Author: Walter Schossig, Germany
Overcurrent Protection
The main element of generator protection was overcurrent protection from the beginning. It protected the generator of inner damages, and was also the backup protection for all further assets such as transformers and lines. Smaller machines were equipped with direct overcurrent release, connected with a lever (Figure 13; Figure 15). The first stand-alone electromechanical relay was designed in1904. Figure 18 shows the first time-overcurrent relay made by ASEA, type TCB, manufactured in 1905. ASEA's (now ABB) first relay of induction type was delivered in 1912 to a hydropower station in the north of Sweden, built to deliver power at 16 2/3 Hz to the railway from Kiruna to Luleå, which was built to transport iron ore. This was the first electrified railway in Sweden.
Overload Protection
The use of thermal relays for protection of generators was introduced exclusively by BBC in Europe and was very successful.
German recommendations required generators with a nominal power of more than 5000 kVA to use six resistancethermometers or thermo elements in the stator to supervise the temperature of winding. Once they were installed it was very difficult to reach them again- that's why they very often were not changed after damage. So it was the decision to supervise the temperature with thermal relays. These devices are equipped with thermo elements (heating relays with current proportional to main current delivered an image of temperature of the machine- Figure 14.
The outer insulating mat "O", was working as a protection against thermal radiation. It encloses the source of heat, the measuring element and the heat storage "P". The heating element - a band made from a resistive material- heated the measuring column and the heat material storage, consisting of changeable measuring boards. The thermal time constant of the relays could be changed with the numbers of boards in 6 stages between 20 and 110 minutes. The upper scale was a display of the temperature to allow a later estimation of temperature. Additionally the relays were equipped with a tripping device.
Later single pole overcurrent relays with a setting of 1.1 In and 10 s have been used for indicating overload.
Short Circuit Protection
To use overcurrent relays between generator and busbar in case of failures inside the generator was active only if other machines were able to deliver short circuit currents. The power of another machine had to be as big as the power of the machine to be protected. In case of its own failure the generator delivered a huge short circuit current but this could not be detected with such a setup. That's why overcurrent relays have been installed in the neutral point of the generator. This setup was the only possibility if the generator was the single source in the grid. Figure 21 shows a three-phase overcurrent protection "S", manufactured by SIEMENS in 1936.
Connecting generators and grids in parallel increased the reliability of power supply but on the other hand created unmanageable short-circuit currents in case of failure. Sometimes the unreasonable guaranties for small voltage drops in the machine required the use of coils for the limitation of the current from generators with high short circuit power (hard machines). Later "soft" generators were used, equipped with a huge short-circuit reactance. The Swiss grid reported good experiences with the usage of a cutback of excitation to achieve a limitation of a sustained short-circuit current. All big power stations have been equipped with such devices. The automatic decrease of excitation could limit the steady-state short-circuit current to a certain value (as 1.4*Inom). After switching off the short circuit the voltages recovers. Out of step machines could be "catched". The current level in case of a short circuit sometimes reached the level of overload current.
Nevertheless, in case of a malfunction of the line protection the generator overcurrent protection should trip. Solowjew,L.E., USSR, proposed in 1932 to use "undervoltage supervision". This allowed a more sensitive setup of the overcurrent startup. An overcurrent relay produced by BBC in the 1960's is shown in Figure 22.
The overcurrent setting must be over the highest possible operating current. Since these devices started up with a fault in the grid already the operating time must be the biggest in the grid. If several generators are working in parallel all overcurrent relays had the same operating time. In case of terminal short-circuit of a generator there was no selectivity anymore. This was only possible if the faulty generator could be tripped faster than the other ones.
Further solutions such as reverse power protection, differential protection, interwinding fault protection and earthfault protection will be covered in a later issue of this magazine.
walter.schossig@pacw.org
 |
 |
|
 |
|
 |