Fire Pump Feeder Sizing

[kitha]

NEC requires 125%xFLA to be the minimum feeder size for the fire pump. But in order to carry the motor Locked Rotor Amps continuously, the breaker is usually sized to 600%xFLA. How is it possible for the feeder to carry Locked Rotor Current continuously if the feeder is under sized( i.e 125% x FLA)? I couldn’t find anything related to this scenario in the code. Really appreciate your help on this.

 

[jraef]

In theory, the conductors feeding the motor will withstand the thermal effects of LRC longer than the motor windings themselves. So in a "run to destruction" scenario, once something gives way in that chain, the capacity of the rest of it becomes moot.

"Will work for (the memory of) salami"

 

[7anoter4]

As jraef said the conductor suitable for a certain ampacity will not reach the emergency temperature in the protection set time for 8 times the ampacity.
Let’s take –for instance-a 100 HP squirrel cage rotor induction motor. Usually the protection time setting for 6-8 times rated current will be 1-5 sec.
The voltage 480 V.FLC as per NEC Table 430.250 is 124 A. The conductor ampacity has to be Amp=124*1.25=155 A and locked rotor current 8*124=992 A.
According to NEC Table 310.15(B)(16) copper conductor XLPE/EPR insulated [90 DGR.C.] 1/0 [ 170 A ampacity]. The maximum temperature for short-time [EMERGENCY] it is 130 oC. The conductor temperature reaches this limit in more than 15 sec. Even for 8*170=1360 A the time to reach the emergency temperature will be 10 sec. [The conductor heating was considered adiabatic phenomenon-no heat dissipation outside]

 

[kitha]

Fire pump is supposed to "run to destruction". Then isn't it ironic to set a protection time setting 1-5 sec? Never heard it before.

 

[dpc]

The simple answer is that you need to meet the NEC. Trying to apply logic to the NEC will waste a lot of your time that you could be using more effectively. This requirement has been unchanged in the NEC for decades IIRC, so if there was an issue, it probably would have come to light by now. As 7anoter4 illustrates, there is generally some engineering basis for the NEC requirements, but it really doesn't matter. You still have to follow it.

Also, as jraef points out if the motor draws LRA for very long, it is going to fail, most likely prior to a cable insulation failure.

 

[kitha]

dpc- Even though NEC says 125% FLA, NFPA 695 calls to size the breaker to run LRA continuously. Having just now discussed with the TornaTech(one of fire pump controller manufacturers) tech support, there's no protection set time in their controllers. I am not contesting NEC but trying to find and answer to a seemingly look like contradiction.

 

[dpc]

If you size it per the NEC, you'll be fine.

Regards,

Dave

 

[John2025]

My guess at their logic is:
A) 125% of FLA for the wires to make sure pump can run continuously without overheating wiring.
B) 600% of FLA continuous rating of protection to make sure pump runs to destruction while still maintaining a small ability to interrupt the circuit in case of a short circuit in the wiring that might occur without a fire being present.
Makes sense to me. Like somebody's tagline, "there's probably a good reason it's always been done that way".

 

[Parchie]

There seems to be a misunderstanding on the "run to destruction" principle there! Simply, the principle requires the fire pumping equipment to be run until destroyed. It doesn't need a rocket scientist to differentiate an overload condition from a short-circuit, the latter being addressed in the fire protection codes while allowing for the motor to run until destroyed "on overload". Hope that clears things up here.

 

[7anoter4]

First of all I said the cable which ampacity corresponds to 1.25*motor rated current will withstand 800% rated current for more than 15 sec. In the case of "usually" 600% of rated current the cable conductor temperature will reach the emergency temperature in more than 100 sec. However it is less than 200 sec. as stated in NFPA 20 but in my opinion it is enough time to burn out the electric motor.
If an actual short-circuit will occur at motor terminals -for more than 600% rated current-the fault clearing time has to be limited to 1-5 sec. In my opinion it is also the spirit of NEC Art.695.4.

 

[John2025]

If I could figure out how to edit I'd line out "small ability" on my post. Protecting 1/0 wire with an 800A breaker seems to increase the possibility of an arcing fault that doesn't clear the breaker, but I guess I'm grasping at straws to justify my initial response

 

[kitha]

Since I initiated this thread, I contacted Tornatech tech engineer. Their controllers can be set for LRA to run up to 20ms. Usually set it to between 8 to 20ms. This means it trips the circuit automatically if LRA flows more than 20ms. It happens both in automatic and manual mode too. In manual mode there's a push button which bypasses pressure signals yet still the pump is protected by its protection settings. In short, apparently LRA wont flow more than 20ms by any means.

Therefore the cable sized for 125% FLA apparently can handle motor LRA for 20ms. I think 7anoter4 has a very good illustration in his first posting.

"Run to destruction" seems to be a myth or perhaps can be applied only when sizing the circuit breaker

You can download one of the Tornatech user manuals from below URL

http://www.tornatech.com/FichiersUpload/Manuals/GPL+GLU-manual-E.pdf

 

[dpc]

kitha,

That explanation doesn't really make any sense. The LRA will last for seconds not milliseconds. Perhaps he was talking about the motor inrush, which is not the same thing as LRA. 20 ms for asymmetrical current might be about right.

An NFPA approved fire pump controller should never trip on locked rotor amps unless there have been some drastic changes in the NEC lately that I am not aware of.

 

[jraef]

The latest revisions to the NEC acknowledge the possible confusion they have created by adding this sentence:

The requirement to carry the locked-rotor currents indefinitely shall not apply to conductors or devices other than overcurrent devices in the fire pump motor circuit(s).

They don't go into why they do it, they just are saying "Do it because we said so." But in some commentary I read on the changes, the people providing input reiterated that the concept is to NOT trip on a Locked Rotor Current scenario in the motor circuit. But LRC is not the same as a short circuit, although we tend to over simplify them in that way. A bolted fault or ground fault will, as we know, allow the total available fault current in the circuit to flow into the fault, which can FAR exceed the motor LRC. But take that idea another step further; HOW is that going to happen? It is going to happen BECAUSE your conductor has ALREADY failed you. In that case, you are already deprived of that fire pump motor; if there is a bolted fault between the source and the motor, the motor is NOT going to get that current.

What might be behind this is an acknowledgement of the reality of what is available out there. Any circuit breaker, regardless of the trip features, comes with an overriding current release point of 10X (1000%) of the breaker FRAME rating. The frame rating is, in essence the maximum current that any breaker can carry; it is the current capacity of the conductive parts inside. The 10X trip feature is inherent even in Molded Case Switches, i.e. there are ostensibly NO trip elements inside at all, at least to the causal observer. But when you read the "fine print" there actually is, and the override release point is 10X the frame rating. So if you have a 600AF molded case switch / breaker, it CANNOT carry more than 6000A for very long not matter what you think the trip setting is (if any). By saying then that the OCPD in the fire pump circuit must be sized for continuously supplying the LRC of the motor, they are indirectly saying that the breaker will NOT trip under any circumstance if the motor shaft is locked, not if there is a short circuit in the conductors. If they did not word it that way and allowed the other OCPD rules to apply from article 430, then it might be possible for someone to install a device that will open up under that 10X override, which is NOT what they want to happen.

"Will work for (the memory of) salami"

 

[kitha]

I found the correct explanation finally I think. Brian J McPartland's NEC 2011 Handbook has correctly explained the scenario. Basically it has conceded that the conductors are undersized to carry the LRA. But since this scenario happens very rarely, it does not pose a significant issue. See my upload in below URL

http://files.engineering.com/getfil...e=Fire_Pump_Conductors-Brian_J_Mcpartland.JPG

 

[davidbeach]

Besides, if prolonged locked rotor current from the fire pump damages your conductors, you probably have far worse issues to deal with than the damaged conductors.

 

[kitha]

davidbeach

That could be true. But since conductor installation should pass all other stringent requirements and the scenario is very rare, the code apparently does not mandate conductors to be larger than 125% FLA and 15% maximum starting voltage drop.

 

[majorminotaur]

I look at it as the Fire Pump is typically (pretty sure its required to be) in at least a 1.5 hour fire rated room. As others have mentioned the motor itself will likely be the point of failure. In a situation where there is a fire, it is worthwhile to try and make the fire pump work to the point of destruction because any water it can pump out will be beneficial in the time frame to assist people that are attempting to escape a building. By the time any fire in the fire pump room escapes from that room, people are given sufficient time for escape. Fire pumps are also typically in basements or on grade, where normal water pressure from a sprinkler system should be enough to extinguish a fire in this room, but this is purely conjecture about the logic used. I think the main point being is it gives people time to escape.