DC fault analysis for modular multilevel converter-based system

J. Mod. Power Syst. Clean Energy DOI 10.1007/s40565-015-0174-3 DC fault analysis for modular multilevel converter-based system Bin LI1, Jiawei HE1, Jie TIAN2, Yadong FENG2, Yunlong DONG2 Abstract DC fault protection is the key technique for the development of the DC distribution and transmission system. This paper analyzes the transient characteristics of DC faults in a modular multilevel converter (MMC) based DC system combining with the numerical method. Meanwhile, lots of simulation tests based on MATLAB/Simulink are carried out to verify the correctness of the theoretical analysis. Finally, the technological difficulties of and requirements for the protection and isolation are discussed to provide the theoretical foundation for the design of dc fault protection strategy. Keywords DC system, DC fault analysis, DC fault protection, Modular multilevel converter CrossCheck date: 18 October 2015 Received: 5 June 2015 / Accepted: 19 October 2015 Ó The Author(s) 2016. This article is published with open access at Springerlink.com & Jiawei HE Bin LI Jie TIAN Yadong FENG Yunlong DONG 1 Key Laboratory of Smart Grid of Ministry of Education, Tianjin University, Tianjin, China 2 NR Electric, Nanjing, China 1 Introduction With the extensive development of distributed generations like the wind power and photovoltaic power [1], as well as the ever increase of electric vehicles [2] and other DC loads, the DC system is drawing growing research interests due to advantages of low power loss, low investment, high reliability and so on [3, 4]. Supported by the evolution of the power-electronic technology [5], the flexible DC system is gaining popularity due to the advantages such as having an independent power control and being immune to commutation failure [6]. In engineering practice, the two-level voltage source converter (VSC) has been acknowledged as a viable device to integrate the distributed generations and DC loads. However, it has drawbacks such as high switching frequency, great switching power loss and poor power quality [7]. In this context, the idea of the MMC technique proposed in [8, 9] based on the modular design has a low switching frequency and better power quality [10]. Presently, several technical challenges to the development of DC system are confronting us, among which is the DC fault protection, including the fault detection and fault isolation. The conventional converters, including the VSC and MMC, are not able to isolate the DC fault by themselves. Even the sub-modules in the converter are all blocked, the freewheeling diodes still act as an uncontrolled rectifier. Nowadays, there are typically three DC fault isolating methods. 1) The most ideal isolating method is interrupting the fault circuit by DC circuit breakers [11], which, however, are still not available for engineering application. Because there is no DC circuit breaker that could meet the requirements of interrupting capacity and action speed. 123 Bin LI et al. 2) Due to the technical difficulty of the DC circuit breaker, an AC-circuit-breaker-based isolating method which could interrupt the DC fault circuit reliably was proposed in [12]. However, drawbacks of this method are obvious: slow response of the mechanical AC switch gear and large blackout area of the system. 3) Because of the reasons above, researchers are looking into the third method, i.e., eliminating the DC fault current by the converter. Reference [13] proposed a new MMC sub-module topology with DC fault current eliminating capability. After a DC fault happens, the fault current could be eliminated due to the reverse voltage from the capacitors in sub-modules. Reference [14] proposed the doublethyristor switch scheme which prevents the AC-side current contribution and allows the DC cable current freely decay to zero. However, the drawbacks like the thyristors withstanding high dv/dt in the normal operation should be taken into account. Reference [15] designed an isolating method which could reduce the level of dv/dt that thyristors have to withstand. There has been a lot of work focusing on the fault isolation, however, the work on fault detection with identification capability is still rarely done. Therefore, it is necessary for us to make sense of DC fault transient characteristics. Reference [16] has done a detailed analysis for the DC faults in the two-level VSC based DC system. Because the topology of a MMC is different from the two-level VSC, it is important to investigate the transient characteristics of DC faults in a MMC-based DC system. The works of this paper are as follows: Firstly, this paper analyzes the transient characteristics of a DC fault in the MMC-based DC system, which are different from the characteristics in a VSC-based one [16], providing the theoretical foundation for the design of the DC fault protection. Secondly, based on the fault characteristics, this paper discusses the technological difficulties of and requirements for the protection against DC faults in the DC system. In addition, the correctness of the theoretical analysis is verified by the simulation tests in MATLAB/ Simulink. SM SM SM SM SM SM T1 D1 T2 D2 C us SM SM SM SM SM SM A phase-unit Fig. 1 Structure of MMC-based DC system Each SM is mainly comprised of two IGBTs (T1, T2), two freewheeling diodes (D1, D2), as well as a cell capacitor (C). Compared with the conventional two-level VSC, the only difference during the normal operation is the modulation strategy. MMC adopts the step pulse modulation which has advantages such as low switching losses and high waveform quality in contrast with the pulse-width modulating (PWM) that the two-level VSC adopts [9]. Fault current isolating technique and protection strategy are vital to the stability and reliability of the power grid. For this reason, it is essential to investigate DC fault transient characteristics and its influence on the DC-side system, AC-side system and converter. There are three kinds of DC faults, i.e., DC disconnection fault, DC poleto-ground fault and DC pole-to-pole fault. Generally, the damage of a DC pole-to-pole fault is the severest [17]. So this paper analyzes the transient characteristics of a DC fault under the DC pole-to-pole fault condition. In the field application, all the IGBTs would be soon turned off after a DC short-circuit fault. Fig. 2 illustrates the fault current path before and after turning off the IGBTs. Consequently, the DC fault transient process could be divided into three stages, namely ‘SMs normal operation stage’, ‘Initial stage after blocking SMs’ and ‘Uncontrolled rectifier stage’. 2.1 SMs normal operation stage 2 Fault analysis for MMC-based DC system Figure 1 shows a symmetrical bipolar DC system connected with the AC system by MMC. A MMC consists of three parallel-connected phase-units. One phase-unit comprises two arms, each with identical sub-modules (SMs) series-connected, as shown in Fig. 1. Owing to a (...truncated)