https://orcid.org/0009-0003-3527-3009
Figures
Abstract
DC circuit breakers (DCCBs) with high breaking capacity and low cost are necessary for quick fault clearance in DC networks. The assembly DC circuit breakers (ADCCBs) have a main breaking section (MBS) and a sub-breaking sections (SBS) for each line, which greatly reduce the cost. But in conventional operation, it bears high voltage for a long time when there is a main switch grounding process in any line fault action. To address this problem, a multiport assembly circuit breaker based on current injection (CI-MPACB) is proposed, which is able to generate a resonant current with increasing amplitude by controlling the duty cycle of Integrated Gate-Commutated Thyristors (IGCTs). Then the resonant current is injected into the SBS to generate current zero crossing and arc extinction. A complex frequency domain circuit analysis is performed on the MBS to describe the action logic as well as the commutation characteristics. In addition, the parameters of each component of the MBS are subject to multiple constraints and reasonable design to ensure the fault current could be cut off quickly and reliably. The cost of existing design is greatly reduced due to the design idea of resonant current injection device parameter selection. Finally, a PSCAD/EMTDC simulation confirms the opening viability of CI-MPACB and the accuracy of the parameter design. The test results show that the designed CI-MPACB can cut off DC fault lines.
Citation: Cui P, Li G, Zhang Q, He Q, Chen Z, Yang W (2024) Improved assembly DC circuit breaker based on resonant current injection. PLoS ONE 19(7): e0304979. https://doi.org/10.1371/journal.pone.0304979
Editor: Arvind R. Singh, HJNU: Hanjiang Normal University, CHINA
Received: October 9, 2023; Accepted: May 20, 2024; Published: July 17, 2024
Copyright: © 2024 Cui et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript.
Funding: The study was supported by “State Grid Corporation Headquarters Science and Technology Project (5500-20220110A-1-1-ZN)” The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
1 Introduction
An efficient method for facilitating seamless access, dependable transmission, appropriate allocation, and flexible consumption of extensive renewable energy is the utilization of DC grids with voltage source converter (VSC) [1]. It is one of the most important technologies to sustain the construction of a new power system with renewable energy sources in China and to achieve the ambitious goal of "carbon neutrality and emission peak " [2–4]. The DC grids with VSCs are "low damping" systems in comparison to the AC systems. Both the fault current and range might expand quickly at the same time [5]. To assure a continuous and reliable operation, it is critical to provide fast fault current blocking of the inverters in DC network after a short circuit fault in line.
DC circuit breakers (DCCBs) are one of the most reliable, rapid, low loss and low cost solutions to achieve fault isolation in the DC grid [6–8]. As there is no natural current in DC system, DCCB isolates fast transient and high short-circuit current from HVDC converter, making DC fault interruption more difficult [9, 10]. Multiple topologies and concepts have thus far been put forth to safeguard multi-terminal DC grids [11–13]. There are three main categories of DCCBs: mechanical DCCBs (MDCCBs), solid-state DCCBs (SSDCCBs) [14–16] and hybrid DCCBs (HDCCBs) [17–21]. Mechanical DCCBs have the advantage of lower losses and cost than SSDCCBs. However, spring actuators have long operating time. SSDCCBs have a high cost due to its large use of power electronic (PE) devices. HDCCBs provide a trade-off between the fast fault interruption and low conduction losses. But a typical HDCCB contains hundreds of PE switches. A mesh HVDC grid consists of numerous converters and extensive transmission lines. As a result, there are much more HDCCBs needed, which raises the cost.
The assembly DC circuit breakers (ADCCBs) consist of a main breaking section (MBS) on each DC bus in the corresponding converter station and the sub-breaking section (SBS) of each line connected to the station. It can cut all types of faulty lines, reduce the number of MBS and the manufacturing costs [22–24]. However, this typical ADCCB has the following problems:
- In normal operation, the MBS of the ADCCB and the auxiliary discharge switches on each line stay in the off state. At this time, each part MBS has to bear the voltage of the system to the ground. The internal PE devices can withstand higher voltage for a long time, which is harmful to the reliable of PE devices.
- After any line fault, the MB is grounded for a long time during the action of the ADCCB.
To solve the above problems, the VSC assisted resonant current assembly DCCB (VARC-DCCB) has been proposed. Due to the shared transfer branch and energy-consuming branch between adjacent lines, the construction cost of the breakers can be significantly reduced. In reference [25], an ultra-high speed actuator and resonant circuits are used. It combines the benefits of MDCCB and HDCCBs for shorter run times, lower conduction losses and better cost effectiveness. A new multiport DCCB (MP-DCCB) based on current injection is proposed in [26]. Its fault isolation time is reduced by detouring the current limiting inductor during dissipation of energy. Accordingly, a further diode branch is built to link the transfer branch to the bus [27]. This branch is capable of eliminating bus faults or fast mechanical switching faults. The bus fault is eliminated by a bypassing branch and thyristor with diode bridge circuit loop design [28].
IGBTs or injection enhanced gate transistors (IEGTs) are frequently used to break current for high-capacity interruptions because of their significant current-breaking capacity [29, 30]. Substantial IGBT series are necessary in the above MP-DCCB transfer branch circuit as switching elements. However, the massive IGBT series give rise to a series voltage unbalance. When the fault current is larger than 15 kA, a number of IGBTs/IEGTs must be linked in parallel in order to achieve a dependable interruption with high cost. Therefore, IGCTs with high surge current withstanding capacity and low cost are seen in rise in DCCBs as an economic alternative.
Due to the existing problems of the current developed CBs, a new multiport assembly DC circuit breaker based on current injection (CI-MPACB) is proposed. Firstly, the expensive HDCCB is widely used in the project at present. It adopts the design of one common MBS and several SBSs to reduce the cost. Secondly, the traditional ADCCBs proposed at present have the problem of the MBS needs to withstand high voltage for a long time. A novel auxiliary resonant current injection method is used. This topology solves the above problems of MBS. In addition, this paper also analyzes the topology principle of MBS and the working process of MBS and CI-MPACB in detail. Thirdly, a large number of series-parallel IGBTs are usually used as PE devices in CBs at present. This also causes high costs. The PE devices which control the current direction of the current injection branch are optimized. We propose to use fewer IGCTs combinations in place of large number of IGBT combinations. Finally, the compactness and low cost of CI-MPACB are realized to meet the requirement of fast and reliable disconnection of circuit breakers. By using multiple constraints to select the device parameters involved in the current injection branch.
The contributions of this article are manifold.
- We propose a new type ADCCB, which is based on the design idea of common MBS. It solves the problem of high cost due to the large number of CBs which are caused by multiple converters and multiple transmission lines in reticulated HVDC system.
- The MBS of CI-MPACB uses resonant current injection to make the line current turn off at zero crossing. It solves two problems in traditional ADCCBs: 1. PE devices bear high voltage for a long time during normal operation, 2. MB is grounded for a long time during fault removal operation. We also perform multi-constraints and cost optimization for the pre-charge capacitor and its voltage, resonant capacitor and resonant inductor in current injection branch. It causes compactness and low cost of CI-MPACB.
- We use IGCTs with high surge current resistance and low cost as the conduction components of the current injection branch. The principle of this new topology is examined in detail, and the rationality of the design is finally verified according to relevant simulation. The problem of high cost caused by a large number of series-parallel IGBT combinations has been solved. And the cost comparison between the PE devices has made the effect of this optimization scheme more intuitive.
The organization of this paper is represented as following. The Topology of the DC circuit breaker section presents the topology and operating principle of CI-MPACB. The Electrical parameter design section presents the optimized design of electrical parameters of the CI-MPACB. The Scheme Comparison and Cost Analysis section compares the technicality and cost of CBs. Based on simulation results, The CASE STUDY OF CI-MPACB IN THE VSC-HVDC section verifies the effectiveness and applicability of the circuit breaker. At last, conclusions are given in The Scheme Comparison and Cost Analysis section.
2 Topology of the DC circuit breaker
2.1 Structure of the RM-DC circuit breakers
In order to realize optional fault isolation, traditional multiport DCCB shall be installed at both ends of each lines connected to the DC bus. This will lead to high construction costs. As shown in Fig 1, the CI-MPACB can take the place of multiport DCCB, which greatly reduce the construction cost of DCCB.
The proposed CI-MPACB topology is shown in Fig 2, which consists of a MBS and SBSs corresponding to the number of lines. The MBS consists of a current injection branch and a energy dissipation branch. The current injection branch is comprised of an inductor LH, a high voltage capacitor CH and a square wave excitation source. The combination of pre-charged capacitor CP and IGCTs is designed as the wave excitation source. The SBS consists of a main breaker (MB) and an auxiliary breaker (AB).
In Fig 2, Idc is the total DC current. IMB is the current of the MB. IC is the current of the current injection branch. IMOV is the current flowing through the energy dissipation branch. UE is the external voltage of the square wave excitation source, with a initial value equals to the pre-charge voltage U0.
2.2 Principle of CI-MPACB fault-division control
As shown in Fig 3, it is the current path under normal operation of the line. When a fault occurs, the CI-MPACB will go through following stages during the breaking process. The opening sequence control logic in case of linei fault is given by:
Stage1: t ≤ tfault: Before the fault occurs, MBs of n lines are in the closing state, and ABs is in the opening state.
Stage2: tfault≤ t≤tCI: As shown in Fig 4(A), the fault occurs at the time instant tfault. After receiving the fault removal command of linei, the control device immediately give offs command to the ABi closing and MBi opening of linei. The MBi should be opened and the ABi should be closed immediately. After detecting that the ABi is closed and MBi is opened in place within 2ms, start the electronic trigger switch of the MBS. The increasing current is generated in current injection branch. The process of Stage2 is shown in Fig 4(B).
(a) Line fault occurs: t = tfault, (b) Current injection branch operates: tfault≤ t≤tCI, (c) Energy absorption branch operates: tCI≤ t≤ tEA, (d) Fault line is disconnected: t = toff.
Stage3: tCI≤ t≤ tEA: When the crosses zero and the arc is extinguished, the Idc transfers to the MBS. When the voltage at the resonant capacitor end exceeds the action voltage of the MOV, the MOV consumes energy and absorbs the line breaking energy. Detect current of
to 0, lock the main electronic trigger switch. The process of Stage3 is shown in the Fig 4(C).
Stage4: tEA≤ t≤ toff: ABi should be broken. The fault line is disconnected at the time instant toff.
The closing signal of the AB is an optical signal, and the communication protocol is consistent with the MB. Closing response time less than 50us
where tCI is the time when the current injection branch starts. tEA is the time when the energy absorbing branch starts.
Both MB and AB use high-performance epoxy resin to seal the pole, which has good vacuum sealing, insulation performance, resistance to high and low temperature impacts, and mechanical impact performance.
2.3 Operating principle of MBS
The waveforms related to the MBS breaking process are shown in Fig 5.
The following expound shows the full operation sequence:
- 0-t0: In advance to the operation of the CB, the controllable electronics are blocked and the capacitor CP is pre-charged by an external DC power supply.
- t0-t1: The fault occurs at time t0. Accordingly, the line current starts to increase, then the fault current limiting reactor (LDC) reduces the increasing rate of the line current. A trip signal is sent to the CI-MPACB at instant t1.
- t1-t2: As the CI-MPACB receives the trip signal at t1, the MB starts to separate the contacts, which reach a sufficient gap distance at t2.
- t2-t3: At t2, the current injection branch is activated. A zero-crossing of arc current is produced by the generation of the resonant current, which steadily grows in amplitude every half cycle.
- t3-t4: IMB crosses zero under the superposition of IC and the arc is extinguished at t3. Idc charges the resonant capacitor CH until MOV action voltage is reached.
- t4-t5: The MOV consumes energy. Finally, complete the entire break process.
Fig 6 shows the topology model when T2 and T1 are switched. As shown in Fig 6(A), T2 is triggered in the first cycle. Then the trigger signal is removed. Cp continuously discharges externally and current injection branch generates IC flowing clockwise. And the current gradually increase while charge CH. also increases. When the IC reaches the clockwise peak, the change rate of the IC is zero, which means that the
at this time drop across zero. According to Kirchhoff’s voltage law,
.
In Fig 6(B), the amplitude of the IC decreases in the clockwise direction due to the continuous current action of inductor LH. It still charges CH. When IC drops to zero, the change rate of IC is the largest, and the gets its maximum. At this time,
reaches the first peak value. And the IC changes direction across zero.
In Fig 6(C), when T1 is triggered, the IC will increase in the counterclockwise direction after crossing zero. When the IC reaches the counterclockwise peak, the polarity of the will change.
In Fig 6(D), the IC drops from the counterclockwise peak until the reaches its peak.
3 Electrical parameter design
3.1 Circuit analysis of the MBS
When the MBS of CI-MPACB starts to operate at t2, the initial voltage of pre-charged capacitor Cp is U0 and UCH is zero when T2 is firstly opened in the current injection branch. In the equivalent circuit, CH is serial connected with Cp, accordingly the equivalent capacitance of the resonant circuit can be calculated by (1). The total equivalent capacitance is approximately equal to CH because of Cp≫CH. The resonance angular frequency (ω) is obtained from (2), and the natural frequency of the commutation branch is fLC (3).
The following is a detailed description and formula expression of the t2-t3 process mentioned above. In the first stage of the first resonant period (T2 turn on, T1 turn off), the resonance of the first half-cycle current () is a second-order oscillatory current discharge process from the excitation source U0 of the Cp to CH and LH. In the second stage (T2 turn off, T1 turn on), there is no pre-charge capacitor Cp in the equivalent circuit, so the resonant current
is generated under the action of CH and LH. In the first stage of the m-th resonant period (T2 turn on, T1 turn off), In the equivalent circuit, CH and Cp are linked in series. At this point when the T2 conduction, the
can be regarded as the second-order resonant current of the excitation source UE charging and discharging LH formed by CH and Cp in series.
In the second stage (T2 turn off, T1 turn on), the can be regarded as the two-order resonant current generated by CH and LH. When the fault current (If) is interrupted at tI, then zero-crossing is generated at tI(4).
The resonant current IC determines the commutation capacity of CI-MPACB. The IC of CI-MPACB in this design is different from that of traditional mechanical DCCB. Its essence is a resonant current with increasing amplitude. The typical waveform is shown in Fig 7. To describe its commutation characteristics accurately, the process of resonant current IC generation will be used to clarify the mathematical model of MBS commutation and serve as the basis for parameter design.
Fig 8(A) shows the topology of the complex frequency domain equivalent circuit when T2 is turned on for the first time Fig 8(A). The formula of this process in the complex frequency domain is listed by using the loop current method:
(5)
After I(s) is resolved, the expression of the release current of the resonant circuit is obtained after inverse Laplace transform:
(6)
(7)
(8)
where
Fig 8(B) is the complex frequency domain equivalent circuit topology diagram when T1 is turned on for the m-th (m≥1) time. The formula of this process in the complex frequency domain is listed by using the loop current method:
(9)
After I(s) is found, the expression of the release current of the resonant circuit is obtained after the pull type inverse transformation:
(10)
(11)
where
Fig 8(C) shows the topology of the complex frequency domain equivalent circuit when T2 is turned on for the n-th (n≥2) time. The formula of this process in the complex frequency domain is listed by using the loop current method:
(12)
After I(s) is sorted out, the expression of the release current of the resonant circuit is obtained after inverse Laplace transform:
(13)
(14)
(15)
3.2 Method of parameter design based on multiple constraints and cost optimization
In order to find the best combination of CH, Cp, U0, and LH for the current injection branch, and meet the requirements for fast and reliable breaking of the MBS, it is necessary to optimize the parameters of the device. So as to realize the compactness and low cost of MBS. Therefore, the following five constraints are given in combination with the commutation characteristics and interrupting process of MBS.
Constraint 1: To ensure reliable commutation of DC current, the peak value of resonant current shall be greater than the maximum value of DC current to be disconnected.
Conestraint 2: In order to realize fast disconnection, the corresponding cut-off time of t1~t3 should be less than 3ms.
Conestraint 3: In order to prevent MB from re breakdown, the growth rate of UCH should be less than the MB medium strength UMB_d recovery speed in t3~t4 stages.
Conestraint 4: The upper and lower limits of resonance frequency are generally 3 kHz to 10 kHz according to [31]. The current slope at current zero and the selection of CH and LH parameters are determined by the resonance frequency.
Conestraint 5: In order to reduce the voltage withstand requirement of components, Cp shall be ensured have no obvious overvoltage. It be obtained according to the voltage division principle of series capacitor.
(19)
where,
is the voltage Cp bears during transient breaking. UMOV is the highest voltage of MOV. k is the reliability coefficient.
For the determined and CH, Cp and other parameters have array values, corresponding to different costs, the minimum total cost is taken as the optimization goal.
To reduce the volume and cost of DCCB, its total capacitance Ws should be as small as possible:
(20)
3.3 Case study of parameter design
Taking DCCB with rated voltage Urate of 50 kV, rated current Irate of 2 kA and maximum breaking current Idc_max of 25 kA as an example. The action time of MB is 3 ms. According to Eq (16), IC can reach 25 kA within 1ms at most.
Irate corresponds to the slowest condition of CH charging. And the maximum value of CH is 40 uF (corresponding to Constraint 2).
Since the dielectric strength recovery rate of MB is about 1 kV/μs [34]. Considering Idc_max is the fastest charging condition of CH, the minimum value of CH is 25 μF according to (17) (corresponding to Constraint 3).
Since it is known that CH is not less than 25 μF. According to Eq (3) (fLC) and the Constraint 4, LH should be greater than 7.2 μH and less than 80.4 μH. The stray inductance of the existing 50 kV voltage level DCCB is about 46 μH. In order to reduce the difficulty of commutation, LH is taken as stray inductance.
The voltage ratio of MOV is about 1.6 times [32]. According to Eq (19), when CH, U0 and k are determined,Cp limit value can be determined.
Starting from the minimum value of CH as 25 μF, set k as 1.2 to obtain the optimal values of Cp and U0 when CH changes. According to the research above, a commutation capability of at least 25 kA is needed to guarantee a dependable interruption fault current. According to (20), higher capacitor pre-charged voltage will result in larger bulk and higher cost. Therefore, the parameters of the current injection branch are recommended as follow, LH = 46 μH, CH = 35 μF, Cp = 4000 μF, and U0 = 4 kV.
4 Scheme comparison and cost analysis
4.1 Technical comparison of circuit breakers
To implement the fault current interruption, the proposed topology makes use of a commutation branch made up of LC and IGCT components. Table 1 compares the performance of the DCCB with traditional interruption technologies.
4.2 Comparison of costs
Additionally, the cost comparisons between the planned and current 50 kV ratings results are provided in Table 2. The design and selection of MOV must take a lot of redundancy into account in order to ensure the availability of the fault energy dissipation. Because this decision is largely influenced by the system parameters, further optimization is challenging. Therefore, the primary factor is the price of PE devices.
1) Topology of Scheme 1
As for scheme 1, there are m load commutation switches (LCSs), m MBS in scheme 1. The IGBT (3000A4500V) module is used for the LCS and MB.
2) Topology of Scheme 2
Only one MBS is required in Scheme 2 because to the H-bridge structure. However, one DC line now has two LCSs linked to it instead of just one. In total, there are 2×m LCSs, one MBS in scheme 2.
3) Proposed topology
A total of 1 MBS and m SBS need to be configured. The total number of IGCTs required for MBS is 12. And in the SBS, m MBs and ABs are required.
Compared with Scheme 1, the proposed CI-MPACB saves ¥(375×m-480) thousand on the cost of PE. When m = 2, the cost saving is ¥2.7million. When m = 3, the cost saving is ¥6.45 million. The proposed CI-MPACB saves ¥(150×m-180) thousand on the cost of PE compared to Scheme 2. When m = 2, the cost saving is ¥1.2million. When m = 3, the cost saving is ¥2.7 million. According to the above analysis, it is obvious that under the current situation of DC power grid with multi outlet network structure, the proposed CI-MPACB can greatly reduce costs on the premise of ensuring reliability.
5 Case study of CI-MPACB in the VSC-HVDC
5.1 Operating waveform of CI-MPACB
As shown in Fig 10, the mathematical model is basically consistent with the PSCAD/EMTDC simulation results, which can verify the current transfer mathematical model and the best parameters in Fig 9. And according to Fig 10 (B) IC can reach 25 kA in 1ms.
(a) Comparison results of UCP, (b) Comparison results of IC. (c) Comparison results of UCH.
In Fig 11(A), fault occurs at t0 = 0.5 s; When t1 = 0.5120 s, the dynamic and static contacts of MB in the fault line are separated and arcing occurs and AB in the fault line starts to close; When t2 = 0.5140 s, AB is closed. Then the MBS of the CI-MPACB takes turns IGCT according to the inherent frequency of the commutating branch. When t3 = 0.5147 s, Idc cross zero and transfer to the commutation branch. When t4 = 0.5149 s, the voltage of the converter branch reaches the MOV reference voltage. Then the MOV is connected and forced to Idc drops. When t5 = 0.5265 s, Idc is disconnected by residual current switch when it is close to zero.
(a) waveforms of current, (b) waveforms of voltage.
As shown in Fig 11(B), residual voltage UMOV of 50 kV DCCB is 80 kV, which is sufficient to attenuate Idc. From the UCH waveform, it oscillates and rises under the charge of the square wave excitation source. This is because that IC can increase the cause of resonance.
As shown in Fig 12, the current superposition of T1 and T2 tubes is the current of the current injection branch. And the voltage amplitude of T1 and T2 tubes decreases with the increase of turn-on times.
(a) waveforms of current, (b) waveforms of voltage.
5.2 Simulation of CI-MPACB
In order to prove the feasibility of the CI-MPACB, a three terminal DC system is used (Fig 13).
As shown in Fig 13, a simulation model with three terminal VSC-HVDC system is built in PSCAD/EMTDC. The voltage source converter is mainly consisted of fully controlled circulating current bridge, DC side capacitor, AC measurement converter transformer or converter reactor, and AC filter. Three-phase and two-level architecture is adopted in the fully controlled converter bridge. And each bridge arm is composed of multiple IGBT or GTO and other turnoff devices. The capacitor in DC side supports the converter’s voltage and buffers the inrush current when the bridge arm is turned off and reduce harmonics in DC side. The function of the AC measurement converter transformer or converter reactor is to filtrate harmonics on AC side. Transmission lines are transmitted by overhead lines, and Bergeron model is used to describe transmission lines. The parameters of the system are shown in Table 3.
Assume t = 0.5s on the transmission line12, the fault occurs at a distance of x(x≤200) km from S1. As shown in Fig 14(A), the system remains steady state, and load current IMBline12 of SBS in CI-MPACB1flow is unchanged from 0s to 0.5s. After the fault occurs, IMBline12 starts to rise rapidly. When t = 0.501 s, the dynamic and static contacts of MBline12 are separated and arcing occurs. Then the ABline12 starts to close for 2ms. When the ABline12 closure is completed, the MBS of the CI-MPACB rotates IGCT according to the inherent frequency of the current injection branch. As the resonant current increases, the fault current is rapidly transferred to the resonant branch. When the Idcline12 charges capacitor CH to the triggering conduction voltage of the MOV, Idcline12 then transfer to the energy consuming branch immediately. Idcline12 gradually decreases to 0. Under three fault conditions, the rising rate and amplitude of Idcline12 is decreased with the increase of distance. When x = 25, the peak value of Idcline12 is 23.38 kA. When x = 50, the peak value of Idcline12 is 17.96 kA. When x = 100, the peak value of Idcline12 is 12.14 kA.
(a) Currents of line Icline12. (b) Currents of current injection branch IC. (c)Currents of energy absorption branch IMOCV. (d) Currents of SBS section IMBline12.
As shown in Fig 14(B), the time of current injection into SBS decreases as the fault distance reduces (i.e., the number of resonant cycles increases). When x = 25, the time when the value of IC reaches the maximum value is 0.51506 s. When x = 50, the time when the value of IC reaches the maximum value is 0.51480 s. When x = 100, the value of reaches IC the maximum value is 0.51453 s.
As shown in Fig 14(A) and 14(C), the time at which MOV is turned on and forces Idcline12 down to zero increases with the distance to the fault. The decreasing rate of IMOV decreases with the fault distance increasing. At x = 25, the value of IMOV reaches zero in 0.5228 s, at x = 50, the value of IMOV reaches zero in 0.5268 s, and at x = 100, the value of IMOV reaches zero in 0.5315 s.
As shown in Fig 14(D), the amplitude of IMB decreases with the increasing distance to the fault. At x = 25, the amplitude of IMB is 45.48 kA. At x = 50, the amplitude of IMB is 36.16 kA. And at x = 100, the amplitude of IMB is 24.84 kA.
As shown in Fig 15(A), the decreasing rate and value of the pre-charged capacitor voltage UCP increases with the fault distance increasing. At x = 25, the value of UCP decreases to 2.858 kV, at x = 50, the value of UCP decreases to 2.433 kV, and at x = 100, the value of UCP decreases to 2.029 kV.
(a) Voltage of: CP: UCP. (b) Voltage of: CH: UCH. (c) Voltage of: CH: UMOV.
As shown in Fig 15(B), the time to reach the peak of the resonant capacitor voltage UCH decreases with the fault distance increasing, and the peak value also decreases with the fault distance increasing. At x = 25, the peak value of UCH is 92.286 kV, at x = 50, the peak value of UCH is 90.189 kV, and at x = 100, the peak value of UCH is 85.032 kV.
As shown in Fig 15(C), the time to reach the peak of MOV voltage UMOV and the peak value decreases with the fault distance increasing. At x = 25, the peak value of UMOV is 84.239 kV, at x = 50, the peak value of UMOV is 83.762 kV, and at x = 100, the peak value of UMOV is 79.247 kV.
A comparison of the line currents Idcline12 and Idcline13 in the three cases is shown in Fig 16. When the line between S1 and S2 is faulty, the CI-MPACB can immediately open the faulty line, and the line between S1 and S3 operates normally.
(a) Current waveforms: Iline12-25km and Iline13, (b) Current waveforms: Iline12-50km and Iline13, (c) Current waveforms: Iline12-100km and Iline13.
The CI-MPACB was applied to the ± 160 kV Nan’ ao multi terminal flexible DC transmission system. And compared with the MCB designed in paper [36]. Adopting a circuit breaker scheme with three 50kV sub-modules connected in series. The parameters of each sub-modules are selected as the optimization value mentioned above. The fault breaking waveforms are shown in the Fig 17.
As shown in Fig 17, the MB of the MCB proposed in the paper [36] starts to open and arc at the moment of 15 ms, and the current flowing through the MB continues to rise. The breaking current of the mechanical switch rises to 10.3kA at 18.5 ms. The commutator branch of the MCB is switched on. The precharge capacitor discharges and injects oscillating current into the MB. At the time of 18.89 ms, the current of MB oscillates to zero. The breaking time is 3.89ms. The peak value of superimposed current is 22.7kA. Under the same short-circuit condition, the MB of CI-MPACB starts to open at 16.5 ms. The current iniection branch is switched on at 18.5ms. The precharge capacitor discharges and T1 and T2 turn on alternately. At 18.8ms, the current of MB oscillates to zero and completes the breaking. The breaking time is 2.8ms. The peak superposition current is 20.07kA. The breaking time of CI-ACB is smaller than that of MCB. And the current stress is smaller.
Therefore, the CI-MPACB has a fast breaking speed and the injection current with an increasing periodic amplitude can form multiple zero crossing points, which can adapt to various short-circuit currents. This circuit breaker has high reliability in practical system applications.
6 Conclusion
A novel CI-MPACB for VSC based DC system is proposed in this paper. Its operating principle is outlined in detailed. A three-terminal HVDC grid with CI-MPACB is simulated in PSCAD/EMTDC, the validity and the feasibility are proved according to simulation results, and following conclusions could be drawn.
- The controlled resonant commutation principle of the MBS of the CI-MPACB creates multiple arc extinction zeros for the rapid mechanical switch. At the same time, the voltage level of power electronic devices can reduce to 10% of the maximum transient breaking voltage, and the breaking current is close to zero.
- The CI-MPACB adopts the design idea of ADCCBs. One current injection branch and energy dissipation branch are shared by every lines connected to the same DC bus. This feature fits well for HVDC applications.
- The CI-MPACB not only reduces the number and cost of power electronics devices by designing the resonant circuit in the transfer branch, uses IGCT to replace IGBT as the conduction device of the resonant circuit. It eliminates the voltage instability caused by a large number of IGBTs in series and further reduces the cost.
There will be an increasing number of multiport HVDC networks in the future. These applications stand to gain greatly from the proposed CI-MPACB.
References
- 1. Tang G., Luo X., and Wei X.: ’Multi-terminal HVDC and DC-grid technology’, Proceedings of the CSEE, 2013, 33, (10), pp. 8–17.
- 2. Yan Fu, Xiaofeng Zhou, Guoan Dai, et al. Research on coordination strategy for an offshore wind power DC chopper device and protection[J]. Power System Protection and Control, 2021, 49(15): 178–186.
- 3. Gaopeng Guo, Qiang Song, Biao Zhao, et al. Seriesconnected-based offshore wind farms with full-bridge modular multilevel converter as grid- and generatorside converters[J]. IEEE Transactions on Industrial
- 4. Qiang Wei, Bin Wu, Dewei Xu, et al. A mediumfrequency transformer-based wind energy conversion system used for current-source converter-based offshore wind farm[J]. IEEE Transactions on Power Electronics, 2017, 32(1): 248–259.
- 5. Pei X., Tang G., Zhang S., et al.: ’Analysis on transient characteristics of short-circuit current for bipolar VSC-based DC grid’, Journal of Global Energy Interconnection, 2018, 1, (4), pp. 403–412
- 6. Huang X., Qi L. and Pan J.: ’A New Protection Scheme for MMC-Based MVdc Distribution Systems With Complete Converter Fault Current Handling Capability’, in IEEE Transactions on Industry Applications, 2019, 55, (5), pp. 4515–4523
- 7. Pei X., Tang G., Pang H., et al.: ’Study on optimized coordination strategies between line protection and DC breakers for VSC-based DC grid’, Proceedings of the CSEE, 2018, 38, (S1), pp. 11–18
- 8. Belda NA, Plet CA, Smeets RPP. Analysis of faults in multiterminal HVDC grid for definition of test requirements of HVDC circuit breakers. IEEE Trans Power Delivery 2017;33(1):403–11.
- 9. Kontos E., Pinto R. T., Rodrigues S., and Bauer P., “Impact of HVDC transmission system topology on multiterminal DC network faults,” IEEE Trans. Power Del., vol. 30, no. 2, pp. 844–852, Apr. 2015.
- 10. Hechong Chen, Jiaxin Yuan, Shunkai Xu, et al. Electrical parameter analysis of a fast energy storage DC current limiter topology[J]. Transactions of China Electrotechnical Society, 2021, 36(14): 3043–3054.
- 11. Tahata K. et al., “HVDC circuit breakers for HVDC grid applications,” in Proc. 11th IET Int. Conf. AC DC Power Trans., Birmingham, U.K., 2015, pp. 1–9.
- 12. Sano K. and Takasaki M., “A surgeless solid-state DC circuit breaker for voltage-source-converter-based HVDC systems,” IEEE Trans. Ind. Appl., vol. 50, no. 4, pp. 2690–2699, Jul.-Aug. 2014.
- 13. Khan U. A., Lee J., Amir F., and Lee B., “A novel model of HVDC hybrid-type superconducting circuit breaker and its performance analysis for limiting and breaking DC fault currents,” IEEE Trans. Appl. Super- cond., vol. 25, no. 6, pp. 1–9, Dec. 2015.
- 14. Rodrigues R., Du Y., Antoniazzi A., and Cairoli P., “A review of solid-state circuit breakers,” IEEE Trans. Power Electron., vol. 36, no. 1, pp. 364–377, Jan. 2020.
- 15. Meyer C., Schroder S., and De Doncker R. W., “Solid-state circuit breakers and current limiters for medium-voltage systems having distributed power systems,” IEEE Trans. Power Electron., vol. 19, no. 5, pp. 1333–1340, Sep. 2004.
- 16. He D., et al.: A novel high-voltage solid-state switch based on the SiC MOSFET series and its overcurrent protection. High Voltage. 1–9 (2022).
- 17. Zhang B, Ren L, Ding JG, Wang J, Liu Z, Geng Y, et al. A relationship between minimum arcing interrupting capability and opening velocity of vacuum inter- rupters in short-circuit current interruption. IEEE Trans Power Delivery 2018;33(6):2822–8.
- 18. Tsukima M, Takeuchi T, Koyama K, Yoshiyasu H. Development of a high-speed electromagnetic repulsion mechanism for high-voltage vacuum circuit breakers. Electr Eng Jpn 2008;163(1):34–40.
- 19. Li B., et al: Transient fault identification method for bipolar short-circuit fault on MMC-HVDC overhead lines based on hybrid HVDC breaker. High Volt.6(5),881–893(2021).
- 20. Yu J., et al.: Hybrid HVDC circuit breakers with an energy absorption branch of a parallel arrester structure. High Volt. 7 (1), 197–207 (2022). https://doi.org/10.1049/hve2.12138
- 21. Chen W., Zeng R., He J., Wu Y., Wei X., Fang T., et al. (2021), Development and prospect of direct-current circuit breaker in China. High Voltage, 6: 1–15. https://doi.org/10.1049/hve2.12077.
- 22. Gaoren LIU, Feng XU, Zheng XU, et al. Assembly HVDC breaker for HVDC grids with modular multilevel converters[J]. IEEE Transactions on Power Electronics, 2017, 32(2): 931–941.
- 23. Gaoren LIU, Feng XU, Zheng XU, et al. An assembled HVDC breaker for HVDC grids [J]. Power System Technology, 2016, 40(1): 70–77.
- 24. Liu G., Xu F., Xu Z., Zhang Z. and Tang G., "Assembly HVDC Breaker for HVDC Grids With Modular Multilevel Converters," in IEEE Transactions on Power Electronics, vol. 32, no. 2, pp. 931–941, Feb. 2017.
- 25. Ängquist L, Nee S, Modeer T, Baudoin A, Norrga S, Belda NA. Design and test of VSC assisted resonant current (VARC) DC circuit breaker.In. the 15th IET international
- 26. Zhang S., Zou G., Wei X. and Zhang C.: ’Multiport Hybrid DC Circuit Breaker With Reduced Fault Isolation Time and Soft Reclosing Capability’, in IEEE Transactions on Industrial Electronics, 2022, 69, (4), pp. 3776–3786
- 27. Zhang S., Zou G., Wei X. and Sun C.: ’ Diode-bridge multiport hybrid DC circuit breaker for multiterminal DC grids’, IEEE Transactions on Industrial Electronics, 2021, 68, (1), pp. 270–281
- 28. He J., Luo Y., Li M., et al.: ’A high-performance and economical multiport hybrid direct current circuit breaker’, IEEE Trans. Ind. Electron., 2020, 67, (10), pp. 8921–8930
- 29. Zhang X. et al., “A state-of-the-art 500-kV hybrid circuit breaker for a dc grid: The world’s largest capacity high-voltage dc circuit breaker,” IEEE Ind. Electron. Mag., vol. 14, no. 2, pp. 15–27, Jun. 2020.
- 30. Chen Z. et al., “Analysis and experiments for IGBT, IEGT, and IGCT in hybrid DC circuit breaker,” IEEE Trans. Ind. Electron., vol. 65, no. 4,pp. 2883–2892, Apr. 2018.
- 31. Lindmayer M., “Mathematical modeling of the high fiequency behavior of vacuum interrupters and comparison with measured transients in power systems,” in Proc. 17th Int. Symp. Discharges Elect. Insul. Vac., 1996, pp. 323–331.
- 32. Weijie Wen, Pengyu Li, Bin Li, et al. Operation strategy and parameter optimization of multiport mechanical DC circuit breaker[J]. Automation of Electrical Power Systems, 2021, 45(11): 86–94.
- 33. Hasegawa R, Kanaya K, Koyama Y, Matsumoto T. Principle Experiment of Current Commutated Hybrid DCCB for HVDC Transmission Systems. 2018 International Power Electronics Conference (IPEC-Niigata 2018 -ECCE Asia), Niigata, Japan, 2018.
- 34. Wen W, et al. Analysis and Experiment of a Micro-Loss multiport Hybrid DCCB for MVDC Distribution System. IEEE Trans Power Electron 2019;34(8):7933–41.
- 35.
[Online]. Available: https://octopart.com/
- 36. Ming CHEN, Xiaolin LI, Shukai XU,et al. Engineering Application Research for Mechanical HVDC Circuit Breaker[J].Journal of Global Energy Interconnection,2018,1(4):423–429 (in Chinese).