Testing Rate of Rise Heat detectors

[PMcCarty]

Hi All,

Just wondering how some of you may have tested combination Fixed Heat and Rate of Rise Heat Detectors?

We use a heat gun and blow the hot air across the device and hope that the Rate of rise goes off before the non-resettable heat detector does. I have thought of placing a wet rag on the thermal part of the device, but not sure that would work.

I look forward to you suggesting and Ideas on this issue.

Phil

 

[mdshunk]

I seem to remember that you don't field test them. You take two out of every 100 down and send them off for lab testing. I vaguely remember reading that in some NFPA document, but I don't remember which one at the moment.

 

[mdshunk]

EDIT: what I wrote above was only for non-resettable heat detectors. For the combo type, the procedure is:

"The rate-of-rise test should be conducted by a qualified fire protection specialist
using a 1,000-watt heat gun (or hair dryer). Under normal ambient conditions
with an air velocity at the detector of 400 feet per minute, or less, the detector’s
contacts should close within 10 seconds when the heat gun nozzle is
located approximately 6 inches from the detector. For every additional 400
feet per minute of air velocity, the heat gun to detector spacing should be reduced
by one inch (e.g. 400-800 FPM, 5 inches; 800-1,200 FPM, 4 inches,
etc.).
Caution must be taken to prevent operating the nonrestorable fixed temperature
element (center disk) during this test. If the fixed temperature element operates,
the detector must be replaced."

Better add the link too, so I don't get accused of stealing again:
http://www.edwards-signals.com/files/i-280b_installation.pdf

It sounds like you'd need to have your heat gun checked with an anemometer to see what the air flow is like so you know how far out to hold it. If you pop the heat thermal element following this procedure, it sounds like the rate of rise component failed, eh?

 

[Lxnxjxhx]

What is the spec'ed value of the rate of rise, in degrees per second or per minute? I guess the rate of rise for the sensing device is also given for a limited temperature range. You have an online spec for this device?

Once you know the specs, you can use a very small (very low thermal time constant) thermistor or temp meter, and heat source, and confirm that your testing method is correct. The response time (thermal time constant) of the temp meter should probably be at least >5x faster than the rate of rise of what you're testing.

 

[Lxnxjxhx]

. . .hope that the Rate of rise goes off before . . .

. . .hope that the Rate of rise goes off before . . .

Here's another citation
http://www.patentstorm.us/patents/5450066-claims.html
that talks about rate-of-rise sensors.

Their test method trips the sensor very fast because the rate of temp rise is much faster.
That's the risk; they use commonly available heat sources that can deliver temps way above these trip points, so instead of cooling the heat source they reduce the time the sensors are exposed to these high temps.

Instead of shielding the hi-temp nonresetable detector, you might make a cardboard funnel to focus only on the rate-of-rise detector.

Assuming you can't do the shielding or funneling, if you have a heat source that puts out less than 130 degrees F you will never trip the non-resetable hi-temp sensor.
To give some safety margin, let's say max 120 F.

Assuming a room ambient of 75 degrees F.
120-75 = 45 degree F rise in temp.
45/(>15) = <3 minutes to get to 120 F.

Using this cooler heat source the rate-of-rise detector should trip in less than 3 minutes. The air from a partially-constricted vacuum cleaner exhaust is probably hot enough to do this. You don't have to worry about air speed with this as long as the air temp at whatever speed it's delivered is around 120 F.

If a working unit doesn't trip within 3 minutes, I assume because the rate-of-rise detector response time is too slow, and you can't use shielding or funnels, then I don't see any totally safe way to test the rate-of-rise sensor. And there may be none, otherwise they might have had you try lower temp settings on the hair dryer.

If you have the 194 F sensor, you can change these numbers around and have a lot less risk of tripping the 194 F sensor.

I hope these things don't have nuisance tripping on hot days in southern climates, or because they are mounted on a ceiling that is the floor of an uninsulated attic.

 

[iwire]

My experience has been you do the test however the fireman tells you regardless of it being right or wrong.

If the FD is happy so am I.

 

[peter d]

Lately we've been testing the addressable ones with a magnet. This is acceptable to our local AHJ's.

 

[Lxnxjxhx]

AHJ vs. quick and dirty

AHJ vs. quick and dirty

I guess if the AHJ signs off on it, you are covered, so this is the best. So why does anyone want to do testing in other ways?

Is the AHJ like The Cable Guy, who doesn't show up? Are they unnecessarily finicky?
It seems to me it is in the interest of public safety to make this whole process reasonably easy, fast and convenient. Otherwise people will look for shortcuts.

Also, I'm surprised these things don't have self-test buttons; technically, they are less complex than other things discussed on this forum that do have self-test buttons. All these things do is measure temperature vs. time, or just temperature, and give a go/nogo indication.

IMHO, the Edwards AHJ-approved procedure seems that it could easily turn from a non-destructive test into the other kind. And I don't think buying some special $5K test unit is such a good option, either, even if someone makes it.

 

[MichaelGP3]

You have a good memory.....

You have a good memory.....

I seem to remember that you don't field test them. You take two out of every 100 down and send them off for lab testing. I vaguely remember reading that in some NFPA document, but I don't remember which one at the moment.

This is how it's presented in the 1999 edition of NFPA 72 for non-resettable spot type heat detectors. You had to do this 15 years after installation. If the lab finds them to be working properly, you have to repeat this procedure every 5 years thereafter. Alternatively, the code directs you to replace all the detectors.

 

[MichaelGP3]

Not knowing what other people might be using for a heat source....

Not knowing what other people might be using for a heat source....

Instead of shielding the hi-temp nonresetable detector, you might make a cardboard funnel to focus only on the rate-of-rise detector.

I wouldn't do this, or advise anyone to do this.

 

[gadfly56]

Here's another citation
http://www.patentstorm.us/patents/5450066-claims.html
that talks about rate-of-rise sensors.

Assuming you can't do the shielding or funneling, if you have a heat source that puts out less than 130 degrees F you will never trip the non-resetable hi-temp sensor.
To give some safety margin, let's say max 120 F.

Assuming a room ambient of 75 degrees F.
120-75 = 45 degree F rise in temp.
45/(>15) = <3 minutes to get to 120 F.

Using this cooler heat source the rate-of-rise detector should trip in less than 3 minutes. The air from a partially-constricted vacuum cleaner exhaust is probably hot enough to do this. You don't have to worry about air speed with this as long as the air temp at whatever speed it's delivered is around 120 F.

You might find yourself waiting next to forever for the ROR portion to respond. Typical spec for the ROR portion is 15 F per minute. It's actually a form of HAD. There is a chamber with a calibrated leak. Heat the chamber up fast enough, and the air can't escape the chamber, causing it to alter shape and bring the contacts together. Otherwise, air moves in and out of the chamber in response to ambient conditions without nuisance trips.

 

[dbuckley]

Just a note on heat guns; I have one that has a variable output, with a knob to adjust the output from lukewarm to $#&^ing hot. This isn't some posh gun, its an el cheapo Chinese thing, that replaced my expensive two heat gun when it passed away. Anyway, the way it controls power is by burst control, so when its on a middle setting my workbench lighting dims as the burts goes on and brightens when the burst goes off.

Thus this thing does not put out anything approaching a constant temperature... So be wary of variable heat guns...

 

[quogueelectric]

I used to do it with a lighter when requested to by the fire marshall.

 

[Weaver Road]

Heat guns, r.o.r, magnet etc.

Heat guns, r.o.r, magnet etc.

I realize this thread has been dead for a couple of days, but while I usually just listen and learn here on the board, i thought I should throw out my thoughts on this one.

As an AHJ, I have learned to fall back on two solid stances on how to test and what to test.
1 the code-whatever code is appropriate for your jurisdiction.
2 the manufacturer's instructions/listing.

Keep in mind that they paid all the money for the testing to get it right, or at least get it legally defined as to what should be right.

I have sometimes been met with funny looks when asking a contractor to test it according to the installation instructions. The testing is sometimes painfully slow to do it right. The proper heat gun in this example is critical. Hair dryers seem to be a good way to get quick heat without popping the fixed element.

As for the "magnet test" we no longer allow those for functional testing of the detection devices. The magnet is only testing the circuitry, digital address, etc. It is not testing the function of the detection device. We do allow the magnet for non-resetable heats, for obvious reasons. We also allow the magnet if we are only testing to see the programming sequence or address changes.

I work with contractors who have excellent reputations, do quality work and now we understand together the value in the long run for example, of 100% testing for situations such as initial acceptance testing of a fire alarm system.

I recently did 750 new initiating devices on a system installed by an excellent contractor and programmed by someone I have trusted for 14 years. We did it (3 days worth!) to be sure and follow the code. We found one bad smoke and two bad addresses, including two devices with the same digital address. The panel had not yet found the duplication, but we did. We both had a great sense of satisfaction that we could really count on the performance after the test and both sleep well at night. Obviously that 1/2 percent error was extremely low, but critical if you are the one in the room with only one detector and it's the one programmed wrong.

Each AHJ does as they see fit, but the rality is that most should be doing the same thing, since we are working from nearly all the same books. The codes in most states allow for some AHJ discretion, but not as much as is commonly used.

Off my soap-box, now I go back to listening and learning from all of you.