Mar 14, 2019

What everyone should know about ground fault protection


EARTfelt Welcome dear friends of protection and control engineering, Today we show you what every protection engineer should know about earth fault protection. We assume that the basic things such as symmetrical components, zero-sequence voltage and the functioning of the ground-fault protection are already known.

What is it about today?

We present the stator ground fault protection in a power plant block and its coordination with the connected transmission network. These basic relationships should have been heard by every protection technician, regardless of the field in which he works.

All those who want to switch off now and call: The coal exit has been decided who needs KnowHow in the field of power plant technology is said: Even if the coal exit is decided in many industrial nations, globally considered it is far far away.

There are currently more than 1,400 coal-fired power stations worldwide and more than 150 nuclear power plants in the planning and construction phases. In total, we are talking about additional 800 gigawatts of conventionally installed generating capacity. Globally, we will have to do a lot of years in this area, especially for European companies that are mainly abroad.

So let's go:

In the classic block circuit, which is always present when each generator is connected to the grid via at least one block transformer, pure zero voltage monitoring is used at the open delta winding. Since the displacement of the voltage in the event of an earth fault only occurs on the low-voltage side of the block transformer, it can be used as a selective criterion.

The problem arises when a ground fault occurs in the connected transmission network. In no case may this cause the operation of the plant's own earth fault protection to be triggered and tripped. Unfortunately, earth faults on the grid side cause disturbances in the zero sequence system of the power plant side.

Despite the galvanic isolation between the upper and lower voltage side of the block transformer, a coupling takes place on the generator side. This is due to the coupling capacities of the unit transformer, which are usually between 5 to 20 nF.

To dampen this disturbance, the zero system is loaded with a high-impedance resistor. There are two variants for connecting this damping resistor:

🌐 Earthing of the generator star point via a single-phase neutral transformer

🌐 or the grounding of the low-voltage side of the block transformer by means of a 3-phase grounding transformer.

Farther on, we must remember that the zero voltage is a function of the fault location in the generator stator core. The further we go in the direction of the machine star point, the lower the tension zero system becomes.

In order to be able to reliably detect ground-faults in the stator of the generator, we would have to choose a very small pick up value close to 0 volts. But now we got to remember:

The earth fault on the high voltage side couples a zero system voltage into the area of ​​the power plant. We have to calculate this disturbance in advance and to over calculate it with a factor of 2. As part of a so-called earth-fault travel, the configured setting values ​​are then tested in practice.

Why is it called "90 % stator earth fault ?"

The usual protection zones for the stator earth fault protection of the generator vary between 80 and 95 % of the stator winding, and the setting values ​​are therefore 5 to 20 V. That's were the term "90 % stator earth fault protection" came up. This 90 % stator earth fault protection function is unable to detect star point earth faults within the last 5% of the stator.

What happens when such a star point ground fault occurs?

As a matter of principle, errors close to the star point practically do not occur. Reason is the reduced voltage stress in the neutral point area. If such errors occur in practice, then mechanical intervention is certainly the cause. For this reason, it is dispensed with on smaller machines every now and then.

However, in our VGB Standard "Electrical Generating Unit Protection" (VGB-S-025-00-2012-11-EN), we also recommend that machines under 5 MVA be equipped with an additional 100% stator earth fault protection.

The 100% stator earth fault protection detects earth faults in the entire block area, even if the fault in question is at the neutral point. In practice, two different measuring circuits come into question here:

🌐 The 100% stator earth fault protection with 20 Hz method

🌐 and the 100% SES based on the 3rd harmonic.

The clearly more complex but technically safest method is the 20 Hz method. It operates independently of the current operating point of the generator, also detects stator earth faults when the machine is at a standstill and is also completely immune to short circuits on the high voltage side.

In addition, earth faults are also detected at the generator terminals and at galvanically coupled resources, such as e.g. on voltage transformers or on the excitation transformer.

The active principle is based on the supply of a 20 Hz current lying below the fundamental frequency. This current comes in case of an earth fault, via the fault resistance to flow. The measuring circuit can be connected either via the star-point transformer in the star point or with the aid of a grounding transformer to the terminal leads.

The second variant based on the third harmonic is not so common in Germany. Not every machine qualifies for this procedure because not enough harmonics are always produced in the individual generator. The function of the measuring method is as follows:

Depending on the design of the poles of a generator, 150 Hz harmonics are produced in the stator winding. Since these 150 Hz components of the three phase voltages are in phase, their absolute values ​​add up in the zero system.

The following phasor graph illustrates this relationship with respect to the 50 Hz fundamental component, which is nearly canceled in the zero system and under symmetric conditions. If the machine in question now produces a sufficiently large third harmonic, it can be comfortably found in the zero system.

When a ground fault occurs in the stator winding of the generator, the distribution of the parasitic capacitances changes. This influences the measured variable.

When a ground fault occurs, the voltage of the 3rd harmonic decreases in the neutral point (breaks down to 0 V) ​​and rises at the generator terminals. Depending on the location of the measurement, it is thus possible to detect ground fault occurrence with a low or overvoltage function or with a voltage comparison.

The big disadvantages of this measurement method are:

🌐 Not every generator generates sufficiently large harmonic components

🌐 Furthermore, the 3rd harmonic is both active and reactive power dependent and thus dependent on the respective operating point of generator, so that in the context of the commissioning a load run test is required to verify the pick up value. In the worst case one notes in the context of the load test run, that there is no suitable setting value, since the machine produces too little 150 Hz.

For these reasons, the 20 Hz method is the better choice in most cases, although it is a bit more complex in the implementation.

Learn how we carry out the commissioning of complex generator protection systems, reading our book "The Bible of Generator Protection Engineering"

Learn how we calculate the coupling via the high voltage fault in detail and how we calculate the ground fault protection, parameterize and put into operation. In our onsite course "Generator protection - basics, concepts and commissioning"  from 25 to 27. November we will discuss this an even more (The course will be held in german language).

Thank you for reading and see you soon!