Your question is confusing. What exactly do you propose to do? Could you post a schematic?

Joshi, As requested.

My commonsense point of view: it is doable. You have to be very careful about the phasing of the C/T1 & C/T2. I must admit, at the outset that I am no expert. Please have patient & wait for the subject experts to comment/share their views. By the way, what is the supply voltage? If you follow the normal practice, you have to pay for one more meter only. Is it not worth it?

This is sheer nonsense, . . . one would NEVER EVER connect the secondaries of CTs in series.

Why not put only one CT after the switches on load side - on the righ hand side of your diagram near the arrow? Curernt is the same anyway. Why should you put 2 CTs and land into more trouble?

There are two feeders, i.e. utility & generator, each having one CT, and both CT are connected in series with correct phasing of CT secondary winding. The meter will read total load current for one phase. The sketch shows only one phase. Other two phases are wired similarly.

Repeat. This is dangerous and sheer STUPIDITY . . . do not connect CT secondaries in series . . . NEVER. CT secondaries may be connected in parallel ONLY.

Well olehwi thanks for the very efficient answer which tells nobody anything useful. If on the other hand the answer was as given by learner42 which details WHY it should not be done then all of us have learned something from it. Thanks and regards to all.

The real reason why NOT in series is:

Each one of the CTs will have the other CT as load in series with the meter (which meter is allways almost a Short circuit). Since the IMPEDENCE of the 2nd CT is significant, it will interfere with the accuracy of the reading and could be dangerous by elevating the voltage at the 1st CT & meter terminals, much more than normally expected!

In Parallel, both CTs are almost Shorted at ther terminals, via the metering device. No danger of high voltages and the accuracy will remain almost as expected from the class of CT used. {the 2nd CT's Impedence will reduce the bypass current into its winding significantly enough not to influence the reading very much}.

First, this is more complex and potentially dangerous than you might realize. If you are having to ask this question, it would be far far better for you to turn this project over to someone with specific experience in switchboard metering.

You must use a switch to switch between the two CTs. You ABSOLUTELY must also short the secondary of the unused CT after switching, you CANNOT leave it open-circuited. There are specially made switchboard switches that will accomplish this for you, but you must know how to read the target operations charts. Again, leave this to someone trained to do it. The consequences of getting wrong can be dire.

Why not just use a set of three phase CTs downstream of the switches and use a standard 3 phase meter?

You have said so yourself, they are interlocked so only one switch can ever be on, your connection method is not set up to monitor individual supplies, and it doesn't sound like this measurement is for the utility metering (otherwise they would have said no due to the non-standard, single-phase and potentially unsafe connection method).

Why would you want to do it your way rather than the common sense way I have mentioned above?

J.O.A.T this is only for panel metering (and it would be 3 phase but I only showed one for simplicity as it does not effect the question) and yes I totally agree that you would normally put the C/T on the common output line after the changeover system. The problem is that in this case there is just no space to put the C/T there, so I wondered if there was any way to achieve the same result by using series connected C/T. My original enquiry was just about what may affect the accuracy but now I am much more interested in why is it potentially dangerous ?? Any info or links to the explanation of the danger would be much appreciated. Regards and thanks.

Dangerous voltages will appear on the breaker terminals which is open since the CT secondary on this open breaker is not shorted. Let us assume it is a 2000/5 CT and let us assume a voltage drop of 10V in the CT secondary with the flow of current from the closed breaker. Theortically 10x2000/5=4000V will appear on the open breaker treminals.

I think that there could be a mistake here:

If the voltage drop in CT secondary =10V and the CT ratio is 2000/5 A, then:

V1/V2 = I2/I1 = 5 / 2000 = 1/400; (not I1/I2)

==> V1 = V2/400 = 10/400 = 0.025V or 25mV {between the contact of the open breaker, if any, depends on the remaining circuit...}

Therefore, there is no Problem except that you must make that the metering device is sound to prevent any possibility of an internal open circuit, which will bring about a huge voltage due to the addition of both CTs Open secondary voltages!

The parallel option is less stressful if this happens. (The device should be suitable for such operation ... Most are ?!)

You could place a resistor across each current transformer secondary.

Place the transformers in series as shown and use a voltmeter instead of a ampmeter to indicate the current.

Like it has been said by others, you have to know what you are doing. If you do it incorrectly, you may destroy the current transformers.

You can use two loop powered CT's.In this way you don't need to worry about the secondary being accidentally opened.http://www.burkert.com.au/ENA/4032.html.

You can do alot more with the 4-20 signals

Put the metering on the load side so that it is measured which ever feeder is on. It is not practical to add two CTs in series as the vector addition is confusing. Your purpose of getting the load from both feeders whichever is on , is not fulfilled.

Assuming the Secondary terminals of the two 2000/5A Current transformers are:-

CTI : CT2:

Spot terminal : 1s1 Spot terminal : 2s1

Non Spot terminal: 1s2 Non Spot terminal : 2s2

Connection:

Link 1s1 to 2s1 (Terminal A).

Link 2s1 to 2s2 (Terminal B).

The terminals A and B will then be wired to the AMMETER. This is the riight connection to do to achieve your desire for measuring the load currents from the UTILITY or Generator separately. (That is ,you connect the CT secondaries in PARRAREL and not in SERIES, taking note of the polarity of each CT as indicated above.)

The application is certainly plausible but you must use the proper/correct type of selector switch such as a GE SBM type.

The chosen selector switch must short circuit the output terminals of each CT being used in order to prevent the induced voltage from getting too high and faulting.

The switch must also short the CT output terminals of the circuit being monitored/displayed before switching to the next CT. (make-before-break)

All wiring must be correctly connected.

This is no different than a 3 phase ammeter switch circuit used to monitor and display current values from two different sources and display each value one at a time.

The CT's output need not be wired together in any manner as the switch provides the correct configuration of terminals.

Below is a link that is invaluable for obtaining answers to your questions.

www.geindustrial.com/solutions/​engineers/docs/sbcat.pdf

Make contact, get the correct wiring diagram, purchase the correct devices from the manufacturer of your choice, follow their direction and the results will be what you are after.

Use one current transformer and run both conductors through it.

GA

It could be the best solution IF BOTH CABLES COULD GO THROUGH THE CT's central hole!

You normally get the CT that is nearest to the size of the cable or BUSBAR you intend to use, so as to keep the losses very small.

Therefore a compromise in accuracy will have to be checked if the CT is selected to accomodate both cables!

qapmoc,

I agree about wiring them in parallel. A problem not discussed is accuracy at low loads. The higher the ratio, the better the accuracy of the CT. At low loads, the accuracy of a typical 200:5 CT is pretty poor, but if you used a 200:1 it would be very good. Also, the two sets of CT's would neeed to be the same ratio--not a 400:5 and 200:5 in parallel (for example).

JMM

This is not true. The accuracy of a CT, under the same burden conditions, depends only on the primary ampere turns , not on its ratio. Therefore a 200/5 and 200/1 CT wound on the same core will have the same accuracy under the same burden (VA) conditions.

Two 200/5 CTs with 1% accuracy, will provide 2% accuracy when paralleled. The same applies to two 200/1 CTs with 1% accuracy, they will provide 2% accuracy.

Friend,

Perhaps I didn't state this best. If you look at the technical specifications and the curve of accuracy vs. metered load and vs. burden, you will generally find that as the burden increases or the load decreases, the accuracy gets poorer. It is much easier to wind a CT for a specified accuracy class when the ratio is higher.

When you state: Therefore a 200/5 and 200/1 CT wound on the same core will have the same accuracy under the same burden (VA) conditions. , this assumes that the two CT's you are comparing have windings designed for the same burden.

If I am putting two CT's in parallel then the device they are connected to either needs to have an input rating equal to twice each one's individual output current (such as two 200/5 into a 10A FS meter), or the CT's must be selected with a ratio twice the expected load and be wired into a meter with the same input as either one's stated output. In the former case, finding an appropriate meter can be a problem (probably a custom scale will be needed). In the latter case, the error is doubled because the stated accuracy is a percent of the full load current the CT is designed to handle, not a percent of the actual current. (In other words, a 1% CT with a 200/5 ratio will have an output accuracy when loaded for less than its stated burden of +/- 0.1A, and when reading a 50A load, the accuracy as a percent of the actual load is only 4%.)

--JMM

You are basing your writings on your opinion and not facts. Put some numbers down and see if they confirm you beliefs.

Why don't you GOOGLE "Paralleling of current transformers for metering applications".

Sir,

You believe I am "basing your writings on your opinion and not facts." I studied this topic recently and spoke at length with one manufacturer of CT's and two manufacturers of third-party revenue meters. Before writing this reply I also did what you suggest. My statements are correct. The accuracy class of a CT is the percent of error when the CT is excited at its stated input current. Thus, a 100:5 CT with a 0.3% accuracy will have an error not more than 5*0.3/100A = 0.015A on its secondary when the primary current is 100A.

From http://www.digikey.com/Web%20Export/Supplier%20Content/CRMagnetics_582/PDF/cr-magnetics-pi-selecting-ansi-class.pdf?redirected=1 I copy the following statement: "It is critical to understand that the accuracy ratings are for a full scale reading. This accuracy will only be maintained from 30% full scale and up. Below this, and the accuracy worsens greatly. Always strive to select transformers so that the majority of readings will be within 20 to 100% full scale range."

Another reference I have read states that when putting CT's in parallel, the accuracy of the total installation is the weighted average of the accuracies of all the CT's, while the burden is the burden of the load multiplied by the number of CT's in parallel. This means that you would have to choose CT's with much higher burden capabilities or a meter circuit with much lower burden.

--JMM