Current Transformers on neutral for electrical meter

[MJ1963]

I've worked with power meters that have typically 3 cts on each of the phases. When would a CT on the neutral bus be needed? Could someone clarify?

Thanks.

 

[GoldDigger]

Mathematically, there is never a need for a CT on the neutral for metering, since the neutral current must be the vector sum of the line currents.
The only thing a neutral CT in addition would be needed for would be for detection of a ground fault.

 

[MJ1963]

thanks.

if installed, it should read 0 correct?

 

[winnie]

The current on the neutral should be whatever is necessary to balance what is running on the 'hot' conductors, and the CT on the neutral should read this.

The vector sum of all 4 CTs (assuming a wye system) should be zero.

For metering, you can use any 3 of the 4, because you can always calculate the 4th value.

So the presence of the 4th CT is for some reason other than metering.

-Jon

 

[Rampage_Rick]

What I can't figure out is how a 2-element meter can properly measure 120/208Y 4-wire service. How do you zero-sum 4 wires with only 2 CTs?

 

[GoldDigger]

I cannot figure it out either. It would work if the utility supplied only the three wires of a delta, even if the service was derived from a wye.

 

[meternerd]

What I can't figure out is how a 2-element meter can properly measure 120/208Y 4-wire service. How do you zero-sum 4 wires with only 2 CTs?

You're right. A two element 3 phase meter (Form 5S) is for 3W Delta only. It uses two CT's and two Voltage inputs plus neutral. To measure 3 phase 4 Wire WYE, you need at least a 2 1/2 element meter (Form 6S). That uses three CT's and two Voltages plus neutral. It assumes the third voltage is the same as the average of the other two. Old technology, though. It was used to save the cost of one Voltage transformer. Now we use 3 element meters (Form 9S) for 3 phase 4W Wye and (Form 8S) for 3 phase 4W delta. Search for "Kwh meter Form diagrams" if you want to see how they're wired.

 

[meternerd]

Mathematically, there is never a need for a CT on the neutral for metering, since the neutral current must be the vector sum of the line currents.
The only thing a neutral CT in addition would be needed for would be for detection of a ground fault.

Yup...no need for a neutral current in metering. If used as a ground fault detector in a protection scheme, it would be unusual to use the utility meter CT's for the phase overcurrent sensing. Most (probably all) utilities do not allow other equipment to be supplied by metering CT's. Adds to the burden and could affect accuracy. If the customer has separate fault protection, it's normal to include both phase and neutral currents. It's also common to use phase imbalance sensing, either electronically or with a CT around all three phases, as a ground fault detector on loads that are normally balanced, such as motors or substation feeders.

 

[ron]

Power quality metering will use a neutral CT, since harmonics add up and may want to be measured.

 

[kwired]

I've worked with power meters that have typically 3 cts on each of the phases. When would a CT on the neutral bus be needed? Could someone clarify?

Thanks.

What exactly is being metered? Many replies are assuming this is for a POCO Watthour meter, if that is not the case or there is more ability then just measuring energy consumed, there may be good reason, ability to measure neutral harmonics was also a suggestion.

 

[Rampage_Rick]

You're right. A two element 3 phase meter (Form 5S) is for 3W Delta only. It uses two CT's and two Voltage inputs plus neutral. To measure 3 phase 4 Wire WYE, you need at least a 2 1/2 element meter (Form 6S). That uses three CT's and two Voltages plus neutral. It assumes the third voltage is the same as the average of the other two. Old technology, though. It was used to save the cost of one Voltage transformer. Now we use 3 element meters (Form 9S) for 3 phase 4W Wye and (Form 8S) for 3 phase 4W delta. Search for "Kwh meter Form diagrams" if you want to see how they're wired.

I found this diagram showing a 2 element Form 5S meter on 4-wire 120/208Y. Maybe that's what I saw? (I watched them swap in a smart meter at one of our facilities and I swear they only connected two CTs)

 

[GoldDigger]

I found this diagram showing a 2 element Form 5S meter on 4-wire 120/208Y. Maybe that's what I saw? (I watched them swap in a smart meter at one of our facilities and I swear they only connected two CTs)

AHA! Yes, note that the middle line is wired in series through both of the other CTs. That will only work to the extent that the load is balanced. It avoids having to incorporate a sqrt(3) calibration factor into the meter, and it will still give an approximately correct answer with a small level of imbalance. But IMHO it has to be incorrect for different cases of loads on one of the phase lines or phases only.
Just as the normal 120/240 3-wire meter assumes that the two line to load voltages are equal, this setup makes assumptions about the loads themselves.
In particular, this circuit apparently assumes that the loads are either delta or balanced, so my earlier qualification about two CTs being sufficient for delta wiring takes effect. If you tried to do the same thing with a wye system, you would lose big.

 

[meternerd]

AHA! Yes, note that the middle line is wired in series through both of the other CTs. That will only work to the extent that the load is balanced. It avoids having to incorporate a sqrt(3) calibration factor into the meter, and it will still give an approximately correct answer with a small level of imbalance. But IMHO it has to be incorrect for different cases of loads on one of the phase lines or phases only.
Just as the normal 120/240 3-wire meter assumes that the two line to load voltages are equal, this setup makes assumptions about the loads themselves.
In particular, this circuit apparently assumes that the loads are either delta or balanced, so my earlier qualification about two CTs being sufficient for delta wiring takes effect. If you tried to do the same thing with a wye system, you would lose big.

Sounds reasonable....as a utility meter nerd, I would never use a 5S meter for a 4W Wye service. I've never seen one wired that way. CT's are cheap, so why bother? That could result in inaccurate measurement and over or under billing. Not good for either side. Most CT services nowadays use bussing instead of individual wires through "donut" CT's, so creative wiring is not an option. But I can't comment on uses other than Kwh metering. Power quality metering is a whole different animal. Our PQ analyzers did come with 4 clamp on CT's. Maybe that's why. I can see where the neutral current could be useful info for upsizing the neutral for solid state loads, etc., but it would have nothing to do with revenue metering. Some newer solid state Kwh meters can record harmonics, but they don't use neutral currents. Unless there's some out there that I've never seen. I used to think I knew everything, but my wife straightened me out!

 

[meternerd]

[Just as the normal 120/240 3-wire meter assumes that the two line to load voltages are equal, this setup makes assumptions about the loads themselves.]

GoldDigger, "line to load voltages"? Not sure what you mean by this? A Form 2S (standard 240 3W 200A) residential meter is 240V and does not require line to neutral voltages (if that's what you meant) to be balanced in order to meter accurately. The two line to neutral voltages always add up to line to line, regardless of imbalance. The voltage coil is line to line across the top jaws. The two current coils are just aluminum bus bars top jaws to bottom jaws. Don't want to hijack this thread, though. Just curious.

 

[GoldDigger]

[Just as the normal 120/240 3-wire meter assumes that the two line to load voltages are equal, this setup makes assumptions about the loads themselves.]

GoldDigger, "line to load voltages"? Not sure what you mean by this? A Form 2S (standard 240 3W 200A) residential meter is 240V and does not require line to neutral voltages (if that's what you meant) to be balanced in order to meter accurately. The two line to neutral voltages always add up to line to line, regardless of imbalance. The voltage coil is line to line across the top jaws. The two current coils are just aluminum bus bars top jaws to bottom jaws. Don't want to hijack this thread, though. Just curious.

Yes, I meant line to neutral voltages.
The problem is that the common meter measures both line currents and assumes that each current corresponds to one half of the line to line voltage. The neutral does not even connect to the meter.
Imagine that there is a severe voltage drop problem in the service wires.
With a one sided load the line voltages at the meter might be 110 and 120.
The meter would calculate as if the power delivered was at 115 volts instead of 110 volts. The error is small, but it will always be in favor of POCO.
If the voltage imbalance at the meter is inherent in the distribution and not related at all to local load, the error can favor either POCO or customer, but will statistically average out to zero over the customer base.

 

[meternerd]

Yes, I meant line to neutral voltages.
The problem is that the common meter measures both line currents and assumes that each current corresponds to one half of the line to line voltage. The neutral does not even connect to the meter.
Imagine that there is a severe voltage drop problem in the service wires.
With a one sided load the line voltages at the meter might be 110 and 120.
The meter would calculate as if the power delivered was at 115 volts instead of 110 volts. The error is small, but it will always be in favor of POCO.
If the voltage imbalance at the meter is inherent in the distribution and not related at all to local load, the error can favor either POCO or customer, but will statistically average out to zero over the customer base.

One thing to think about is that residential meters measure KWh. Utilities supply KVAh. Any power factor less that unity results in more current being supplied than what is measured. Probably makes up more than the difference. In your example, if you have a service that's 120V @ 30A and 110V @ 50A, (for an hour) the meter would not register 9.2 KWh only if PF was unity. Actual would be 3.6 KVAh + 5.5KVAh = 9.1KVAh. But at a .989 PF, the KWh would be 9.1 KWH. Actual PF is probably lower than that on most services. But, too early in the morning to think about this stuff. Gotta go get another cup of coffee.

 

[meternerd]

OOPS...Edit

OOPS...Edit

One thing to think about is that residential meters measure KWh. Utilities supply KVAh. Any power factor less that unity results in more current being supplied than what is measured. Probably makes up more than the difference. In your example, if you have a service that's 120V @ 30A and 110V @ 50A, (for an hour) the meter would register 9.2 KWh only if PF was unity. Actual would be 3.6 KVAh + 5.5 KVAh = 9.1KVAh. But at a .989 PF, the measured value would also be 9.1 KWh. Actual PF is probably lower than that on most services. But, too early in the morning to think about this stuff. Gotta go get another cup of coffee.

I went past the 15 minute edit time (had to wait for my coffee to get done) , so it wouldn't let me fix my typo.

 

[GoldDigger]

One thing to think about is that residential meters measure KWh. Utilities supply KVAh. Any power factor less that unity results in more current being supplied than what is measured. Probably makes up more than the difference. In your example, if you have a service that's 120V @ 30A and 110V @ 50A, (for an hour) the meter would not register 9.2 KWh only if PF was unity. Actual would be 3.6 KVAh + 5.5KVAh = 9.1KVAh. But at a .989 PF, the KWh would be 9.1 KWH. Actual PF is probably lower than that on most services. But, too early in the morning to think about this stuff. Gotta go get another cup of coffee.

The CT in the meter, or the current coil in a series connected meter, will always detect current, but they will include the timing information as well, so that the meter can record the vector product of that with the voltage.
Industrial meters also do this for the proportional part of the bill. The difference is that they also record the maximum PF (or the maximum KVA) for separate use in assessing a non-proportional penalty or fee as part of the monthly bill.

 

[meternerd]

The CT in the meter, or the current coil in a series connected meter, will always detect current, but they will include the timing information as well, so that the meter can record the vector product of that with the voltage.
Industrial meters also do this for the proportional part of the bill. The difference is that they also record the maximum PF (or the maximum KVA) for separate use in assessing a non-proportional penalty or fee as part of the monthly bill.

I was kinda talking about mechanical meters, but electronic are similar. PF, VAR, demand, etc. are not recorded by the average $60 residential meter. I assume you're talking solid state meters. I'm not sure how the meters calculate watts, but it can't be too complex at that price. We test the KWh meters at 50% power factor and they must still meter to less than 0.2% error, so it works, however they do it. Three phase commercial are 4 quadrant meters which can record as much as you want to pay for. They all record gobs of values, and you pay for firmware codes to "turn on" various features. But they aren't cheap. Usually $400 plus. Most utilities I know of bill KWh and add charges for KVARh and KW peak demand. The KWh rate is usually lower than for non-demand customers. Since ours is a small utility with very little large commercial load, we don't bill for KVARh. Our system PF is pretty close to unity, so we pay no penalties to our suppliers for poor PF, thus no charge to the customer. We do add PF correction equipment if the PF is too bad at any particular location.

It's always been my opinion that the industry should bill based on KVA, since that's really what we sell. That would eliminate all the other junk as far as PF. But it's been KWh since metering was invented, and they aren't about to change.

 

[GoldDigger]

Mechanical meters make use of induced currents in the aluminum disk to provide a force on the disk which is proportional to the product of the instantaneous current and the instantaneous voltage. With the resistance supplied proportional to the speed of the disk, you now have the disk rotating at a speed proportional to the kW, not the kVA.
and without any additional mechanical devices in the dial chain, the disk will rotate backwards during reverse power flow (such as from local generation.)
And what the utility sells is energy produced by its generators, and a higher kVA delivered to the customer does not cost more fuel. It just requires larger distribution lines.