Zero Sequence CT

Status
Not open for further replies.
T

timm333

Senior Member
Location
Minneapolis, MN
Occupation
Electrical Design Engineer
What is the rule for sizing a zero sequence CT. For example if we have a 100A 3-phase circuit (coming from a 100A breaker) and we want to place the primary of a zero sequence CT on it, what would be the size of the primary of this zero sequence CT? Thanks
 
timm333 said:
What is the rule for sizing a zero sequence CT. For example if we have a 100A 3-phase circuit (coming from a 100A breaker) and we want to place the primary of a zero sequence CT on it, what would be the size of the primary of this zero sequence CT? Thanks
Click to expand...

Do you have a brand name for ther CT's?, that would make it easy to look up their ratings.
 
I guess the question is whether you want to be able to measure the maximum possible zero sequence current or just detect a fairly low level current and not worry about saturating at the maximum value.
In a GFCI, for example, it does not matter exactly what the output current is for a 15A GF current.
But the threshold requires linearity in the 6ma range and increased output current in the higher unbalance range that is good enough to allow it to match the UL trip curve.
 
I don't have the model number at this stage. For a 100A three phase three wire circuit, we put total four CT's. The first three CT's are regular CT's and are placed one on each phase, each of these three CT's is sized as 100A:5A, this makes sense.

The fourth CT is the ZSCT (zero sequence CT) which is placed around all three phases; should this ZSCT be also sized as 100A:5A, or should it be 300A:5A ?
 
timm333 said:
I don't have the model number at this stage. For a 100A three phase three wire circuit, we put total four CT's. The first three CT's are regular CT's and are placed one on each phase, each of these three CT's is sized as 100A:5A, this makes sense.

The fourth CT is the ZSCT (zero sequence CT) which is placed around all three phases; should this ZSCT be also sized as 100A:5A, or should it be 300A:5A ?
Click to expand...
The answer to that question is given in part by the knowledge that if the source voltages are balanced the maximum unbalanced current with loads with the same displacement power factor but worst case unbalance will be roughly 100A.
If one phase has a leading power factor near zero and another has a lagging PF near zero, you could get 200A, I think.
 
GoldDigger said:
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . .
If one phase has a leading power factor near zero and another has a lagging PF near zero, you could get 200A, I think.
Click to expand...
More than 200A, even closer to 300A! :)
I played with your scenario on a 400V, three phase imaginary system-->with one line to neutral load at 0.01PF leading (231 VA) and the other line to neutral load at 0.01PF lagging (231 VA) and the lone UPF load (23.1kVA); I got 272A!
Do we consider that scenario above? I think most design for an imbalance when losing load on one phase.
 
topgone said:
More than 200A, even closer to 300A! :)
I played with your scenario on a 400V, three phase imaginary system-->with one line to neutral load at 0.01PF leading (231 VA) and the other line to neutral load at 0.01PF lagging (231 VA) and the lone UPF load (23.1kVA); I got 272A!
Do we consider that scenario above? I think most design for an imbalance when losing load on one phase.
Click to expand...
Since the OP mentioned a three wire delta, I did not consider line to neutral loads. Yes, that case could get above 200A on the ungrounded lines, but would give you zero if you include the neutral.
Meanwhile, even with unbalanced line to line loads, the only way to get zero sequence current in a three wire delta would be a fault, yes?
It was late last night.... :(
 
timm333 said:
What is the rule for sizing a zero sequence CT. For example if we have a 100A 3-phase circuit (coming from a 100A breaker) and we want to place the primary of a zero sequence CT on it, what would be the size of the primary of this zero sequence CT? Thanks
Click to expand...

What exactly are you trying to measure? For example: Is this a feeder to a motor? Or maybe a feeder to a delta primary transformer? Or a feeder to a panelboard?

The loads or load power factor do not matter. As Gold finally noted, if there is no fault current, the zero sequence CT has no output. And if there is a fault, the current could easily be up in the 5000A - 1000A range.

Zero sequence CT are generally for ground fault protection. What exactly are you trying to protect? A few amps on a motor going bad? A few milliamps on an instrument going bad? Short answer: Size the CT for what you need to see - or what the gf relay needs to see.

ice
 
If you are using a ZSCT, you must be using a Zero Sequence input to a protective relay of some sort. Depending on the sensitivity you are looking for and/or the type of system it is connected to, the ratio of the ZSCT is not related to the primary current, it is related to the possible zero sequence current that you will allow before tripping the relay. So for instance if you are using a Multilin 750 FPR or a 469 MPR, the ZSCT recommendation is 50:5 for solidly grounded systems, 50:0.25 for resistance grounded systems, because your current to ground will be restricted by the NGR.

Don't confuse this with Residual GFP, or GFP vs GFCI, those are all different issues. If you are required to provide ZSGFP, you must have a system that is looking for the input from it, and THAT system is what should dictate the selection of the ZSCT. Once you know that, the only other issue is the size of the window opening so that you can get all of your cables through it.
 
GoldDigger said:
...
Meanwhile, even with unbalanced line to line loads, the only way to get zero sequence current in a three wire delta would be a fault, yes?
It was late last night.... :(
Click to expand...
Yes, with a ZSCT, only current that is going to ground* will be seen. The vectorial sum of even unbalanced phase currents will still come out to zero, because regardless of the current balance, all current originated from another phase and goes back through another phase, and since they all go through the ZSCT, it nets out as zero.

* The possible exception is cable capacitive charging current, which may flow out differently in each phase because of differences in capacitance for long enough to cause a sensitive relay to trip. That's why placement of the ZSCT near the relay / breaker and routing of cables and shields is critical on MV and HV systems. But not something to worry about on most normal LV installations such as described here.
 
As jraef indicated, the first criteria for selecting a zero sequence CT is the amount of ground fault current available on the system. If you system is solidly grounded use a 50:5 CT. If it is resistance grounded it depends on the let through current of the resistor. Most low resistance grounded systems can also use a 50:5 CT, while high resistance grounded systems need a more sensitive CT, often described as a 2000:1 or 50:0.025

The second consideration in applying a zero sequence CT is the type of relay that will be used with the CT. This is significant with electromechanical relays due to the energy the CT must provide to operate the relay, referred to as the burden of the relay. The common 50:5 zero sequence CT only has ten turns of wire around the core and is not able to transmit very much energy from the primary to the relay connected to the secondary of the CT. When zero sequence relaying became popular the relay manufacturers and some others published lists of acceptable CT and relay model combinations that would work together - i.e. a particular CT that could deliver enough energy to operate a particular relay. This CT and relay matching is not a concern when zero sequence CTs are applied with solid state or microprocessor protective relays that have a separate control power source and therefore require very little energy from the CT.
 
Thanks, this is all very good information! This is a feeder for a VFD and ZSCT is required to operate Multilin relay. So is 2000:1 (50:0.025) the only standard ratio, or are other ratios (like 2000:5) also available?

Also if we have a 100A circuit fed by a 100A breaker, and the circuit has a short circuit current of 10000A: then what would be the theoretical maximum ground fault current. Will it be 100/1.732=57.7A, or will it be 10000/1.732=5773.6A ?
 
It depends on where, if at all, the three phase delta is grounded.
If ungrounded, the ground fault current from s single fault will be zero.
If corner grounded, the fault current could be 10,000A.
 
timm333 said:
Thanks, this is all very good information! This is a feeder for a VFD and ZSCT is required to operate Multilin relay. So is 2000:1 (50:0.025) the only standard ratio, or are other ratios (like 2000:5) also available?
Click to expand...

50:5 and 50:0.025 are the only two CTs that are commonly used for zero sequence applications. CTs are available in many other ratios, but the other ratios are generally not used for zero sequence applications. For more information on available CTs, this is a good reference from one manufacturer: http://www.gedigitalenergy.com/products/brochures/OEM_Guide.pdf


timm333 said:
Also if we have a 100A circuit fed by a 100A breaker, and the circuit has a short circuit current of 10000A: then what would be the theoretical maximum ground fault current. Will it be 100/1.732=57.7A, or will it be 10000/1.732=5773.6A ?
Click to expand...

The ground fault short circuit current has to be calculated for the particular application just like the 3 phase fault current is calculated for each application. There is not a simple way to determine ground fault current if we know the 3 phase fault current. In general the ground fault current can range from 30% of the 3 phase fault current up to 110% - a wide enough range that selecting and setting ground fault protection requires a calculated ground fault short circuit value.

With resistance grounded systems, where a resistor is inserted between the transformer X0 terminal and ground, the ground fault current will be determined based on the system line to neutral voltage, the resistance value of the resistor, and ohms law. Generally a 50:5 CT is used with solidly grounded systems and any resistance grounded systems that have a high enough fault current for the relay to sense the ground fault current. The 50:0.025 CT is applied when the 50:5 does not have sufficient sensitivity to detect the available ground current.
 
The 50:0.025A CT?s (for example GE part number 140-SD-13526) and 50:5A CT?s (for example GE part # 143-500) have 600V insulation. As they only contain ground fault current, so can they be used on 4160V high resistance grounded (25A) system?
 
600V window type CTs are used in MV switchgear and motor starters when the cable or bus insulation that passes through the window CT is rated for the full voltage of the primary conductors. The CTs you listed would be appropriate. As another example, GE/ITI series 780 CTs, rated 600V, are commonly used for phase CTs on several lines of MV switchgear.

One of the few places a 600V CT is not used in common MV equipment is for the phase CTs in some motor starters. On some starters the field cables are landed on one pad of the bar type CT and the cables from the load side of the contactor are on the other side. In this case the CT primary has line voltage on it and must be rated appropriately.
 
I would size the CT so that the maximum ground-fault current, divided by the CT ratio, does not excced the input rating of the Multlin. For example, if the Multlin can take up to 100A on the ZSCT input channel, then a 50:5 CT is will not overload the relay input channel for any ground fault up to 1,000A. It's a good idea to make sure you don't saturate the CT for all currents up to the available ground-fault current. This won't be too hard to achieve because the electronic relays have such low impedances compared to the old electromechanical ones. You need to know the input channel impedance. If the CT leads are long (say, the CT is not in the same cubicle with the relay), you'll need to account for them too.
 
Status
Not open for further replies.
Top