is responsible for the assurance the aerosol smoke is acceptable by the device manufacturer as required by the above referenced code. The Michigan building code and IFC or NFPA 1 reference the 2010 edition of 72 and the Handbook clarifies the testing methods and the smoke test is not considered a sensitivity test. The five year interval is only applicable if the detector is within the acceptable range after two consecutive annual calibrated tests in accordance with [14.4.5.3.3] and [14.4.5.3.4] if the fail they get cleaned and recalibrated or replaced in accordance with [14.4.5.3.5].

The key is this….. (magnet tests) do not simulate smoke entering the chamber as required and this allowance would be similar to one pulling down the pin on a flow switch or jumping connections to trip a device. Detector cleaning is also covered in the code for situations were they are dirty. Code compliant testing will "pick up" failing ranges.

 

[cda]

I may be thinking if two different things

Sensitivity testing and annual test

I need to look at the books

Sensitivity Testing: Any of the following tests shall be performed to ensure that each smoke detector is within its listed and marked sensitivity range:

1) Calibrated test method.

2) Manufacturer’s calibrated sensitivity test instrument.

3) Listed control equipment arranged for the purpose.

4) Smoke detector/control unit arrangement whereby the detector causes a signal at

the control unit when its sensitivity is outside its listed sensitivity range.

 

[cda]

I am curious where this area will divert in the future. Testing smoke alarms with smoke guarantees that the device activates as desires. Although, residue is left behind. Alternative testing (manufactures seem to like) is activation of device via magnet. Problems have arose that devices wont activate with smoke but they will with a magnet. My curiosity is, isn't the sensitivity test being done every 5 years picking up that loss of range?

NFPA 72 requires annual testing with smoke

 

[Scott Wytosick]

Thanks for the discussion!

The jurisdictional party responsible for verification of fire alarm system alarm initiating device testing for occupancies there in Holland should be requiring reporting verifications from system service providers. In these reports, should be areas indicating sensitivity testing results and needs by detector if the service provider is a reputable one in western Michigan. Are you a fire inspector?NFPA 72, 2010 as referenced requires system component testing and inspection in accordance with [Table 14.4.2.2 (g)] and [14.4.5] to my knowledge and understanding, the magnet test does not verify smoke entering the chamber as required by the code. We do not allow magnet tests and the alarm service provider is responsible for the assurance the aerosol smoke is acceptable by the device manufacturer as required by the above referenced code. The Michigan building code and IFC or NFPA 1 reference the 2010 edition of 72 and the Handbook clarifies the testing methods and the smoke test is not considered a sensitivity test. The five year interval is only applicable if the detector is within the acceptable range after two consecutive annual calibrated tests in accordance with [14.4.5.3.3] and [14.4.5.3.4] if the fail they get cleaned and recalibrated or replaced in accordance with [14.4.5.3.5].

The key is this….. (magnet tests) do not simulate smoke entering the chamber as required and this allowance would be similar to one pulling down the pin on a flow switch or jumping connections to trip a device. Detector cleaning is also covered in the code for situations were they are dirty. Code compliant testing will "pick up" failing ranges.

I am a firefighter/fire inspector for a volunteer fire department (recently became a state inspector for Indiana).

The fact that so much smoke is sometimes needed to test a device makes me concerned about the sensor. I agree with that magnets aren't a reliable way to test because they don't test the sensor. Proof of sensitivity tests and cleanings seem often ignored.

Sensitivity Testing: Any of the following tests shall be performed to ensure that each smoke detector is within its listed and marked sensitivity range:1) Calibrated test method.

2) Manufacturer’s calibrated sensitivity test instrument.

3) Listed control equipment arranged for the purpose.

4) Smoke detector/control unit arrangement whereby the detector causes a signal at

the control unit when its sensitivity is outside its listed sensitivity range.

That's what I check via paperwork.

 

[fireguy]

The manufactuer's magnet test does NOT test for smoke entry. But it does meet NFPA 72. We have convinced one local State FM to allow us to test 20% per year. At the end of the 5 year cycle, we have tested all the smokes. Two problems with a 5 year cycle

1.is convincing the client to OK the cost every 5 years. Especially with the turnover in maintenance staff.

2. We replace a smoke, because of failure. Next year we test the sensitivity, and it is OK. We skip a year and test for sensitivity the third year. If that test is OK, we next test the smoke in 5 years. Every one who thinks that will happen, wave your hand.

3. We probably lost the account at 5 year, to someone who says they never heard of such a test and we are running a scam.

4. My tester cost $5,000.00, plus $500.00 every couple of years to send it to the manufacture so they can make sure it is in calibration.

72-2013 Section 14.4.3.2 (g) Page 258 The commentary gives a good explanation.

When we are done with this discussion, I will start a new topic, duct detectors.

 

[cda]

The manufactuer's magnet test does NOT test for smoke entry. But it does meet NFPA 72. We have convinced one local State FM to allow us to test 20% per year. At the end of the 5 year cycle, we have tested all the smokes. Two problems with a 5 year cycle1.is convincing the client to OK the cost every 5 years. Especially with the turnover in maintenance staff.

2. We replace a smoke, because of failure. Next year we test the sensitivity, and it is OK. We skip a year and test for sensitivity the third year. If that test is OK, we next test the smoke in 5 years. Every one who thinks that will happen, wave your hand.

3. We probably lost the account at 5 year, to someone who says they never heard of such a test and we are running a scam.

4. My tester cost $5,000.00, plus $500.00 every couple of years to send it to the manufacture so they can make sure it is in calibration.

72-2013 Section 14.4.3.2 (g) Page 258 The commentary gives a good explanation.

When we are done with this discussion, I will start a new topic, duct detectors.

""""NFPA 72 requires annual testing with smoke"""

What is the reference for the magnet test being acceptable?

 

[FM William Burns]

But it does meet NFPA 72.

Please show me where in NFPA 72 a magnet test is permitted since 72 requires a test method "that ensures/simulates smoke entry" for system detectors. One can't convince me that a magnet "simulates" smoke entry............

 

[cstanford]

Not for annual test, but for sensitivity. NFPA 72 2016 : Table 14.4.3.2 - 17(h) -- (i) [at the bottom of the table]

--pulled directly from the book--

(h) Smoke detectors —
sensitivity testing
In other than one- and
two-family dwellings, system
detectors
N/A See 14.4.4.3 (i) Perform any of the following tests to ensure that each smoke detector is
within its listed and marked sensitivity range:
(1) Calibrated test method
(2) Manufacturer’s calibrated sensitivity test instrument
(3) Listed control equipment arranged for the purpose
(4) Smoke detector/control unit arrangement whereby the detector causes
a signal at the control unit when its sensitivity is outside its listed
sensitivity range
(5) Other calibrated sensitivity test method approved by the authority

--below table--
(i) There are some detectors that use magnets as a manufacturer’s calibrated sensitivity test instrument.