Enhancing the distribution grid resilience using cyber‐physical oriented islanding strategy
2019; Institution of Engineering and Technology; Volume: 14; Issue: 11 Linguagem: Inglês
10.1049/iet-gtd.2019.0184
ISSN1751-8695
AutoresAn Yu, Dong Liu, Bo Chen, Jianhui Wang,
Tópico(s)Power Systems Fault Detection
ResumoIET Generation, Transmission & DistributionVolume 14, Issue 11 p. 2026-2033 Research ArticleFree Access Enhancing the distribution grid resilience using cyber-physical oriented islanding strategy Yu An, School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai, People's Republic of ChinaSearch for more papers by this authorDong Liu, Corresponding Author dongliu@sjtu.edu.cn orcid.org/0000-0001-6990-2473 School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai, People's Republic of ChinaSearch for more papers by this authorBo Chen, Energy Systems Division, Argonne National Laboratory, Lemont, IL, USASearch for more papers by this authorJianhui Wang, Department of Electrical and Computer Engineering, Southern Methodist University, Dallas, TX, USASearch for more papers by this author Yu An, School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai, People's Republic of ChinaSearch for more papers by this authorDong Liu, Corresponding Author dongliu@sjtu.edu.cn orcid.org/0000-0001-6990-2473 School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai, People's Republic of ChinaSearch for more papers by this authorBo Chen, Energy Systems Division, Argonne National Laboratory, Lemont, IL, USASearch for more papers by this authorJianhui Wang, Department of Electrical and Computer Engineering, Southern Methodist University, Dallas, TX, USASearch for more papers by this author First published: 22 April 2020 https://doi.org/10.1049/iet-gtd.2019.0184AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onEmailFacebookTwitterLinked InRedditWechat Abstract The increasing penetration of distributed generations enables an innovative operation paradigm that allows islanded operation to enhance the resilience of the distribution grid. In this study, a cyber-physical oriented islanding strategy is proposed by coordinating centralised and distributed control to achieve seamless islanding transition and operational flexibility in emergency conditions. A cyber-physical control structure is developed to mitigate various disturbances (e.g. emergencies or fluctuations) according to different operation conditions. Specifically, the distributed fault isolation and seamless islanding transition are coordinated to mitigate the outage caused by unplanned islanding, while a secondary control is proposed to support primary control by reducing the power fluctuations during islanded operation. With a rapid response speed, the local cyber-physical devices are coordinated to accomplish islanding separation by selecting a feasible islanded area even under an unplanned islanding situation. A field test was conducted on a practical distribution network in China, and the results demonstrated the effectiveness and feasibility of the proposed islanding strategy. 1 Introduction With the increasing expansion of the electrical distribution grid, unexpected failures on power lines may cause widespread social disruptions and massive economic losses. These failures, resulting from extreme weather or artificial attacks, seldom occur but exert a high consequence, which arises a growing global need in enhancing the power grid resilience [1]. In the context of the power system, resilience is the ability of the power grid to withstand extraordinary and low-probability high-impact incidences, mitigate the interruption of power supply, and recover to the normal operation with a rapid response speed [2]. Many researchers have presented and discussed the framework of power grid resilience, including modelling challenges, limitations and impact assessment [3–8]. The existing resilience studies can be divided into two categories, infrastructure reinforcement or appropriate operational measures. The operational resilience measures include a black-start scheme with backup source and distribution reconfiguration, while infrastructure measures focus on upgrading components with stronger materials and building redundant transmission paths. With the increasing penetration of distributed generations (DGs), a promising operational resilience measure is to deploy the islanded operation of a portion of the distribution grid with power support from DGs. The islanded operation can improve the resilience of the distribution grid by providing an uninterruptible power supply for critical loads and reducing the outage area. The islanded operation was incipiently not allowed for several systematic conflictions in traditional protection settings [9, 10]. When a fault occurs, all the DGs are required to shut down immediately because the back feed short-circuits current provided by downstream DG can damp the sensitivity of traditional fault location methods. In addition, asynchronous reclosing of the DG will also cause severe transient oscillations to the power grid and trigger protective relays. Therefore, the microgrid is introduced to address the DG's integration. The microgrid is a small-scale power system consisting of interconnected DG and loads with a defined electrical boundary [11]. The microgrid typically represents a controllable entity with respect to the main grid. When the upstream grid is de-energised, the microgrid can be disconnected from the main grid by opening the point of common coupling (PCC), and operate as a stand-alone system with self-supply. The islanding methods of a microgrid are widely studied in the literature [12–20]. In [13], the operating margin of islanded microgrid with constraints is evaluated to ensure a sufficient power supply of critical loads. A detection method based on rate-of-change-of-frequency is proposed in [14] to recognise the islanding situations and avoid asynchronous reclosing of DG. The seamless transition strategy of inverter-based microgrid between grid-connected operation and islanded operation is described in [16, 17], where the inverter controller should switch the control mode once the microgrid is disconnect from the main grid. According to the studies in [18–20], the power sharing of multiple DGs in islanded operation can be coordinated via droop control. An optimisation approach combining particle swarm and the multi-agent system is introduced in [21] to improve the microgrid reliability and economy. Although the research studies in microgrid provide the technical basis for islanding implementation, there are still some limitations on microgrid to further enhance the distribution grid resilience. As mentioned above, the microgrid's boundaries are pre-defined because the loads and DGs have to be aggregated together under a fixed PCC [11]. Therefore, the microgrid should be applied to the region with concentrated DGs and loads. However, in most practical distribution grids, the DGs and loads are placed so dispersedly that the islanded operation cannot be implemented from a microgrid perspective. The pre-defined PCC will restrict the coverage of islanded microgrid when DGs are scattered. For instance, as shown in Fig. 1, diversified DGs are interconnected in a practical distribution grid in China. The DGs, including combined cooling and heating power (CCHP) system, battery energy storage system (BESS), photovoltaic (PV), and wind turbine (WT) are scattered and are connected to different 10 kV buses. Among these DGs, CCHP is an internal combustion engine (ICE) connected to the grid with a synchronous generator, and the other DGs are inverter-based. Another practical issue is that only the CCHP can perform as a voltage source of an islanded system, while the other DGs are brought in as constant power sources. The microgrid method is not applicable to this network condition, because it is hard to define appropriate microgrid boundaries, especially considering the continued expansion of the distribution grid. For an emergency case (e.g. unplanned islanding), as shown in Fig. 1, a part of the feeder may be isolated from the distribution grid due to a system fault. It is unrealistic to pre-define a microgrid for islanded operation because the fault location and load distribution cannot be predicted. Therefore, to improve the resilience of the distribution grid with scattered DGs, a more flexible islanding strategy should be investigated. Fig. 1Open in figure viewerPowerPoint Practical distribution system with the proposed cyber-physical control structure In this study, a flexible islanding transition strategy is proposed to improve the resilience of distribution grids by mitigating the outage in unplanned islanding condition. The contributions of this study can be summarized as (i) a cyber-physical control framework based on the coordination of local devices is established to accomplish seamless islanding transition under unplanned islanding situation; (ii) a master–slave-based power sharing method is proposed to reduce power fluctuations during islanded operation; (iii) a field test was implemented in a practical distribution network in China, which validated the effectiveness and feasibility of the proposed method. The rest of the paper is organised as follows: Section 2 presents the framework of the proposed cyber-physical control structure. Section 3 details the resilient islanding control procedure with fault isolation, islanding separation, islanding transition, islanded operation, and grid reconnection. Section 4 presents the field test results in a practical distribution grid to show the feasibility of the proposed strategy. Section 5 concludes the paper. 2 Cyber-physical control framework The proposed cyber-physical control structure, as well as a practical distribution grid, is shown in Fig. 1. The control system consists of applications and database servers in the control centre, coordinator, DG controller, and feeder terminal unit (FTU). The interdependent communication data flows and control signal flows in this cyber-physical control structure are illustrated in Table 1. Table 1. Content of control flow and data flow Flow Function and content Type s1 local relay control control s2 local DG control; anti-island protection control s3 operation point setting; control mode switch control s4 islanding separation control control s5 optimal control; topology update control d1 local measurements from FTU data d2 local information from DG controller data d3 shared information of fault current data With comprehensive real-time data acquired from local sensors, the control centre evaluates and regulates the overall operation of the substation by running algorithms in multiple applications (e.g. state estimation, load forecasting, and optimal generation dispatching) and transmitting the operational decisions to the field devices. The field devices, including FTUs and DG controllers, are local intelligent devices placed at each bus and each DG, respectively. The FTUs normally serve as local sensors providing updated measurements for the upper controllers (control centre and coordinator). The FTUs are also equipped with distributed feeder automation (FA) for automatic fault detection, fault location, and fault isolation. The DG controllers are developed to regulate DG outputs by receiving remote control signals from upper controllers to modify specific operational parameters, such as droop slopes of CCHP, power references, voltage set point etc. In normal operation, the field devices are under centralised control by receiving periodical control signals from the control centre, which, however, may fail to meet the response speed requirements in some emergencies (e.g. unplanned islanding). In unplanned islanding condition, the fault may occur at any section of the feeder so that the islanded area cannot be predicted. The feasibility of islanded operation should be identified rapidly. However, the algorithm in the control centre is time consuming as well as the field devices are unable to estimate the power deficit with limited local data. Therefore, a coordinator is developed in response to emergencies by taking over centralised control of the field devices on a feeder. The coordinator will estimate a feasible islanded area and implement the islanding transition. During the islanding transition, if necessary, the coordinator will also execute load shedding to minimise power deficit. 3 Proposed method With the cyber-physical control framework introduced above, a detailed islanding strategy, as well as the control sequence, will be presented in this section. As shown in Fig. 2, to improve the resilience of distribution grids by providing the uninterruptible power supply, the attempt of islanded operation should be prior to the load transfer procedure. Therefore, a time delay of anti-island protection is set to each DG to avoid DG disconnection during the islanding transition. The fault isolation, islanding separation, and islanding transition should be finished within this time delay. In case the islanding transition fails, all the DGs will be shut down, and the control centre will implement load transfer. In this section, the proposed islanding strategy is detailed, while the load transfer method is not discussed as it is not the focus of this study. Fig. 2Open in figure viewerPowerPoint Resilience control sequence 3.1 Fault isolation FA is widely used in the distribution network for its fast fault identification, fault location, and fault isolation [22]. As the first step in the proposed resilience control sequence, fault isolation will be implemented locally with a distributed FA system equipped in each FTU (Fig. 3). Owing to the delay of anti-island protection, downstream DGs will provide back feed short-circuit current. According to Kirchhoff's law, the direction of fault currents at terminals of the fault line is opposite, while those of the non-fault line are the same. Therefore, a differential-activated algorithm is used in distributed FA to locate the fault area by comparing the direction of the fault current [23]. The short-circuit current flowing on each circuit breaker is synchronously sampled by each FTU with the global positioning system, and transmitted to the neighbour FTU based on a peer-to-peer process. The microcomputer in distributed FA computes the sampled data from both sides and triggers the protection locally with an embedded fault location algorithm. If a fault is located, the FTUs will automatically open the circuit breakers at both sides of the fault line to isolate the fault area. In the proposed resilience method, fault isolation will activate the procedure of the islanding transition, which will be performed by the coordinator. Fig. 3Open in figure viewerPowerPoint Identification of a feasible islanded area 3.2 Islanding separation As shown in Fig. 3, after fault isolation, the coordinator will be activated to implement the islanding transition. However, considering the unplanned islanded situation, the capacity of DGs and total load consumption in the downstream area may not be perfectly matched. The islanded system will collapse if the CCHP fails to balance the power deficit. Besides, the power deficit in the islanded area should be as small as possible to achieve a seamless transition. Therefore, necessary islanding separation will be processed by the coordinator to minimise the power deficit. Islanding separation is a load shedding process during islanding transition. The coordinator will identify a feasible island area and execute load shedding by opening a specific circuit breaker based on the current power deficit. The approximate power deficit in the potential islanded area can be estimated based on the prefault exchange power at each circuit breaker (1)where denotes the prefault exchange power at upstream boundary i, and is the prefault exchange power at downstream boundary j. As shown in Fig. 3, there can be multiple potential islanded areas for upstream boundary i and downstream boundary j. Based on (1), the coordinator will compare the power deficits of all potential islanded areas and determine island boundaries i and j with minimal . The proposed algorithm is simple that can meet the requirement of response speed in islanding transition. After identification, the coordinator will send control signals to FTUs to open specific circuit breakers. 3.3 Islanding transition The seamless transition of the isolated area from grid-connected operation to islanded operation is achieved by the primary control of the CCHP, as shown in Fig. 4. The CCHP's prime mover, ICE, can provide heavy rotational inertia during islanded operation. The active power control and reactive control of CCHP are decoupled with the governor and automatic voltage regulator. The basic idea of the primary islanded control is to regulate the frequency and voltage by adjusting the active and reactive power output based on the concept of droop method, which can be expressed as (2)where and are the nominal rotating speed and the nominal voltage of the generator, and and are the active and reactive droop slopes, respectively. Fig. 4Open in figure viewerPowerPoint Primary control blocks in CCHP When the feeder of CCHP is connected to the main grid, the frequency and voltage are dominated and remain stable (, ). Then the output of the CCHP is subject to constant power control with and during grid-connected operation. When the CCHP is isolated from the main grid caused by the upstream fault, the droop control takes effect as the rotating speed and voltage of the CCHP occur deviation from its nominal value. The active power output of the CCHP will increase as the frequency drops, and vice versa. The power deficit in the islanded system can be compensated by the CCHP through droop control. After the coordinator approves the feasibility of islanding action, the CCHP switches its control mode to V/f control by gradually reducing the parameter values of droop slopes and until and . Therefore, the CCHP works as a slack bus and the islanded system is under steady frequency and voltage control with , . During islanding transition, other DGs (except CCHP) will maintain as constant power sources without any contribution to voltage transient. 3.4 Islanded operation During islanded operation, the CCHP is the only voltage source because the other DGs (PV, WT, BESS) have no droop characteristics. However, these constant power sources can still make a contribution to islanded operation with a secondary control. To smooth the output curve of CCHP by utilising these constant power sources, a secondary control scheme with a master–slave structure is proposed in this section, as shown in Fig. 5. Fig. 5Open in figure viewerPowerPoint Framework of the secondary control In secondary control, CCHP will work as a master, while other DGs will be slaves. The outputs of slaves will be re-dispatched by the coordinator according to the output of the master. Generally, the output of the master () will change with the voltage and frequency deviation caused by load fluctuations as well as intermittent DGs. In other words, by regulating the output of slaves, the total net load (the sum of load consumption and constant power sources) can be shaved to a stable value, which will be finally balanced by CCHP. In this way, the output of CCHP (same to total net load) can be smoothed around a benchmark value (). The benchmark value () for CCHP is the same value as the power reference of CCHP prior to fault isolations. Since during grid-connected operation, the CCHP is already provided a power reference by the control centre based on the calculation results of optimisation algorithms, which is also recorded and updated in the coordinator. When CCHP is switched into islanded operation, this power reference will continue to be the benchmark for the coordinator to perform secondary control. As shown in Fig. 5, the control error of secondary control is defined below (3)where the control error is the difference between current CCHP's output and the provided benchmark value . If the control error exceeds the threshold, the secondary control will be activated to re-dispatch the power references for slaves (PV, WT, BESS), otherwise, the power references for slaves remain unchanged (4)where is the dead-band coefficient for control threshold, which can be selected as 0.05–0.1, is the power reference of the slave i in the next control period, is the current power output of the slave i, is the compensation weight of the slave i, which can be determined in (5) and (6) (5) (6)where is the maximum available capacity of the slave i, n is the number of the slaves, is the current maximum power point tracking value for intermittent DGs (PV and WT), is the maximum rated discharging power of BESS, is the maximum rated charging power of BESS, E is the rated storage capacity of BESS, SOC is the current state of charge of BESS, T is the estimated time duration of islanded operation. It should be noted that , , and all represent positive values, which indicates the compensation weight is also positive. However, the power reference of BESS can be negative when it is under charging mode. According to (4), when the control error is negative and exceeds the threshold, the power reference of BESS may be decreased from a positive value to a negative value, which indicates the BESS will be switched to charging mode. The BESS can be switched between charging mode and discharging mode according to its power reference . 3.5 Grid reconnection There are two options for the grid reconnection of the islanded feeder. One is to connect to another feeder with available power capacity, the other is to reconnect back to the previous grid if the fault is cleared. For both options, the islanded system should be synchronous to the main grid in advance to avoid severe transient problems caused by unsynchronised reconnection. The synchronisation is approved when the voltage magnitude difference and the voltage phase angle difference at the reconnection point satisfy (7) (7)where and are the pre-defined values according to the grid code requirements. The phase synchronisation is accomplished by an additional synchronisation controller in the governor, as shown in Fig. 6, where , , and , , are the instantaneous values of the three-phase voltages from the main grid side and the islanded feeder side, respectively. Through Clarke's transformation (8)–(10) (8) (9) (10)where , and , are the voltages on the -axis from the islanded feeder side and the main grid side, respectively; and are the voltage phase angles of the islanded feeder and the main grid. As the voltage magnitude of the islanded feeder synchronous to the main grid side (), we can derive the difference of the voltage phase angle using the equation below (11)where is the voltage phase difference between the islanded system and the main grid. When the phase angle difference , . Then can be used as input to correct the control error in phase synchronisation control (as shown in Fig. 6). Under a feedback control of a proportional–integral controller, a new reference rotated speed of CCHP is corrected for CCHP's primary control. Fig. 6Open in figure viewerPowerPoint Control blocks of the voltage phase synchronisation For voltage magnitude synchronisation, the method is to reset a new voltage reference for CCHP. As shown in Fig. 4, the CCHP is under V/f control with a fixed voltage output . When is replaced by the voltage magnitude of the main grid , the voltage of CCHP, as well as the voltage of the feeder side, will synchronise to the main grid through primary control. After the successful reconnection of the islanded feeder, the CCHP's control mode was switched from V/f control to PQ control by resetting the parameter values of the droop slopes and . 4 Field test 4.1 Preparation The field test of the proposed resilience method was performed and evaluated in a practical distribution network in China. The demonstration system was developed with complete sets of cyber-physical devices, including control centre, coordinator, DG controllers, FTUs, and communication networks. The original topology of the distribution network is shown in Fig. 1, while the whole islanding procedure with topology changes was displayed through the monitoring interface, as shown in Fig. 7. The field test was implemented on the Feeder SP, where the basic parameter of DG generation and load consumption is listed in Table 2. The CCHP has a minimal active power value (50 kW) for stable operation, while other DGs can decrease their outputs to 0 kW. The maximum charging power and discharging power of a single BESS are −50 kW (negative value represents charging) and 50 kW, respectively. The rated storage capacity of a single BESS is 50 kWh. Before the field test, the output of CCHP was 250 kW, each BESS was under charging mode of 40 kW with SOC = 0.5, the total output of intermittent DG (PV and WT) is around 120 kW, the consumption of critical loads was about 280 kW, and the consumption of flexible loads was around 100 kW. The field test aimed to verify the system's resilience in the emergency condition (e.g. unplanned islanding), so we generated a fault at the head-end of the Feeder SP (as shown in Fig. 7) to observe if the system could cope with this unbalanced situation and transferred to islanded operation successfully. It should be noted that the fault did not actually occur, but was generated by injecting analogue fault current waveform into each FTU via a feeder fault producer, where the fault current waveforms were referred to the simulation result in DIgSILENT software. The fault current waveforms captured by the FTUs at both ends of the fault line were recorded in Fig. 8. Fig. 7Open in figure viewerPowerPoint Decoupling of the islanded area Table 2. Parameters of DGs and loads Name Current status (total) Power range (single) CCHP 250 kW 50–500 kW × 1 BESS −80 kW (SOC = 0.5) −50 to 50 kW × 2 PV 50 kW 0–50 kW × 4 WT 70 kW 0–100 kW × 1 critical load (CL) 280 kW — flexible load (FL) 100 kW — The amplitude of the upstream fault current had exceeded the recording limits, while the downstream one was recorded completely. The reason is that the fault current provided by the main grid was greater than the fault current provided by the DGs. As shown in Fig. 8, the directions of fault currents at the fault line were totally opposite, which assisted distributed FA to locate and isolate the fault area by opening circuit breakers (CB1, CB2). Fig. 8Open in figure viewerPowerPoint Sampled current waveforms at the fault line 4.2 Result of islanding transition As shown in Fig. 7, after FTUs tripped the circuit breakers and isolated the fault area, the islanding transition was initiated and processed by the coordinator. The coordinators estimated the feasible islanded area by calculating the current power deficit (about 90 kW) downstream of the fault area and executed the islanding separation to minimise the power deficit. The CB3 was opened by FTU with an indication from the coordinator. At the same time, the coordinator transmitted the control signal to the DG controller of CCHP to transfer CCHP into V/f control. After islanding separation, the area 2 (annotated with blue) with flexible loads and two PVs (about 30 kW) was abandoned, therefore, the final power deficit in area 1 (annotated with red) was around 20 kW. As a result, area 1 was secured with an uninterruptible power supply in islanded operation, while area 2 was inevitably abandoned for load reduction. The DGs in area 2 were shut down due to anti-island protection. The loads in area 2 were forced into outage and waited for load transfer procedure. In our demonstration, by closing the tie breaker, area 2 was finally connected to Feeder SH, which had available capacity. With successful islanding transition, most of the critical loads and renewable energies on Feeder SP were secured to ride through the fault. The voltage transient of CCHP during islanding transition is shown in Fig. 9. After fault isolation, the CCHP was disconnected from the main grid so that the voltage dropped severely at the beginning. This decline stopped when the coordinator implemented islanding separation by tripping CB3. During this period, the voltage of the islanded area was maintained by the heavy rotational inertia of CCHP as well as its droop characteristics. After CCHP was switched to V/f control, the islanded system's voltage was recovered back to the nominal value. During the islanding transition, the implementation of islanding separation was critical because the instant power deficit was minimised, which contributed to the voltage transient of CCHP. Fig. 9Open in figure viewerPowerPoint Voltage of CCHP during islanding transition 4.3 Result of islanding operation and reconnection The power sharing result under secondary control in islanded operation is shown in Fig. 10, where Fig. 10a presents the overall power sharing of the DGs and Fig. 10b shows the performance of the secondary control. Fig. 10Open in figure viewerPowerPoint Performance of the secondary control during islanded operation (a) Power outputs of DGs during islanded operation, (b) Control error in the secondary control During islanded operation, the CCHP regulated the voltage and frequency of the islanded system with its primary control, while the other constant power sources were under secondary control based on a master–slave framework. The aim of the proposed secondary control was to reduce the fluctuations of total net load (the sum of load consumption and constant power sources) by re-dispatching the power references of constant power sources. With shaved total net load, the output of CCHP was smoothed around a benchmark value (). In the field test, the benchmark value in secondary control was 250 kW, which was the same as the power reference of CCHP before islanding. The dead-band coefficient was selected as 0.05. Therefore, according to (4), the upper bound and lower bound of the control error in secondary control were ±12.5 kW. When the difference between CCHP's output and benchmark value exceeds 12.5 kW ( > 12.5 kW), the secondary control would be activated by the coordinator to re-dispatch the output of slave DGs (BESS, PV, and WT). The performance of the secondary control can be evaluated from the result in Fig. 10b, where the control error was maintained in the range of ±12.5 kW. When the control error was beyond the range of ±12.5 kW, the secondary control would change the total net load back to the benchmark by re-dispatching slave DGs according to (3)–(6). The result shown in Fig. 10 indicates that, compared to the load variation, the output curve of CCHP was smoothed because of the secondary control In grid reconnection, as shown in Fig. 11, there were two options: one was to get back to the Feeder SP if the fault was cleared and the upstream lines were under normal operation, the other option was to join another available feeder interconnected. In the field test, the islanded feeder was finally reconnected to the Feeder SH by closing the circuit breaker CB3. Without the restriction of PCC, the grid reconnection in the proposed strategy was more flexible. Fig. 11Open in figure viewerPowerPoint Grid reconnection of the islanded area The frequency performance of the islanded system is displayed in Fig. 12, where the islanded system was decoupled from the Feeder SP at 17:15 and reconnected back to the distribution grid at 17:53. As shown in Fig. 2, the frequency fluctuated before the islanded operation because of the high penetration of intermittent DGs on Feeder SP. The frequency also fluctuated after grid reconnection because almost all the loads and DGs on Feeder SP were transferred to Feeder SH, which was brought with an additional burden in frequency regulation. During islanded operation, the frequency fluctuations were more significant, because the inertia, as well as the capacity of the islanded system, was much smaller than that of the distribution grids. Some typical ranges of frequency limits for the Chinese power system [24–27] are listed in Table 3. With comparison, the frequency performance in the field test was far from excellent, but acceptable, especially when considering that the tested system was not a typical microgrid with pre-planned structure. The proposed method was more like an emergency measure that aimed to mitigate the outage in unplanned islanding condition. When the DGs and loads in the islanded area were scattered, it was difficult to optimise the system's frequency based on a single V/f source. Further improvement of frequency control should be investigated in future work. Fig. 12Open in figure viewerPowerPoint Frequency of CCHP during islanded operation Table 3. Typical frequency limits for China Type Level, kV Ranges, Hz distribution grid [24] ≥ 35 49.5–50.5 grid-connected microgrid [25] 6–10 48–50.5 islanded microgrid [26, 27] 6–10 47.5–52.5 5 Conclusion A cyber-physical oriented islanding strategy is proposed in this study to enhance the distribution system's resilience under unplanned islanding situations. The seamless islanding transition was implemented locally by the coordination of cyber-physical devices based on shared measurement data. During islanded operation, a master–slave-based power sharing method was proposed to mitigate the fluctuations. The field test was accomplished in a practical distribution grid in China, and the results demonstrated the effectiveness and feasibility of the proposed strategy. Compared to microgrid [11], some characteristics of the proposed strategy are summarised in Table 4. (i) The islanding region in the proposed strategy is more flexible regardless of PCC, while the islanded operation in a microgrid is limited in a pre-defined area. Besides, the reconnection point is more flexible in the proposed strategy, while the PCC is the only point for a microgrid in grid reconnection. Table 4. Comparison between microgrid method and proposed strategy Characteristic Microgrid Proposed strategy flexibility pre-defined microgrid area with fixed PCC flexible islanding transition with no fixed PCC stability more stable with a pre-planned structure depend on the practical network condition resilience secure its own microgrid region boost the resilience of the distribution grid (ii) The islanded operation of a microgrid is more stable due to its pre-planned and well-designed structure. However, the performance of frequency control in the proposed strategy cannot be guaranteed when the DGs are scattered, or there is a high penetration of intermittent DGs in the islanded area. (iii) The microgrid is mostly applied to a small region with concentrated DGs and loads, while the proposed islanding method is designed for common distribution grids. The proposed strategy is an emergency measure, which can boost the resilience of the distribution grid by mitigating the outage under an unplanned islanding situation. In conclusion, the proposed islanding strategy is more flexible and universal for the distribution grid with scattered DGs and loads, while the overall coordination framework, as well as distributed control algorithm, still requires further improvements and optimisations in the future. 6 Acknowledgments Yu An and Dong Liu acknowledge the support of the National Key Research and Development Program of China (Basic Research Class: 2017YFB0903000) and the National Natural Science Foundation of China (with issue number: 51677116). 7 References 1SouthWell, P.: ' Disaster recovery within a Cigre strategic framework: network resilience, trends and areas of future work', CIGRE Study Committee C1, 2014Google Scholar 2Chaudry, M., Ekins, P., Ramachandran, K., et al.: ' Building a resilient UK energy system'. UK Energy Research Centre, London, U.K., 2011Google Scholar 3Panteli, M., Mancarella, P.: ' Influence of extreme weather and climate change on the resilience of power systems: impacts and possible mitigation strategies', Electr. Power Syst. Res., 2015, 127, pp. 259– 270CrossrefWeb of Science®Google Scholar 4Panteli, M., Mancarella, P.: ' Modeling and evaluating the resilience of critical electrical power infrastructure to extreme weather events', IEEE Syst. J., 2015, 11, pp. 1– 10Web of Science®Google Scholar 5Panteli, M., Mancarella, P.: ' The grid: stronger, bigger, smarter?: presenting a conceptual framework of power system resilience', IEEE Power Energy Mag.., 2015, 13, pp. 58– 66CrossrefWeb of Science®Google Scholar 6Wang, Y., Chen, C., Wang, J., et al.: ' Research on resilience of power systems under natural disasters – a review', IEEE Trans. Power Syst., 2016, 31, pp. 1604– 1613CrossrefWeb of Science®Google Scholar 7Liu, Y., Singh, C.: ' A methodology for evaluation of hurricane impact on composite power system reliability', IEEE Trans. Power Syst., 2011, 26, pp. 145– 152CrossrefWeb of Science®Google Scholar 8Billinton, R., Singh, G.: ' Application of adverse and extreme adverse weather: modelling in transmission and distribution system reliability evaluation', IEE Proc. Gener. Transm. Distrib., 2006, 153, pp. 115– 120CrossrefGoogle Scholar 9' IEEE standard for interconnecting distributed resources with electric power systems', IEEE Std. 1547-2003, 2003Google Scholar 10' IEEE guide for design, operation, and integration of distributed resource island systems with electric power systems', IEEE Std. 1547.4, 2011Google Scholar 11Chowdhury, S., Chowdhury, S.P., Crossley, P.: ' Microgrids and active distribution networks' ( The Institution of Engineering and Technology, Stevenage, UK, 2009), pp. 3– 8Google Scholar 12Peiravi, A., Ildarabadi, R.: ' A fast algorithm for intentional islanding of power systems using the multilevel kernel k-means approach', J. Appl. Sci., 2009, 9, pp. 2247– 2255CrossrefGoogle Scholar 13Tsikalakis, A.G., Hatziargyriou, N.D.: ' Operation of microgrids with demand side bidding and continuity of supply for critical loads', Eur. Trans. Electr. Power, 2011, 21, pp. 1238– 1254Wiley Online LibraryWeb of Science®Google Scholar 14Gupta, P., Bhatia, R.S., Jain, D.K.: ' Active ROCOF relay for islanding detection', IEEE Trans. Power Deliv., 2017, 32, pp. 420– 429CrossrefWeb of Science®Google Scholar 15Rocabert, J., Luna, A., Blaabjerg, F., et al.: ' Control of power converters in AC microgrids', IEEE Trans. Power Electron., 2012, 27, pp. 4734– 4749CrossrefWeb of Science®Google Scholar 16Balaguer, I.J., Lei, Q., Yang, S., et al.: ' Control for grid-connected and intentional islanding operations of distributed power generation', IEEE Trans. Ind. Electron., 2011, 58, pp. 147– 157CrossrefWeb of Science®Google Scholar 17Gaonkar, D.N., Pillai, G.N., Patel, R.N.: ' Seamless transfer of microturbine generation system operation between grid-connected and islanding modes', Electr. Power Compon. Syst., 2007, 37, pp. 174– 188CrossrefWeb of Science®Google Scholar 18Ashabani, M., Yasser, A.R.I.M., Mirsalim, M., et al.: ' Multivariable droop control of synchronous current converters in weak grids/microgrids with decoupled dq-axes currents', IEEE Trans. Smart Grid, 2015, 6, pp. 1610– 1620CrossrefWeb of Science®Google Scholar 19Lu, L.Y., Chu, C.C.: ' Consensus-based droop control synthesis for multiple DICs in isolated micro-grids', IEEE Trans. Power Syst., 2015, 30, pp. 2243– 2256CrossrefWeb of Science®Google Scholar 20Hu, J., Zhu, J., Dorrell, D.G., et al.: ' Virtual flux droop method – a new control strategy of inverters in microgrids', IEEE Trans. Power Electron., 2014, 29, pp. 4704– 4711CrossrefWeb of Science®Google Scholar 21Xu, X., Mitra, J., Cai, N., et al.: ' Planning of reliable microgrids in the presence of random and catastrophic events', Int. Trans. Electr. Energy Syst., 2014, 24, pp. 1151– 1167Wiley Online LibraryWeb of Science®Google Scholar 22Staszesky, D.M., Craig, D., Befus, C.: ' Advanced feeder automation is here', IEEE Power Energy Mag., 2005, 3, pp. 56– 63CrossrefGoogle Scholar 23Jiaming, W., Dong, L., Ning, L., et al.: ' Distributed processing based fault location, isolation, and service restoration method for active distribution network', J. Modern Power Syst. Clean Energy, 2015, 3, pp. 494– 503CrossrefWeb of Science®Google Scholar 24' Power quality – frequency deviation for power system', Chinese national standard GB/T 15945-2008, 2008Google Scholar 25' Technical requirements for connecting microgrid to power system', Chinese national standard GB/T 33589-2017, 2017Google Scholar 26' Technical requirements for operation and control of microgrids', IEC/TS 62898-2, 2017Google Scholar 27' Technical connection rules medium voltage', VDE-AR-N 4110, 2017Google Scholar Volume14, Issue11June 2020Pages 2026-2033 FiguresReferencesRelatedInformation
Referência(s)