Current Transformers for High Primary Currents
When current transformers are used in heavy current applications, such as 4000 amps and above, the magnitude of the magnetic field effect from the primary conductor configuration becomes so significant that it requires consideration in the design of the transformers. Two phenomena require attention. Air flux associated with adjacent phase or neutral conductors can pass through part of the core and cause magnetic saturation at that point. Also, placement of the primary conductor away from the central axis of the primary aperture can cause uneven magnetisation of the CT core.
For detailed analysis, complex electromagnetic field calculations would be required for each application. However, published work shows that the stray flux caused by adjacent conductors may be calculated approximately from the formula:-
| B = | 10-6IP (R + 0.5W) | Log10 | (D + R) | Tesla |
| --------- | --------- | |||
| A | ( D - R ) |
where
W is the width of the core (m)
R is the outer radius of the core (m)
D is the distance between phases (m)
A is the cross section of the core (m2)
(RA Pfuntner, The accuracy of current transformers adjacent to heavy current buses. AIEE Trans., vol. 70, pp 1656-61, 1951)
Insertion of typical design figures shows that local saturation of the core material occurs with currents around 7000 amps when the primary conductors are reasonably spaced, and with lower currents where spacing has been minimised. For example, with a 4000 amp busbar assembly of 100 x 50mm bars at 115mm centres, local saturation would occur in a typical ring type metering CT at full load current and at 3 times overcurrent in a protection CT.
The effect of asymmetric placement of the primary conductor is similar to but not as severe as the adjacent conductor effect. Typically a 4000 amp CT might show a variation of 0.06% when the primary conductor is moved from the centre to the edge of the primary aperture. This can be greatly increased if the CT is not located on a straight run of the primary conductor, but against a right angle bend in the bar.
Though both of the above cause errors in accuracy, of much more significance is the long term damage caused by localised heating of the CT core, a situation which can be aggravated by heat from the resistive losses of a dense CT secondary winding, heat from the adjacent primary bars and the thermal insulating properties of the CT finish. It is also worth noting that stray flux effects are more pronounced when the fundamental frequency is higher. We have seen a conventionally wound 1500/5 amp 50Hz CT completely burnt out when used on a 2kHz induction furnace.
In order to overcome the problems outlined, it is necessary to provide the CT core with some form of magnetic shielding from uneven and stray flux effects. There are three methods in common use:
1. By shrouding the core and secondary winding in a heavy gauge U section copper shield.
2. By providing an additional flux equalising or balancing winding comprising a number of
equal winding segments distributed around the core and connected in parallel.
3. By winding the secondary winding, all or in part, as a flux equalising winding of at least
four sections comprising a pair about the horizontal and a pair about the vertical plane to
the main CT axis and connected in parallel and series.
Method 1 is bulky and expensive. Method 2 is the simplest to produce but is not as efficient as method 3. It also can be overworked by severe conditions, resulting in overheating of the balancing winding. We favour method 3 and it is a matter of policy with very few exceptions that, to overcome the heating effects of stray flux, we wind all 50Hz current transformers at 4000 amps and above in this way. We also use this method for class 0.1 accuracy CTs at currents of 1000 amps and above to overcome variations in accuracy with primary conductor location. Our range of high frequency CTs for induction heating applications at 0.7-10kHz are wound with flux equalising windings above 1000 amps.
A consequence of this policy is that our high current CTs may be larger and more expensive than those wound as a simple single winding. From the foregoing it should be clear why we deem this necessary. The temptation to fit a conventional CT between closely spaced high current busbars should be resisted. If it is not possible to fit a flux equalised design, then the selection of a single winding design may only provide a short term solution.