Impact of distributed generation in protection and fault management . 26

An increasing amount of distributed generation is changing the nature of the electricity generation from centralized power plants into smaller decentralized power generation units connected into medium and low voltage networks. Power flow of the distribution network is becoming more and more bidirectional instead of traditional convention, when power flows were unidirectional from the primary substation towards customer. [18] Con-nection of the DG with active network management (ANM) provides many possibilities e.g. avoiding or delaying network reinforcements, reducing network losses, and micro grid operation during major disturbance situations. However, distributed generation

causes also challenges in network protection, fault management and voltage rise in weak networks. [35]

3.6.1 Protection blinding

The principle of an overcurrent protection is based on the minimum fault current that causes the relay to trip. Protection zone of a certain CB must cover the fault current level even from the end of the feeder, where fault current has its minimum value. The fault current contribution of the DG unit can disturb the operation of the protection relay by delaying the operation or even block the operation completely. This phenomenon is called protection blinding and it is caused by the DG unit feeding the fault current parallel with the feeding substation. [18] The parallel fed fault current by the DG unit is presented in the Figure 13.

Figure 13. Fault currents fed by the primary substation and the DG unit.

Adapted from [35]

Even if the operation of the relay is not completely interrupted, the protection-blinding effect can delay the operation of the relay that can cause the thermal limits of conductors and components to exceed. The blinding problem can be overcome by configuring the relay with more sensitive tripping values, but this can lead to a sympathetic tripping by fault on the adjacent feeder, tripping caused by starting currents of the DG units or trip-ping by load current during maximum load conditions. The protection blinding can also be avoided by reinforcing the distribution network, but it usually makes connecting the problematic DG unit economically unfeasible. [18]

3.6.2 Adjacent feeder tripping

A distributed generation can cause unnecessary tripping of the feeder, when fault occurs in the adjacent feeder. In situation like these, DG unit located on the adjacent feeder feeds a fault current trough a substation to the faulted feeder and causes an unwanted operation of the feeder protection. This problem occurs when the fault and the DG unit are located near the substation, fault current capacity of the generation unit is large enough and feeder is not protected with a directional protection. [18] Protection selectiv-ity problem, or so-called sympathetic tripping, is presented in the Figure 14.

Figure 14. Tripping of the adjacent feeder. Adapted from [35]

Selectivity problems can usually be avoided with adequate protection relay settings. This can be overcome by changing the current settings of the relay located in the DG feeder, but in most cases, it cannot be done to ensure suitable protection in all of the fault situ-ations. [18] Instead of changing the current limits, the operation time of the relay can be set slower in the feeder, including the DG unit. If a delayed operation time of the protec-tion is used, precauprotec-tions must be taken that thermal limits of components are not ex-ceeded. In case these solutions are not possible, directional protection can be applied.

Directional protection is the most definite solution, but it requires replacement of the old relays with new ones, which causes expenses. [35]

3.6.3 Failed reclosing and loss of mains protection

Temporary faults are typically cleared by automatic reclosing, in which a fault arch is extinguished during non-supplied time. If the first reclosing fails, couple of more reclos-ings are made before the circuit breaker is opened permanently. If distributed generation units are connected to the faulty feeder, voltage is sustained, and the fault arch is not extinguished. Therefore, all the DG units must be disconnected before the reclosing se-quence is applied. Figure 15 illustrates the failed reclosing supplied by the tDG unit and the successful case of DG units disconnected before the reclose.

Figure 15. Failed reclosing due to DG unit [35]

To coordinate the protection scheme with the feeder and DG protection, distributed gen-eration units must be equipped with a loss of mains (LOM) protection. Thereby, the DG unit is disconnected instantly after the connection to the grid has been lost. In addition, after the fault has been cleared, the reconnection of the distributed generation must be done carefully to avoid harmful stress to the DG units. [36] The DG unit disconnected from the network must also be considered in a supply restoration. After the fault, distrib-uted generation is not supporting the feeder and loading conditions differ from the pre-fault situation. Thereby, supply restoration may not be possible even if constraint viola-tions are not exceeded the load flow calculation of the pre-fault state. Modelling the DG units to the distribution management system is essential due to the increasing penetra-tion of distributed energy resources. [18]