Comparison of Performance of the PPS Against the "Handshaking"

The "handshaking" method has been introduced in [24] [72]. In this method, the opening operation occurs in a limited number of FDs. Then, the network restoration occurs with a waiting time after the opening of a disconnector. This time is due to the operation of ACCBs where a delay was assumed as 100 ms (in [24] [72] ). This means that the fault must be isolated within 100 ms after opening the first dc FD in the network. If the fault isolation is not ensured, the grid re-energisation might re-initiate the dc fault. This restoration failure event is likely to happen if an FD opens more than 100 ms later than another FD.

Fig. 5.12 shows the dc current and state of FDs with the handshaking method upon a P2P fault on link 12 at 20 km from relay 12. It should be noticed that grid restoration is not simulated. The "handshaking" method would require the opening of ’potential’ FDs that, in this fault case, are FDs 12, 21, 32 and 41 (in open state in Fig. 5.12 (b)). Then, ACCB 2 would re-close 100 ms after the opening of ’FD21’ which is at ≈ 0.85 s. However, such an operation would re-initiate ac infeed current to the dc side fault. As the current increases, the

5.5 Summary 95

FDs opening operation would fail. Hence, a time delay larger than 100 ms, for re-closing ACCB 2 in this case, would be necessary to ensure dc fault isolation.

-2 0 2 4 6 8 10 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 I12p I21p I32p I41p Time [s] D C Cu rren t [pu] 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 FD12 FD21 FD23 FD32 FD24 FD42 FD14 FD41 Time [s] (a) (b) -0.2 0 0.2 0.4 0.6 0.8 0.8 0.9 1 1.1

Fig. 5.12 (a) DC link current and (b) FD opening state with the "handshaking" based strategy without ACCB reclosing.

In the PPS approach, the re-closing of ACCB 2 would occur only after the opening of all FDs associated to the busbar station 2, and after the re-closing of one of these disconnectors. For this reason, the grid restoration would not compromise the fault isolation process.

It should be highlighted that while the "handshaking" method was not conceived to deal with the failure of isolation devices, the PPS includes back-up protection (see Section 5.4.1). This capability ensures a higher reliability of the PPS in comparison to the "handshaking" method.

5.5

Summary

When ACCBs are used to clear dc side faults, P2P and P2Gnd faults lead to a temporary outage of the whole dc network. A novel protection strategy has been proposed to reduce the grid outage time through the progressive restoration of the dc grid.

The proposed strategy is able to detect a dc fault, discriminate the faulty link and restore non-faulty links in a progressive manner. Each FD is ordered to reclose in a step-by-step basis. The opening and reclosing operation of ACCBs and FDs is controlled in a busbar unit while link communication is not required. Hence, such unit could be easily adapted to different dc grid configurations, making the PPS a robust and flexible strategy.

Benefits of the PPS include fault clearance with economic ACCBs, avoidance of dc link communication channels, a limited time of power outage in the dc network and a guaranteed back-up protection. Grid restoration is faster in comparison to the "handshaking" method as it does not rely on the faulty link isolation time. The PPS has the ability to restore non-faulty dc links even if the faulty link is not completely isolated from the network. This reduces grid outage time.

Simulation results have illustrated the application of the PPS for two types of faults. For a P2Gnd fault, the dc network can be restored within 300 ms. In the case of a P2P fault, the grid outage ranges from hundreds of milliseconds to seconds as this depends on the dc fault current decay behaviour.

Chapter 6

Proposed Protection Strategy for MTDC

Grids equipped with FB Converters

6.1

Introduction

The design of MMC submodules is essentially restricted to HB or FB configurations [103]. HB submodules employ a design with 2 IGBTs, which incur of lower power losses in comparison to FB submodules (4 IGBTs). As a result, MMCs built entirely with HB submodules are more attractive in terms of investment and operational costs. However, the main drawback of HB based converters is the fault current that flows uncontrollably through the free-wheel diodes following a dc fault. Therefore, grids equipped with HB converters must rely on ACCBs or DCCBs to achieve dc fault current clearance.

Converters with FB submodules are able to quickly interrupt the converter fault current contributions [19, 73]. In this case, ACCBs or DCCBs are not required as fault clearance is ensured by the converters. Other advantages of the FB converter include dc voltage controllability and flexibility; i.e. possibility of a smooth voltage ramp up, voltage level reduction to mitigate atmospheric conditions, and voltage polarity reversal for fast de- ionisation of the arc [104].

To the knowledge of the author, only one protection strategy for MTDC grids using FB converters has been reported in the open literature [29]. In spite of its potential, dc protection with FB converters constitutes an under-researched topic due to higher investment

and operation cost (in comparison to HB converter) and due to the expected DCCBs, devices that would selectively isolate a dc faulty link. However, FB converters might have advantages that should be carefully analysed. To bridge this research gap, this chapter proposes a protection strategy for MTDC grids equipped with FB converters and link end FDs.