Transformer Overload

[That Man]

A client is putting me in an awkward position, and I have to explore a theoretical scenario that I can't predict.

In an emergency, what will happen to a 300KVA transformer if 900KVA of load is running? Will it run at all? How long will the transformer survive?

My client does not want to upgrade their service, because it costs too much. They have to install a fire pump, and the code clearly states that the service must have adequate supply for the locked rotor current of the fire pump. The client is taking the position with the utility that it's OK if the transformer is destroyed, so long as the fire pump runs for the duration of the fire. I'm sure there are code references that disallow this, but if the client was interested in the code, we wouldn't be in this situation in the first place, so I'm approaching this from another angle. What will physically happen to a transformer when put under 3X nameplate capacity? Are there any steps we can take to mitigate the extreme overload? Water spray might help cool it off, but I doubt it will help that much. I have a meeting with the client about this on Tuesday, and I would like to at least say I explored this scenario. Your thoughts?

 

[jumper]

Is the transformer customer or utility owned?

 

[Jraef]

In this situation when the fire pump needs to start, the starting current will likely saturate the transformer, the voltage will drop, the motor torque will drop by the square of the voltage drop and the pump may not start at the EXACT time you need it the most. That's exactly why that rule exists.

It's a FIRE pump, not a transfer pump! :slaphead:

 

[That Man]

It's utility owned. That's why I don't understand the point of my meeting on Tuesday. There is no way around this.

But I still want to explore the scenario, if only to say that I did.

 

[That Man]

In this situation when the fire pump needs to start, the starting current will likely saturate the transformer, the voltage will drop, the motor torque will drop by the square of the voltage drop and the pump may not start...

The pump is a 150HP pump. I would think it would be able to start without incident. But if it entered a continuous locked rotor state, which 695 says needs to be accounted for, would the transformer simple be unable to supply the power at all, like you are describing? Is there an upper limit to transformer power transference for a given rating?

 

[jeremy.zinkofsky]

Model the system in SKM or ETAP and perform a power flow study. Look at the voltage drop across the motor terminals under full load conditions. If you can prove that the voltage drop is below the normal operating tolerance of the fire pump then you have a case. Basically, prove that the fire pump will not work sufficiently enough to provide water to the sprinklers.

 

[That Man]

Model the system in SKM or ETAP and perform a power flow study. Look at the voltage drop across the motor terminals under full load conditions. If you can prove that the voltage drop is below the normal operating tolerance of the fire pump then you have a case. Basically, prove that the fire pump will not work sufficiently enough to provide water to the sprinklers.

I hope this works. I have already modeled the system with a 1000kva transformer (the size that is actually needed). Why don't I just run the scenario again with the smaller transformer. I'm kicking myself right now. I'll let you know if it works.

 

[jumper]

I hope this works. I have already modeled the system with a 1000kva transformer (the size that is actually needed). Why don't I just run the scenario again with the smaller transformer. I'm kicking myself right now. I'll let you know if it works.

The NEC does not cover utility owned transformers, your inspector/AHJ/someone else should be looking at the utility side for conformence to whatever is required on that side by the utility. NESC or some such code would be applicable.

Sounds like you getting shafted to do someone else job IMO.

 

[iwire]

It's utility owned.

That being the case I do not see the issue.

 

[That Man]

I remember now. I already tried this, but I wasn't confident in the results. Is there a way in ETAP to declare a pump is in locked rotor state? I just overrode the current to be 962A, but I wasn't sure if that would be accurate, so I began pursuing other methods. Maybe it is accurate enough. It shows motor voltage as 78.42%, and everything is highlighted red, but that doesn't tell me what the actual behavior of this circuit is. It doesn't matter if the electrical system is damaged. Is it running or not? The fire pump controller will throw an alarm, but will not stop the pump. This is where my quesiton comes from.

 

[gar]

161111-1416 EST

Jraef:

I don't believe you should describe a current overload on a transformer as saturating the transformer.

A transformer can be described by its linear equivalent circuit. These values should be relatively constant up to some excessive input voltage that does cause substantial core saturation. Or for some core materials at very low voltage excitation levels.

As you increase load current on a transformer this will reduce the voltage that is exciting the magnetic core. The volt-time integral determines how far the core is driven into saturation. Because of the transformer's primary source impedance and whatever impedance exists before the transformer a high load current means the voltage to the volt-time integral will be less.



That Man:

If you load a transformer to 3 times its rated current, then the transformer will have an output voltage that drops 3 times more than at rated load current. But power dissipation will be 9 times greater.

Maximum power transfer from any source occurs when the load impedance equals the internal impedance of the source.

If a motor has a locked rotor, then there is no power output, but there can be a lot of torque.

Is your 3 times overload based on locked rotor current, or running current?

If your transformer can supply sufficient current throughout the motor's starting needs by a reasonable margin, like 2 to 3 times the needed current, then you likely have a reliable system. Further if the running load on the transformer is less than or near the transformer rating, then there is no real heat problem in the transformer.
Starting time won't be that long to produce a lot of transformer internal heat..

.

 

[That Man]

gar,

Thanks for the response. I'm certain that under normal load conditions, the existing 300KVA system is adequate. But the code for fire pumps mandates that the power circuit must work under locked rotor current indefinitely. Under these conditions, the transformer will be loaded to around 900KVA, while only being sized to 300KVA. Etap says the transformer, the bus, and the cables are all inadequately sized, but Etap doesn't know that this is a fire pump under locked rotor conditions, and that the code allows for this, even though it will cause permanent damage to the system. Etap tells me that voltage will drop to 78%, but doesn't indicate whether the transformer can physically withstand such a heavy load, or if the pump will actually be able to spin. This sort of extreme behavior is not something that I can predict.

Let me put this into perspective. If this system we correctly sized, the transformer would increase to 1,000KVA, but all other elements of the system would remain in place. The fire pump code allows for 15% voltage drop, and this system would drop 11%.

You know what guys. I know this whole exercise is pointless. I don't know if the system can physically run or not. I don't want to find out either. Maybe I can use the voltage drop number to convince the client to abandon this line of reasoning.

 

[Phil Corso]

That man...

Core-saturation is a function of Magnetizing-current not load-current. Hence, any power-xfmr should handle a 300% load-current! The key is not magnitude, but, duration!

Regards, Phil Corso

 

[topgone]

A client is putting me in an awkward position, and I have to explore a theoretical scenario that I can't predict.

In an emergency, what will happen to a 300KVA transformer if 900KVA of load is running? Will it run at all? How long will the transformer survive?

My client does not want to upgrade their service, because it costs too much. They have to install a fire pump, and the code clearly states that the service must have adequate supply for the locked rotor current of the fire pump. The client is taking the position with the utility that it's OK if the transformer is destroyed, so long as the fire pump runs for the duration of the fire. I'm sure there are code references that disallow this, but if the client was interested in the code, we wouldn't be in this situation in the first place, so I'm approaching this from another angle. What will physically happen to a transformer when put under 3X nameplate capacity? Are there any steps we can take to mitigate the extreme overload? Water spray might help cool it off, but I doubt it will help that much. I have a meeting with the client about this on Tuesday, and I would like to at least say I explored this scenario. Your thoughts?

Assume your 300 kVA transformer to be having a 4.5% impedance. Also assume your utility has large distribution wires that it is stiff enough to start your 900 kVA load:

• Your secondary voltage will drop to about 77% of the rated secondary voltage.
• That would result into a motor torque reduction to (0.77)^2 ~ 60%.
• If your motor has a starting torque of 150%, the motor torque available at starting will just be 1.5 x 0.6 = 0.9 or 90%.

Please check if the motor load torque (if it's the fire pump, the pump speed torque characteristics) doesn't exceed that value, else, there's no way your fire pump motor can run successfully.

 

[Phil Corso]

ThatMan & TopGone…

Although I agree that a problem exists your “worst case scenario” is much too severe. Your Pre-Fault V-drop across the Xfmr was for a Running-load of 300-kVA! But, if the Fire-pump is 150-Hp, then the Running-load can’t be more 150 kVA! That said, then the extreme load on the Xfmr is 750 kVA, not 900-kVA!

A second point is related to voltage-regulation. Paralleling of the Running-Load with the motor’s LR-condition, will change the total-load Pf from say 80% to around 30%! It may not have such a deleterious impact on V-drop, as you thought!

My third point is your worst-case scenario assumes the Xfmr’s primary is connected to an “Infinite-Bus”! If it isn’t then it could be the deal-breaker!

Regards, Phil Corso

 

[Phil Corso]

161111-1416 EST
But power dissipation will be 9 times greater.
.

Copper-Loss will be increased 9 times the value at 1/2 load, not 9 times that of full-load!

Phil

 

[Phil Corso]

Copper-Loss will be increased 9 times the value at 1/2 load, not 9 times that of full-load!
Phil

Correction: Copper-Loss will be 6.25 times that of full-load power loss, not 9 times!

Phil

 

[Cow]

They have to install a fire pump, and the code clearly states that the service must have adequate supply for the locked rotor current of the fire pump. Your thoughts?

It's utility owned. That's why I don't understand the point of my meeting on Tuesday. There is no way around this.

But I still want to explore the scenario, if only to say that I did.

Why are you trying to apply the NEC to utility equipment that you know isn't covered under the NEC?

 

[topgone]

ThatMan & TopGone…

Although I agree that a problem exists your “worst case scenario” is much too severe. Your Pre-Fault V-drop across the Xfmr was for a Running-load of 300-kVA! But, if the Fire-pump is 150-Hp, then the Running-load can’t be more 150 kVA! That said, then the extreme load on the Xfmr is 750 kVA, not 900-kVA!

A second point is related to voltage-regulation. Paralleling of the Running-Load with the motor’s LR-condition, will change the total-load Pf from say 80% to around 30%! It may not have such a deleterious impact on V-drop, as you thought!

My third point is your worst-case scenario assumes the Xfmr’s primary is connected to an “Infinite-Bus”! If it isn’t then it could be the deal-breaker!

Regards, Phil Corso

Forget about the side issues. From what I gathered, his problem is basically to provide the client with a professional opinion regarding whether the additional fire pump can run successfully given the available 300kVA transformer capacity.
Like I said earlier, ThatMan needs to look at the fire pump specs and see the pump torque requirements and compare with the available motor torque at a 77% voltage level of the source at starting.

 

[Phil Corso]

TopGone...

While you seem to have no interest, others may be interested! So, here is the conclusion of my analysis: V-drop, across the transformer is in the order of 13%, not 23%!

Parameters used:

o Running-load is 150kVA (1/2 Xfmr Rating) @ 0.8 pf.

o Pump motor locked-rotor inrush is 600kVA @ 0.3 pf (4x motor rating).

o Resultant connected-load is ~ 730kVA @ 0.6 pf ( 2.4x Xfmr rating).

o Xfmr: kVA, 300; %Z, 4.75; Pcu, 4,000W @ FL; Pfe, 400W @ FL.

o % V-drop for running-load is ~ 2%.

o % V-Drop for combined loads is ~ 13%.

Regard, Phil