Fire alarm

Status
Not open for further replies.
T

texie

Senior Member
Location
Fort Collins, Colorado
Occupation
Electrician, Contractor, Inspector
Existing building and fire alarm. Call it system "A". An addition is put on the the building. There are no firewalls and both the old and new sections are 1 building. The FA designer not realizing this, specified a separate FA system for the addition. Call this system "B". System "A" and system "B" are different vendors. The FA contractor has proposed tying the systems together using supervised relay circuits so that alarms, troubles, etc will pass both ways to act as one system. The original system "A" will be the one with the central station communicator.

We have already discovered in testing that there is 1 issue that will have to be addressed, but I'm not sure if it is worth discussing at this point if this whole arrangement can't be done in a compliant manner.

I'm wonder if this can be done in a manner that complies with NFPA 72?
 
texie said:
Existing building and fire alarm. Call it system "A". An addition is put on the the building. There are no firewalls and both the old and new sections are 1 building. The FA designer not realizing this, specified a separate FA system for the addition. Call this system "B". System "A" and system "B" are different vendors. The FA contractor has proposed tying the systems together using supervised relay circuits so that alarms, troubles, etc will pass both ways to act as one system. The original system "A" will be the one with the central station communicator.

We have already discovered in testing that there is 1 issue that will have to be addressed, but I'm not sure if it is worth discussing at this point if this whole arrangement can't be done in a compliant manner.

I'm wonder if this can be done in a manner that complies with NFPA 72?
Click to expand...

Absolutely. "Master/Slave" configurations are done all the time. Usually it's between buildings, but there's no code limitation I know of that prohibits it within a building.
 
gadfly56 said:
Absolutely. "Master/Slave" configurations are done all the time. Usually it's between buildings, but there's no code limitation I know of that prohibits it within a building.
Click to expand...

Yes, understand master/slave arrangements. The issue here is we have 2 different FA panel vendors.
 
texie said:
Yes, understand master/slave arrangements. The issue here is we have 2 different FA panel vendors.
Click to expand...

It makes no difference whether they are Silent Knight/Silent Knight, FireLite/Silent Knight, or any other pair of vendors, or whether they are both conventional, both addressable, or a mix. The key is, can they be made to work in concert, and unless you are dealing with conventional panels already at their zone limits the answer will be "yes".
 
gadfly56 said:
It makes no difference whether they are Silent Knight/Silent Knight, FireLite/Silent Knight, or any other pair of vendors, or whether they are both conventional, both addressable, or a mix. The key is, can they be made to work in concert, and unless you are dealing with conventional panels already at their zone limits the answer will be "yes".
Click to expand...

OK then, let me explain the "issue" discovered in testing.
System "A" goes into alarm and system "B" does as well. If you silence system "B" and then get another alarm from system "A", system "B" does not go back into audible alarm.
 
Got to run...I'll be back to this post at 7 am mountain time.
Please keep the comments coming as I have an 8:30 am mountain time conference call to discuss this with some of the brass.
 
texie said:
OK then, let me explain the "issue" discovered in testing.
System "A" goes into alarm and system "B" does as well. If you silence system "B" and then get another alarm from system "A", system "B" does not go back into audible alarm.
Click to expand...

Are you cross zoning with conventional zone inputs, or addressable monitor modules?
 
texie said:
OK then, let me explain the "issue" discovered in testing.
System "A" goes into alarm and system "B" does as well. If you silence system "B" and then get another alarm from system "A", system "B" does not go back into audible alarm.
Click to expand...



It's a horrible idea, IMO. It rarely flies around here as iWire said, but it can work. No AHJ I deal with would allow both panels to require reset when only one is detecting.

It is a programming issue, they may need to add more modules. You could try using a modules that are mapped as silenceable and auto resetting That is-

when system A is in alarm the module triggers system B's input module. However system B is not in an alarm condition, It activates NAC's but the red light isn't on. Now you don't have to reset two panels when only one is in alarm.

when system A is silenced, the module quiescences. System B's input module sees the the restoral and silences it's devices.

2nd alarm on A repeats the process.


Keeping the strobes going when A is silenced but still in alarm (until a reset) is a little more complex but can be done.



And cross supervision and circuit integrity is absolutely necessary. Ensure the input zones/modules are supervised to their end.
 
There is another issue not yet covered. You must have a standalone DACT (in it's own enclosure with it's own backup batteries) and seperate inputs for alarm, supervisory, and trouble FOR EACH SYSTEM. This way a catastrophic failure of one fire alarm panel will not affect a signal generated by the other.
 
MichaelGP3 said:
There is another issue not yet covered. You must have a standalone DACT (in it's own enclosure with it's own backup batteries) and seperate inputs for alarm, supervisory, and trouble FOR EACH SYSTEM. This way a catastrophic failure of one fire alarm panel will not affect a signal generated by the other.
Click to expand...

It appears they are attempting to make one system (which I think it's a bad idea), the AHJ could allow it (which I think is a bad idea).

Is it a compliance/listing reason not to allow the master/slave configuration?
 
No. Its because I would not want my name brought up when the fire department did not respond to a fire due to an admittedly rare case of a catastrophic failure (think plumbing leak) affecting system 'A' that powered the DACT.

It isn't my idea. A Radionics sales engineer explained that their firm wouldn't support monitoring multiple fire alarm systems with their 3 zone D2071A DACT, and the reason above is why.
 
Last edited:
MichaelGP3 said:
No. Its because I would not want my name brought up when the fire department did not respond to a fire due to an admittedly rare case of a catastrophic failure (think plumbing leak) affecting system 'A' that powered the DACT.

It isn't my idea. A Radionics sales engineer explained that their firm wouldn't support monitoring multiple fire alarm systems with their 3 zone D2071A DACT, and the reason above is why.
Click to expand...

Excellent point. So that would probably mean the DACT would be required to be in the same area of protection that it's reporting for? There is probably something in UL864 about that. (Wondering if it would be required to be within 20' and in conduit too)

Then it stand to reason you would need to have a smoke from system A and B covering the DACT. This is fun!!! What else?
 
MichaelGP3 said:
No. Its because I would not want my name brought up when the fire department did not respond to a fire due to an admittedly rare case of a catastrophic failure (think plumbing leak) affecting system 'A' that powered the DACT.

It isn't my idea. A Radionics sales engineer explained that their firm wouldn't support monitoring multiple fire alarm systems with their 3 zone D2071A DACT, and the reason above is why.
Click to expand...

Let's think about this a little bit. Keep in mind the OP has ONE building in two parts. Suppose you had a single panel covering this building and there was a plumbing leak that disabled the panel and there was a fire. Do you suppose you'd be held accountable in this instance? Now we have an "A" panel with the DACT and a "B" panel without. The plumbing leak destroys panel "A". A fire breaks out in the panel "B" area. Notification appliances sound and occupants are safely evacuated from the area in which the fire actually occured. Think anyone's coming after you with a hatchet? I don't think so either. Your objections would be more telling if you were trying to monitor several physically separate buildings. In that case, I suspect the AHJ wouldn't allow it.

The A/B panel setup isn't really very different than installing a Silent Knight IFP-1000 with an RPS-1000 remote enclosure sporting a 2815XL SLC expander card. Put a bullet in the RPS-1000 and the area it covers is unprotected. Put a bullet in the IFP-1000 and BOTH areas are unprotected; the RPS-1000 doesn't have a "degrade" mode available. The A/B arrangement effectively gives you degrade operation. Normally you'd have to move up to a high end panel like Siemens XLS for that. True, it's not as well supervised, but there are partial solutions for that, especially if both panels are addressable.
 
Last edited:
This arrangement happens all the time in pre-action and clean agent releasing systems.

The existing panel might not be listed for releasing, or not listed for releasing a particular solenoid, so a sub-panel gets installed. With the interconnecting relays and monitoring modules, so each panel knows what the other is doing, the releasing panel does its thing, lets the main panel know and all is wonderful in the world.
 
gadfly56 said:
Let's think about this a little bit. Keep in mind the OP has ONE building in two parts. Suppose you had a single panel covering this building and there was a plumbing leak that disabled the panel and there was a fire. Do you suppose you'd be held accountable in this instance? Now we have an "A" panel with the DACT and a "B" panel without. The plumbing leak destroys panel "A". A fire breaks out in the panel "B" area. Notification appliances sound and occupants are safely evacuated from the area in which the fire actually occured. Think anyone's coming after you with a hatchet? I don't think so either. Your objections would be more telling if you were trying to monitor several physically separate buildings. In that case, I suspect the AHJ wouldn't allow it.

The A/B panel setup isn't really very different than installing a Silent Knight IFP-1000 with an RPS-1000 remote enclosure sporting a 2815XL SLC expander card. Put a bullet in the RPS-1000 and the area it covers is unprotected. Put a bullet in the IFP-1000 and BOTH areas are unprotected; the RPS-1000 doesn't have a "degrade" mode available. The A/B arrangement effectively gives you degrade operation. Normally you'd have to move up to a high end panel like Siemens XLS for that. True, it's not as well supervised, but there are partial solutions for that, especially if both panels are addressable.
Click to expand...

Occupant evacuation was never the issue. Automatic fire department response, which may be needed at night or on a weekend when there is no one present is the issue.
 
ron said:
This arrangement happens all the time in pre-action and clean agent releasing systems.

The existing panel might not be listed for releasing, or not listed for releasing a particular solenoid, so a sub-panel gets installed. With the interconnecting relays and monitoring modules, so each panel knows what the other is doing, the releasing panel does its thing, lets the main panel know and all is wonderful in the world.
Click to expand...

Of course, you are correct.
 
We call this a "Ring Around". I thought they went out of style with addressable systems.

I think most systems can be programmed to be resettable in this arrangement. You will need some collaboration between both vendors to make it work.

The panel with the DACT should be considered the master panel and the other one the slave. As long as the slave reports all alarms, troubles and supervisories to the master, it's ok.
 
We have this in my buildings... separate tenant-installed pre-action on one floor, separate Viking panel, reports to the main FACP (Gamewell). New tenant installed A/V's in the ceilings, had to install new power extenders, used Silent Knight, tied into my existing Wheelock's on each floor. Worked with both the old Siemens panels, as well as my new Gamewell's. When the new FACP/system was installed, my contractor installed a separate panel with relays, both panels were live, which enabled them to swap out one floor at a time... ingenious, and homebrew.
 
ron said:
This arrangement happens all the time in pre-action and clean agent releasing systems.

The existing panel might not be listed for releasing, or not listed for releasing a particular solenoid, so a sub-panel gets installed. With the interconnecting relays and monitoring modules, so each panel knows what the other is doing, the releasing panel does its thing, lets the main panel know and all is wonderful in the world.
Click to expand...

In any of theses cases I have worked on the pre-agent systems were above and beyond the main system.

In other words the main system took care of all the required coverage while the precaution systems were optional and provide some redundant coverage.

Even with a failure of the pre action panel the building would still have all the required coverage.

I see this as different than using two panels to handle different areas of the same building.
 
Status
Not open for further replies.
Top