Abstract-Fast and reliable DC fault detection is one of the main challenges for modular multilevel converter (MMC) based DC grid with DC circuit breakers (DCCBs). This paper extracts the high frequency components of transient voltages by wavelet transform and proposes a fault identification method based on the difference of transient voltages to identify the faulted lines for DC grids using overhead lines. Meanwhile, a faulted pole discrimination method based on the difference between the change of positive and negative pole voltages is presented. A line protection scheme including detection activation, fault identification, faulted pole discrimination and post-fault re-closing is designed. Using only the local measurements, the scheme can realize the protection of the whole line without communication and has the capability of fault resistance endurance and anti-disturbance. The proposed method is tested with a four-terminal MMC based DC grid in PSCAD/EMTDC. The selection methods of threshold values are presented and the impact of DCCB operation on the reliability of DC fault protection is analysed. Simulation results verify the fast detection and reliability of the designed DC line protection scheme. Index Terms-DC fault protection, fault identification, transient voltage, hybrid DC breaker, modular multilevel converter (MMC), post-fault re-closing.This paper is a post-print of a paper submitted to and accepted for publication in IEEE Transaction on Power Delivery and is subject to Institution of Electrical and Electronic Engineering

Modular multilevel converter (MMC) based high voltage direct current transmission (HVDC) is an effective solution for large-scale renewable power integration over long-distance. In the overhead MMC-HVDC systems, the high voltage DC circuit breakers (DCCB) are implemented to interrupt the DC fault current. Subsequent to fault isolation, the DCCBs are required to automatically re-close to restore power transmission quickly. However, when the DCCBs are re-closed to permanent faults, they will be tripped again, resulting in a high requirement of interruption capacity for DCCBs and second overcurrent strikes on the HVDC systems. To overcome the drawbacks of the conventional auto-reclosing scheme, this paper proposes an adaptive reclosing scheme based on the active pulse injection from the converter associated with the coordination control of hybrid DCCBs. The single-end adaptive reclosing scheme as well as the two ends adaptive reclosing scheme dedicated to two-terminal HVDC systems and meshed DC grids are presented respectively. By applying this method, the location of faults can also be achieved in the case of permanent faults. In order to verify the effectiveness of the proposed adaptive reclosing schemes, extensive simulations have been conducted under PSCAD/EMTDC.

Abstract-To detect the DC faults for MMC based DC grids using overhead line transmission, many protection methods in phase-domain have been proposed. These existing protection methods suffer from incomplete function, weak theoretical basis and sensitivity to fault resistance and noise disturbance. To overcome these shortcomings, this paper proposes an improved DC fault protection algorithm using the modal-domain approach for the MMC based overhead DC grids, which decouples interaction between positive and negative poles and mitigates the strong frequency-dependency of the characteristic impedance in phase-domain. The DC fault equivalent circuits are established in modal-domain and the fault characteristics during the initial stage are analysed. Based on the modal-domain analysis, the line-mode reactor voltage which combines fault characteristics of negative and positive reactor voltages, is employed to identify the internal faults. The zero-mode reactor voltage which enlarges the differences between faulty and healthy poles, is employed to select the faulted pole. This method is robust to fault resistance and noise with high detection speed. In addition, it is not affected by power reversal, AC faults and DCCB operation, which are validated and evaluated by simulations in PSCAD/EMTDC.

Fast and reliable protection is a significant technical challenge in the modular multilevel converter (MMC)-based DC grids. The existing fault detection methods suffer from the difficulty in setting protective thresholds, incomplete function, insensitivity to high-resistance faults and vulnerable to noise. This study proposes an artificial neural network (ANN)-based method to enable DC bus protection and DC line protection for DC grids. The transient characteristics of DC voltages are analysed during DC faults. On the basis of the analysis, the discrete wavelet transform is used as an extractor of distinctive features at the input of the ANN. Both frequency-domain and time-domain components are selected as input vectors. A large number of offline data considering the impact of noise is employed to train the ANN. The outputs of the ANN are used to trigger the DC line and DC bus protections and select the faulted poles. The proposed method is tested in a four-terminal MMC-based DC grid under PSCAD/EMTDC. The simulation results verify the effectiveness of the proposed method in fault identification and the selection of the faulty pole. The intelligent algorithm-based protection scheme has good performance concerning selectivity, reliability, robustness to noise and fast action.

To protect the converters and minimize the power transmission interruption during DC line faults, it is necessary to detect the faults at ultra-high-speed for the MMC based HVDC grids. This paper reviews the state-of-the-art of DC fault protection methods of MMC HVDC grids, and summarizes the underlying principles of each method. On this basis, the DC fault characteristics analysis in terms of modal-domain, time-domain and frequency-domain are analyzed, which direct the protection design. Typical boundary protection and non-boundary protection schemes are reviewed. The advantages and disadvantages of existing fault protection methods are compared. A two-terminal MMC-HVDC prototype is developed to test the effectiveness of three fault protection methods. Comprehensive quantitative assessments of the protection methods discussed above are carried out in a four-terminal MMC HVDC grid. Finally, the future trends of the protection schemes are discussed and the findings are concluded.

For modular multilevel converter (MMC) based DC grids, current-limiting reactors (CLRs) are mainly employed to suppress the fault current and provide boundary effects to detect internal faults. Thus, most existing protection schemes are highly dependent on the larger CLRs to guarantee high selectivity. However, in existing MMC based HVDC projects, the size of CLRs is restrained by the cost, weight, and system stability under normal state. Thus, boundary protections may fail to detect high-resistance faults and pole-to-ground faults under weak boundary conditions. To overcome these shortcomings, this paper proposes a fast and selective DC fault detection scheme independent of boundary components. The propagation characteristics of line-mode backward traveling-waves (TW) are analyzed to identify external and internal faults. The polarities of zero-mode backward TWs are employed to select faulted poles. The directional overcurrent based pilot protection is adopted as a complementary criterion to detect remote faults. The proposed method can be applied in MMC-HVDC systems with small CLRs that cannot provide strong boundary conditions. Besides, the detection speed is fast (less than 1.5ms). Moreover, it is robust to fault resistance and immune to noise. Various simulation results in PSCAD/EMTDC verifies the effectiveness of the proposed scheme.

Primary single-end line protection strategies of MMC HVDC systems are difficult to make a tradeoff between fast detection speed and high reliability. To improve reliability, the pilot protection schemes based on communication and data exchange can be adopted. However, the communication-based schemes suffer from potential communication failure problems, such as data error, data loss and time synchronization error. To avoid blocking of protection devices during communication failures, a resilience-oriented differential pilot protection method is proposed in this paper. To address the problem of synchronization error, a startup element based on the multiresolution morphological gradient (MMG) of traveling wave is proposed. For the problems of data error and data loss in communication, the sampled data are preprocessed by the morphological filtering (MF). And the correlation of traveling waves is used to identify the internal and external faults; the ratio of the morphological gradient of pole voltages is adopted to discriminate the faulty poles. A four-terminal MMC based DC grid model is built in PSCAD/EMTDC interfaced with the optical fiberbased communication system built in MATLAB/Simulink. The simulation results show that the protection scheme can effectively identify the faults against serious communication problems of 1% bit error rate and 5% data loss.

In the hybrid modular multilevel converter (MMC) based high voltage direct current (HVDC) systems, the fault current can be actively suppressed by the converter itself, which endows a smaller requirement for current-limiting reactors (CLR) and a larger time margin for fault detection algorithms, comparing with the half-bridge MMC. But the robustness to fault resistance and noise disturbance of existing boundary protection schemes will be deteriorated with small CLRs. Moreover, the fast response of the fault current-limiting control will change the output DC voltage of hybrid MMC, which affects the fault characteristics and may cause mal-operation of existing protection algorithms. Thus, a single-end protection scheme considering the impacts of the active current-limiting control is proposed for the hybrid MMC based DC grids. The travelingwave characteristics under different fault stages are analyzed to evaluate the impacts of the fault current-limiting control. In addition, a coordination protection strategy versus different fault conditions is adopted to improve reliability. Various cases in PSCAD/EMTDC are simulated to verify that the proposed method is robust to fault resistance, fault distance, power reversal, AC faults, and immune to noise.