Jul 21, 2017
 in 
Ground-fault

Protection-Revolution?

W

ELCOME Dear friends of protection and control engineering! Is there a protection revolution for us in the closer future? This question can be answered "Yes" immediately at the beginning of the article. We also have many indications of the direction in which the technical solutions are developed. The systems are adaptive, a high degree of digitization is on the agenda, we have Goose, Sampled Values ​​and the IEC61850. But do we really have everything on the screen? Today we present you a technically highly interesting solution, which certainly has not been heard by everyone.

We will take you on a protection journey to the other end of the world. It goes to Australia. Even though the summer is slowly starting up in Germany, we are talking about quite different temperatures in many parts of Australia. But in Down Under it is not only significantly warmer than in Central Europe, it is also much drier. And so we are already in the middle of the topic: the vegetation completely dried up is very easily flammable and thus triggers rapidly devastating bush fires, which represent a high danger for humans and animals, but also for material values. Not infrequently such fires are caused by earth faults. There are already a few amps to ignite a fire (see videos below).

For this reason, in 2015, the Victorian government, together with the network operators, carried out numerous tests on the consequences of earth faults regarding the cause of bush fires. The following video shows very impressively how an earth fault with a few amps (in this case 18 to 24 A over 0.9 s) can ignite the dry grass.

From the experiments carried out, the following requirements were derived from the network in order to reduce arcs and their energy as far as possible:

🌐 Detection of fault impedances up to 25 kOhm

🌐 Error clarification within a maximum of 85 ms

🌐 maximum ground fault current of 500 mA

These requirements placed new challenges on protection technology. Several methods have been developed that are summarized under the well-known designation "REFCL" (Rapid Earth Fault Current Limiter). The following video shows the repetition of the experiment from the first video with a REFCL system. In this case, the earth fault current is reduced from originally 15 A to almost zero amperes within 55 ms.

One of these methods comes from our European colleagues from the Swedish company "Swedish Neutral" and is called "Ground Fault Neutralizer" (GFN). The GFN is based on the classical measurements of the phase variables, displacement voltage, zero currents and their transients in the fault entry moment. The difference to a conventional protection system is, however, that in case of fault detection, the faulty line is not switched off, but a current is supplied at the star point of the network which corresponds to the magnitude of the fault current, but is 180 ° opposite in its direction. Thus, not only the reactive component of the fault current, such as, for example, in quenched networks with the Petersen coil, but also its active component can be compensated for. The fault is cleared through this protection system in less than 3 net periods, ie under 60 ms, without interrupting the supply. Thanks to the integrated fault locator, the fault can also be quickly found and remedied by the operating personnel.

The system can be retrofitted in almost all networks from 6 to 100 kV nominal voltage, but does not provide protection for phase errors. The existing switches can be reused and controlled via the GFN cabinet and its remote control.

The manufacturer also promises an integrated partial discharge measurement, which monitors the condition of the cables in a cable network and prevents arcing before they ignite. The upper harmonics (up to the 7th) can also be detected and eradicated.

Does this system now also replace conventional protection in Germany? First of all, this system makes sense in fire-affected regions. For this reason, the urgency of retrofitting in our latitudes is rather low. Since an additional protection for phase errors must be installed anyway, REFCL is not the "Swiss Army Knife" anyway. Nevertheless, the GFN or generally REFCL systems have great potential and will certainly attract attention in the coming years. We remain excited and inform you how it goes!

Greetings Hannes Heiden and Alexander Muth