Get down to the basics, electrical power transformer has a more than one winding connected to an electrical system. The primary winding has the inflow of load currents whereas the secondary winding/s have the outflow of currents normally to the loads in spur connections. As the vectorial sum of the currents of a power transformer is zero based on the infamous Merz Price principle, it can naturally be protected by comparing the inflow and outflow of currents. If there is no fault in the transformer, there would be the same magnitude of per unit full load currents passing through the current transformers on the high and low sides of the transformer and there will be no residual current measured by the differential protection relay.
Protection engineers should make a distinction to the high impedance protection scheme with low impedance although they work around the same Merz Price principle. In a low impedance protection scheme, the differential protection can have the protection characteristic set typically with a two slope restraint characteristic.
So how does this sloped characteristic works? To understand the operating principle, consider a case when the differential protection relay sees a low fault current with a significant residual current by direct measurement. These fault currents should give a restraint current lower than the IRES setting and provided that the setting of SLOPE 1 is less than 100%, the compensated operating current internal to the protection relay will be “scaled down”. This causes the compensated IOP forced to be smaller than the measured IOP. This will cause the comparator to have a greater tendency to be logical TRUE condition. This makes the differential protection with an improved sensitivity to operate the differential protection. In typical applications, the unbalance current as seen by the protection relay for a fault of this nature is always considered as a TRUE fault. To improve the security, the SLOPE1 setting is always set to the maximum percentage error on the current transformers (plus a margin) caused by the saturation of CTs during fault.
On the other hand, the SLOPE2 setting could be set higher than 100%. If the transformer is subjected to a through fault with a significant pass through currents, the differential protection relay will observe high currents on high or low sides. The restraint current will exceed the setting of IRES causing the compensated IOP to have a tendency of higher than the measured IOP. The logical output of the comparator is forced to be a FALSE condition and the protection relay is resilient to trip under this condition. The SLOPE2 setting is said to improve the security and reliance of differential protection on a external fault to the transformer. One can say that this restraint characteristic improves the reliability of the protection relay by making the differential settings higher when there is a high pass through currents and lower on a low pass.
Unrestraint differential protection elements also exist but they simply measure the difference of per unit operating current on high and low sides of the transformer for comparison. Without the influence of the slopes of the restraint characteristic, the unrestraint protection can be made to isolate an internal fault quickly.
(Click Image To Enlarge)