The most frequent errors in specifying current transformers (CTs)

Proper specification of CTs

Current transformers (CTs) form an essential link within the protection chain of electrical networks. Their specification, even if it is handled by specialists, often includes errors and is insufficiently optimized.

This often leads to technological impossibilities, operating delays, extra costs, incorrect operation of protections and can even jeopardize the safety of installations and people.

The most frequent errors in specifying current transformers
The most frequent errors in specifying current transformers (on photo: CTs installed in MV switchgear type MCset 1250A, 10kV, 31kA; credit: Edvard Csanyi)

Proper specification of CTs requires sound knowledge of:

  • Electrical installation diagram,
  • Electrical data (voltage, nominal current, short-circuit current, etc.),
  • Associated protections,
  • Overall network protections (protection plan, the load that they represent for the CTs, as well as wiring and their settings.
Often, due to lack of data or even ignorance of how a CT shall be used, a CT manufacturer says “these features are not feasible”, while a standard CT may be suitable.

Common engineers errors

Many errors normally lead to oversizing of the CTs which increases costs and can be dangerous. Many CT definition errors stem from ignorance of their operation and from the unknown or incomplete characteristics of the network component to be protected and of the associated protections.

The better informed the CT manufacturer, the fewer errors and the more the CT will be optimised.


  1. Protections and conventional CTs
  2. Differential protections and class X

Protections and conventional CTs

For these protections which do not require class X defined CTs, the most frequent “errors” are:

Error #1 – Different accuracy limit factor or classes…

Using two CTs or one CT with two secondaries for two protective relays the manufacturers of which recommend different accuracy limit factor (ALF) or different accuracy classes.

As the CT manufacturers can translate 10P to 5P (according to the corresponding induction levels), and can move from one ALF to another by adjusting power, they can find a CT matching the needs of both relays.

Go back to contents ↑

Error #2 – Adding up the powers…

Taking into account the wiring resistance although the protection manufacturer has already integrated it into the needs formulated for the CT. Let us take an example of two relays whose technical data indicate for 1 A CTs:

  • CT1 for relay 1: 5 VA 10P15 (assuming 2 RL < 1.5 Ω),
  • CT2 for relay 2: 10 VA 5P15 (assuming 2 RL < 2 Ω).

A single CT may be suitable for both relays: in theory a 10 VA 5P15. You need to:

  • Avoid adding up the powers (5 + 10 VA) required for each relay. In point of fact, for the CT2, relay 1 only represents a load (just like wiring) and vice versa.
  • Check, in this case for the CT2, that: 2 RL + Rp1 ≤ 2 Ω;
    and if CT1 was selected, that: 2 RL + Rp2 ≤ 1.5 Ω. If this was not verified, the relay supplier can suggest that “x” VA be added per additional ohm.
Adding the specified power for several protections linked to an application results in CTs that are often impossible to manufacture or that jeopardise safety during short-circuits. Use of multifunction numerical relays avoids such errors. You only need to size the CT for the most restrictive protection.

Go back to contents ↑

Error #3 – Changing the required characteristics…

Changing the required characteristics without verifying the consequences.

Case 1 – Low ratio CT…

A CT manufacturer cannot make a low ratio CT and suggests increasing this ratio. Let us take an example:

  • Requested: 30/1 CT – 2.5 VA 5P20,
  • CT manufacturer’s proposal: 60/1,
  • With motor In = 16 A and minimum thermal protection setting: 40 % of CT In (i.e. 60 × 0.4 = 24 A).

The protection setting at 16 A (normal thermal protection setting at motor In), is then impossible. The solution is to increase the rating and lower the ALF requirement: 40/1 – 2.5 VA 5P10. This CT, feasible, allows the required setting (40 x 0.4 = 16 A).

Case 2 – Thermal withstand…

A buyer accepts a thermal withstand of 0.1 s proposed by the CT manufacturer instead of 1 s.

What is likely to happen is that, on a short-circuit, if the real fault duration exceeds 0.1 s, thermal and probably electrodynamic withstand will be insufficient and may result in CT destruction.

Go back to contents ↑

Error #4 – Lack of real information

Due to lack of information on real requirements. Let us take the following case, relatively educational: a CT with two primaries and three secondaries (200-1000/1-1-1) is requested with:

  • First secondary: 1 A, class X (given ),
  • Second secondary: 1 A 15 VA class 0.5 for metering,
  • Third secondary: 1 A 10 VA 5P20.
The supplier can propose a CT with three magnetic cores and secondary tappings to meet the 200 A or 1000 A need at the primary. However, such a CT is hard to manufacture because to obtain 15 VA class 0.5 and 10 VA 5P20 on 200/1 ratios, you need 5 × 15 VA class 0.5 and 5 × 10 VA 5P20 on 1000/1 ratios!

Moreover, the supplier must comply with class X for both ratios! In point of fact, class X concerns only the 1000/1 ratio (for busbar differential protection). The 200/1 ratios concern metering and the traditional protections (see fig. 1).

Example of poor understanding between the customer and the CT manufacturer
Figure 1 – Example of poor understanding between the customer and the CT manufacturer

The CT to be manufactured is then easier, less bulky, cheaper and definitely feasible. This example shows that the lack of information shared between those involved is a source of errors and of non-optimization.

A consultation that does not begin properly may result in a CT that cannot be manufactured!

Go back to contents ↑

Error #5 – Taking into account relay impedance…

Taking into account the relay impedance Rh for calculation of real load (see fig. 2) in the CT calculation for overcurrent or in the calculation for CTs in class X.

Internal and load impedances of a CT
Figure 2 – Internal and load impedances of a CT

A word of warning! Rh is only considered when calculating CTs for zero sequence current lh.

For high impedance differential protections, in the calculation of Knee Point Voltage Vk given by:

2 If (Rct +2 RL + Ra)

where Ra = other loads, Rh must not intervene.

This is the load of one phase (we assume that no current flows through the neutral). Vk is indeed calculated for relay stability conditions, i.e. no phase or earth fault, in the protected zone, no incorrect unbalance, therefore, in the differential connection I = 0 and the voltage of that connection = 0.

Knee point voltage defined by the BS 3938 standard: Vk for class X (PX in IEC 60044-1). Vk is determined by the point on the curve Vs (Im) from which a 10% increase in voltage Vs leads to a 50% increase in the magnetizing current.

Go back to contents ↑

Differential protections and class X

For these applications, the most usual errors are:

Error #1 – CTs with the greatest knee point voltage…

Asking the CT manufacturer to supply CTs with the greatest knee point voltage Vk that he can build using a standard mould. This occurs when the differential protection relay (make, type) is not defined.

There are three consequences:

  1. Overcost,
  2. Possibility of high overvoltages and overcurrents at the CT secondary which can lead to destruction of the circuit and the relay,
  3. With no requirements for the CT secondary winding resistance Rct, it is not certain that the Vk expression corresponding to the relay used, will be complied with.
To illustrate this case, let us take the example of a high impedance busbar differential protection. The CT supplied is a 2000/5 where Vk = 400 V and Rct = 2.5 Ω.

For the relay used, the expression to be satisfied is: Vk ≥ 200 Rct + 20, i.e. 520 V. The Vk = 400 V is not sufficient!

More serious still, the requirement of too high a Vk may lead to the manufacturing of a non-standard CT (see the first two consequences above) requiring a specially designed stabilising resistance and an overvoltage limitor as well as the use of a deeper panel!

Go back to contents ↑

Error #2 – Error on the through current…

Error on the through current is very common. Let us take the example of a high impedance differential protection where the switchboard Isc is taken into account instead of the maximum through current. The aim is to protect a motor, the CTs have a ratio of 100/1.

  • Result obtained with the through current (7 In of CT):
    Vk ≥ 14 (Rct + 2 RL).
  • Result obtained with switchboard Isc (Isc = 40 kA):
    Vk ≥ 800 (Rct + 2 RL)
It is not necessary to go into too much detail to understand the importance of choosing the right parameter!

The table 1 gives the through current values to be taken into account when the through current is the CT calculation base.

Determining the through current properly
Table 1 – Determining the through current properly

Go back to contents ↑

Error #3 – Pilot wires in the calculation…

With line differential protections, taking into account the pilot wires in the calculation of Rwiring. In point of fact, RL is given by the wiring linking the CTs to the relay located on the same side (end) of the line (see fig. 3).

You must not take into account the length of the pilot wires which run from one end to the other of the protected line.

RL is given by the wiring between the CT and the relays located on the same side of the line
Figure 3 – RL is given by the wiring between the CT and the relays located on the same side of the line


With respect to high impedance differential protections:

  • For the calculation of min. Vk, take account of the through current (see Table 1).
  • Calculation of the stabilising resistance Rst is a function of min. Vk and of the relay setting current.
  • Calculation of peak voltage (Vp) is based on the internal Isc of the protected zone and on the real Vk of the CT.

Go back to contents ↑

Reference // Current transformers: Specification errors and solutions by P. Fonti (Schneider Electric)

SEARCH: Articles, software & guides //

About Author //


Edvard Csanyi

Edvard - Electrical engineer, programmer and founder of EEP. Highly specialized for design of LV/MV switchgears and LV high power busbar trunking (<6300A) in power substations, commercial buildings and industry fascilities. Professional in AutoCAD programming. Present on

One Comment

  1. Faisak
    Apr 26, 2017

    Thank you very much Eodvard , I study Electrical power system . I am very intersting and connect always your webset it is very helpful for me , so i want to friend with you because you are expert in my field .
    best regards

Leave a Comment

Tell us what you're thinking... we care about your opinion!

Premium Membership

Premium membership gives you an access to specialized technical articles and extra premium content and many more...
Get Premium Now ⚡