EARTfelt Welcome dear friends of protection and control engineering. Today we are looking forward to an exciting guest article on "backup protection in open ring nets". Our current guest author Sören Sturm is Team Leader of Service Strom at wesernetz Bremen GmbH, an excellent specialist and also a longtime companion of mine.
Have fun reading, we pass on Sören Sturm.
Backup protection problem in open ring networks
Frequently medium-voltage networks are operated as open ring networks. Especially in urban areas these structures are very common.
The open rings are supplied from a substation or a center station. The protection of a half-ring takes place only from this point. Particular attention must be paid to compliance with the backup protection function of the feeding switch.
In our example, we feed the ring network via a 40 MVA 110/10 kV transformer. The 10 kV busbar then feeds several ring networks. The circuit breaker of the transformer takes over the backup protection function in order to back up a breaker failure of a fault on the half ring.
Each protection technician now should see what's the matter. The minimum short-circuit current, in the worst-case constellation of the open ring, must still be safely above the trip level of the transformer-sided over-current protection.
In the following example we determine the problem with real numbers:
From the substation, the half rings A and B are supplied directly. Half Ring A (right in the picture) is about 6 and half Ring B about 4 km long (1 km = 0,62 miles). In the worst case, the Ikmin 2-minimal is 3.5 kA (see figure).
This minimum short-circuit current is sufficient to trip the 10 kV transformer protection, even though the safety factor for back up protection should be better at 1.5 (in our case 1.25). In this constellation, it is still possible to assume a functioning backup protection for the 10 kV circuit breakers.
The most unfavorable case is usually given when the complete ring is fed in on one side, e.g. as part of a circuit breaker maintenance in the substation. Depending on the cable type and installed cross-sections, the minimum short-circuit currents at cable lengths of approx. 8-10 km may already be in the range of the rated current of the transformer.
In the following we see such a "worst-case" calculation for the previous example:
In the present maintenance state, the minimum short-circuit current is even below the operating current of the feed. Here, with a classic over-current function, no meaningful set value for the backup protection can be found.
Now somebody could say yes, OK, but there are other methods of detecting low-current short circuits, such as the impedance starting or the voltage-dependent overcurrent starting. However, in urban cable networks in particular, limits are also reached for reasonable setting values. In the substation, the voltage dip is often low. Load and error angles can often hardly be distinguished because of the cable impedances.
In the worst case, in our example, backup protection has data of the following order:
Ik = 2.1 kA
Zk = 2.06 Ω phi = 27 °
ULLmin = 9.1 kV
There are several ways to get this problem under control, here are 4 suggestions:
1. The rings are designed so that the minimum short circuit current is always safely above the tripping current of the transformer. In principle, the disadvantage here is that the primary technology of the secondary technology must adapt. Either the nets must be kept correspondingly small or short or very large, for normal operation actually not necessary cross sections are used.
2. The correspondingly large half-ring is divided into two protection sections. An additional station, including circuit breakers and protection technology, must be installed on the newly created half-ring. Again, this is not a purely secondary solution and the primary technology needs to be touched.
3. An additional circuit breaker failure protection is used. However, this does not help with a protection relay outage of the main protection.
4. Maximum adaptation of the protection technology? What ideas do you have?
If now arises the question of whether this cost is justified for such a rare case of a protection or line breaker, you have to worry about the consequences of a system failure.
A two-phase fault, several kilometers away from the substation, is often only a high operating current for the protection of the transformer. Now, if the protection or the circuit breaker fails, a first extension to a three-phase short circuit is likely. As a result, the starting value of the back up level in the feed can be reached under certain circumstances and everything would be fine.
And what if not? In the case of old three-phase lead sheath cables, the short circuit burns back within a very short time until the tripping level is reached. But what about today's XLPE cables?
If a short circuit then reaches the switchgear of the network station, a fire is the most probable consequence. The switchgear of the network stations are completely overloaded and theoretically over minutes. The entire cable system from the substation to the fault location is thermally irreparably damaged. The hazards can have dramatic effects. The danger of fire, if there are network stations in this area in buildings, can not be imagined. Therefore, a responsible approach to this topic is essential.
Many thanks to Sören, for the interesting contribution. From my point of view, a fully redundant ring protection with communicative feedback signal to a breaker failure protection function of the protective relay in the transformer feed-in is the solution. This would be the third variant paired with fully-redundant ring protection.