The need of transmission grid expansion has drawn the attention towards high voltage DC (HVDC) as a solution.This brings new challenges for power system control application such as state estimation that are originally designed for AC power system. This papers studies centralized and distributed architecturesfor hybrid AC/HVDC state estimator (SE). Furthermore, the benefit of having bad data detection (BDD) and deploying phasor measurement units (PMU) in the state estimation are investigated. The method used for the SE is the weighted least square (WLS) method. The SE will be developed based on the power grid model ’The CIGRE B4 DC Grid Test System’. Theresults of the tests show that the addition of BDD and PMU improved the error of the estimated values. Furthermore the distributed architecture offered slightly less accurate AC values than the centralized.

Recent issues such as priority access of renewable resources recommended by European energy directives and increase the electricity trading among countries lead to new requirements on the operation and expansion of transmission grids. Since AC grid expansions are limited by legislative issues and long distance transmission capacity, there is a considerable attention drawn to application of HVDC transmission grids on top of, or in complement to, existing AC power systems. The secure operation of HVDC grids requires a hierarchical control system. In HVDC grids, the primary control action to deal with power or DC voltage deviations is communication-free and local. In addition to primary control, the higher supervisory control actions are needed to guarantee the optimal operation of HVDC grids. However, the implementation of supervisory control functions is linked to the arrangement of system operators; i.e. an individual HVDC operator (central structure) or sharing tasks among AC system operators (distributed structure).

This thesis presents distributed control of an HVDC grid. To this end, three possible supervisory functions are investigated; coordination of power injection set-points, DC slack bus selection and network topology identification. In this thesis, all three functions are first studied for the central structure. For the distributed solution, two algorithms based on Alternating Direction Method of Multipliers (ADMM) and Auxiliary Problem Principle (APP) are adopted to solve the coordination of power injection. For distributed selection of DC slack bus, the choice of parameters for quantitative ranking of converters is important. These parameters should be calculated based on local measurements if distributed decision is desired. To this end, the short circuit capacity of connected AC grid and power margin of converters are considered. To estimate the short circuit capacity as one of the required selection parameters, the result shows that the recursive least square algorithm can be very efficiently used. Besides, it is possible to intelligently use a naturally occurring droop response in HVDC grids as a local measurement for this estimation algorithm. Regarding the network topology, a two-stage distributed algorithm is introduced to use the abstract information about the neighbouring substation topology to determine the grid connectivity.

This paper presents the results of an analysis of distributed two-stage coordination of network topology processor for HVDC grids. In the first stage of the two-stage processor, the substation topology is analyzed locally using an automated graph-based algorithm. Thereafter, a distributed algorithm is proposed to used the neighboring information to realize the grid connectivity. For distributed islanding detection, the connectivity problem is formulated as a set of linear equations and solved iteratively using successive-over-relaxation method. The performance of the proposed methods versus conventional one-stage method has been tested in an islanding scenario for a 5-terminal HVDC grid.

This paper presents the results of an analysis of distributed two-stage coordination of network topology processor for HVDC grids. In the first stage of the two-stage processor, the substation topology is analyzed locally using an automated graph based algorithm. Thereafter, a distributed algorithm is proposed to used the neighboring information to realize the grid connectivity. For distributed islanding detection, the connectivity problem is formulated as a set of linear equations and solved iteratively using successive-over-relaxation method. The performance of the proposed methods versus conventional one-stage method has been tested in an islandinv, scenario for a 5-terminal HVDC grid.

This study addresses the problem of voltage variations in active low-voltage distribution networks caused by distributed photovoltaic (PV) generation. Three strategies based on model predictive control (MPC) are introduced to flatten the voltage profile in a cost-optimal way. The compared strategies are the business as usual approach that manipulates a controllable on-load tap changer at the primary substation, the problematic feeder control strategy (CS) that adds an additional degree of freedom by controlling the critical secondary substations (SSs), and finally the compensation strategy, which controls the primary substation and compensates the non-critical SSs. A sensitivity analysis on the CSs has been conducted comparing the voltage variation reduction and the asset utilization with regard to the accuracy of the prediction models and the forecasted disturbance data. The results show that better (and more costly) characterisation of these parameters only provide a marginal improvement in the reduction of the voltage variations due to the restriction caused by the heavy tap change penalisation. Moreover, the tested case-study shows that the problematic feeder CS outperforms the compensation strategy in terms of larger voltage variation reduction for similar asset utilisation.

This paper proposes a method to be used by a DSOto optimally place sensors at MV/LV substation and some lowvoltage cable distribution cabinets. This method aims to improvethe estimation of the grid states at low voltage distributionnetworks. This method formulates a multi-objective optimizationproblem to determine the optimal meter placementconfiguration. This formulation minimizes the low voltage stateestimation error and the cost associated to a particular meterdeployment configuration. The method uses Binary ParticleSwarm Optimization (BPSO) to solve the optimization problemand it has been tested on a network based on the Cigré LVbenchmark grid. The simulation results show that the methodcan be applied to both situations where smart metermeasurements are available and situations where they are not. Inthe latter situation the measurements are replaced by pseudomeasurements,which represent meter readings by using smartmeter historical data and prediction models.

Secondary frequency control plays a vital role inpower systems and has therefore been the focus of muchresearch. Recent focus is being targeted towards developingdistributed solutions. This paper proposes a fast converging,distributed solution for secondary frequency control on the basisof the Alternating Direction Method of Multipliers (ADMM).For economic benefits the proposed solution integrates the useof rapid start units. Rapid start units are fast response and highramp offline units, which can provide the reserve power in caseof need. The proposed control scheme is a distributed solution tocombine secondary frequency control, economic dispatch and theRS start up process. Finally, the developed algorithm is tested fora power system model on a real-time co-simulation platform. Theresults show a fast converging algorithm that provides secondaryfrequency control compliant with ENTSO-E requirements, andthe economical benefits of the inclusion of a rapid start unit.

Smart Grid integrates communication and computationtechnologies into power systems. Research on interdisciplinarytopics between power system and communicationsystems requires models for both systems. There are severalpower system models available representing standard powersystems and real power systems. However, lacking of theircorresponding communication infrastructure information leadsthe difficulty of communication system modeling. In this paper,a communication infrastructure model compatible to Nordic 32power system model has been developed and presented. In thismodel, Wide Area Monitoring System (WAMS) is incorporatedwith Supervisor Control and Data Acquisition system. Theparameters of this model are based on the data from a Nordicelectric power utility. Validation of the model is performedthrough comparison of results from simulation and a previousempirical data study performed on the same utility network. Inadditional, this model is simulated by using different Qualityof Service (QoS) mechanism. The obtained results show thatcongestion management mechanisms introduce additional delayto the highest priority traffic slightly. But traffics with lowerpriority can be benefit from congestion management mechanisms.The proposed model can be widely used for other research onNordic 32 power system.

Being able to estimate the AC grid strength using Short Circuit Capacity (SCC) fromthe perspective of a connected HVDC station allows adjustment of converter control parameters orto select the converter’s operational control mode. The short circuit capacity can be calculated byestimation of the grid’s impedance and its equivalent voltage. This paper presents the operationaland practical challenges in real-time implementation of the grid estimation algorithms especiallyfor HVDC applications. This paper shows that the Recursive Least Square (RLS) algorithm can bevery efficiently used for the real-time estimation of SCC. This technique forms a regression prob-lem using algebraic complex equations with an objective to minimize the error between estimatedand measured parameters. The algorithm has been reformulated and simplified to make it non-complex without the use of matrices, in order to facilitate further implementation on an industrialreal-time controller. The performance of the real-time implementation has been evaluated usinga HIL platform. A sensitivity analysis has been also carriedout to study the impact of differentparameters and operational conditions on the performance of the estimation algorithm. Finally, theactual application of real-time SCC estimation for the HVDCsystems has been demonstrated.

This paper proposes a real-time quantitative evaluation of HVDC converters’ in an HVDC grid to select the suitable DC slack converter. This real-time evaluation considers the strength of connecting AC grid and the converter's on-line capacity margin as selection metrics. The strength of AC grid is evaluated in real-time by the estimation of grid short circuit capacity using recursive lease square algorithm. Given these selection metrics, the credibility of HVDC stations in controlling the DC voltage can be offered to the system operator in real-time for further operational decisions. This paper also studies the practical aspect of the estimation algorithm regarding selection of the operating points. As major contribution, it suggests to intelligently use a naturally occurring droop response in HVDC grids as a second operating point in the estimation algorithm to calculate the short circuit capacity. The method has been tested through set of scenarios using a real-time co-simulation platform. This platform includes real-time power system simulator to model AC/DC grid, industrial HVDC controllers and corresponding ICT systems. The results show that the proper selection of DC slack station can improve the AC system response and DC voltage drops during disturbances.

This paper studies various operational aspects ofrecursive least square algorithm as a potential solution for theestimation of short circuit capacity in HVDC application. Inthis work, low computational requirement, minor operationalcomplication and acceptable accuracy of the estimated gridparameters have been considered as the performance metricsfor the selection of the suitable estimation algorithm. The chosenalgorithm forms a regression problem using at least two algebraiccomplex equations based on two different operating points. Theappropriate selection of the second operating point which fulfillsthe minimum required accuracy and convergence rate is of greatimportance, particularly in the practical implementation. Thispaper proposes and analyzes several choices of providing thesecond operating point for the estimation algorithm adopted toHVDC application. Furthermore, a sensitivity analysis has beencarried out to assess the impact of different parameters andoperational conditions such as execution time on the performanceof the estimation algorithm for the HVDC control purposes.

This paper studies centralized and distributed schemes for the coordinationof power injection in an HVDC grid connected to large wind generation.This coordination of power injection aims to keep the power balanced withinthe HVDC grid especially during large wind disturbances. Furthermore, thecoordination tries to follow the converters’ schedules set by the connectingAC-TSOs every 15−minutes. This schedule comes from an overall combinedAC/DC economic dispatch calculation with lower resolution compared to thiscoordination. In this paper, the coordination of power injection has been formulatedas a non-linear constrained optimization problem for the centralizedarchitecture. Next, the centralized optimization problem is decomposed tosub-problems using the Auxiliary Problem Principle (APP) method for thedistributed architecture. This distributed optimization problem is solved byexchanging the required information between the AC TSOs. These two differentapproaches have been evaluated for a 5-terminal HVDC grid. Furthermore,a sensitivity analysis has been carried out to find the optimal updatingrate of power injection set-points. The result shows that the more frequentupdating of power injection coordination in the centralized architecture improvesthe generation reserves in each AC area.

This paper proposes a two-stage network processorfor the supervisory control of multi-terminal HVDC grid thatconnects different AC areas. The proposed network processorprocesses the DC substation topology locally in the first stageand analyzes the HVDC grid connectivity at the supervisorycontroller. The processor uses the k-means clustering methodto detect the islands in HVDC grid and prepares requiredinformation to carry out the converter control mode assignation.The performance of the method has been tested for an isladingscenario in a 7-terminal HVDC grid using a real-time cosimulationplatform. The output of network processor helps thecontrol application at the supervisory level to make decisionsbased on the control modes of the available converters in theislands. This two-stage architecture has been compared with thecentralized architecture in terms of computational complexity.

This paper proposes different architectures for the coordinationof power injection in interconnected hybrid AC/DC gridswith Multi-TSO. The power dispatch problem is first formulatedfor the HVDC grid with the aim of keeping the balancein real time and try to follow the 15-min schedules introducedby AC TSOs considering the ongoing changes in the HVDCgrid. Next, the Auxiliary Problem Principle (APP) method isproposed to decompose the dispatch problem and distributedbetween connected AC operators who control the local converters.Two different approaches are tested and the sensitivityof the OPF updating rate has been evaluated for a 5-areaHVDC grid with the wind production.

This chapter presents a cyber-physical approach to design of HVDC control system architectures and evolving HVDC grid operation and control modes. In addition, the chapter describes the communication system architectures needed for centralized and distributed operation and control of HVDC grids. Modeling and analysis methods suitable to analyze such systems using graph theoretic concepts, and also the design of distributed control systems utilizing a Multi-Agent approach and its dependence on the information graph theory. The chapter is concluded with a description of an application for distributed control of DC grids utilizing the concepts introduced. The application is presented both with regards to comparison with other design choices and analysis of performance and robustness of the algorithm versus communication metrics.

This work aims to study different ancillary servicessuch as frequency support, rotor angle stability and voltagestability that can be provided to independent AC areas throughMulti-Terminal HVDC grid. An agent-based control scheme hasbeen suggested to support the frequency of the imbalanced ACarea using other intact areas during a disturbance. The frequencysupport is carried out through some agents/controller, wherethe objective of the agents is to reach a common value fora given state, i.e. frequency, using the local and neighbouringinformation. In this scheme, the concept of Adaptive Graph (AG)has been introduced to the consensus problem to deal with thelimitation of original scheme in considering the rated capacityof HVDC converters. In this case, the frequency consensus canbe reached to a new stable value by removing one or severalconverters from the communication graph when reaching ratedactive power capacity. In order to test the proposed scheme,a model of 7-terminal dc-grid and corresponding AC areas isdeveloped in real time simulator OPAL-RT. The results showthat the adaptive graph scheme improves the reliability of thesystem during distributed frequency support.

Increasing the integration of renewable energy resources has introduced new requirements to the structure of electric power transmission grids. In addition to distributed renewable production, electric power exchange between neighboring countries also leads to development of transmission grids. Since AC grid solution is limited by legislative issues and long distance transmission capacity, High Voltage Direct Current (HVDC) technology with its different benefits compared to AC such as lower power losses, controllability and visual impact is being considered as appropriate alternative solution. Different aspects of HVDC technology from technical and economic challenges regarding the selection of proper HVDC grid topology to detailed control schemes are addressed extensively in the literature. Regarding the control of HVDC grids, the concept of centralized DC slack converter has been presented with the task to control the DC voltage level of the grid within the defined value and keep DC power flow in the grid balanced in real-time. Beside the centralized DC slack concept, a distributed DC slack concept called DC voltage droop is proposed and investigated. Similar to primary frequency control in AC grid, the controller uses the local DC voltage deviation as indication of power mismatch and tune the power production based on this local signal. An alternative agentbased control scheme has been also proposed to coordinate the power sharing in HVDC grids. On top of this fast control design, a supervisory controller is needed to guarantee the optimal operation of the HVDC grid in normal or abnormal states. This supervisory controller can be integrated to available AC Supervisory Control and Data Acquisition (SCADA) in the case of HVDC grid embedded in a single AC area. When it comes to HVDC grid connecting different AC areas, an “independent HVDC SCADA” can be defined which coordinates with connecting AC areas’ SCADA or it again can be integrated to one of the AC Area’s SCADA system called “integrated HVDC SCADA”. Regardless of system operation architecture, in order to monitor and operate the HVDC grid, supervisory control needs to be equipped with state estimation including the topology processor. Topology processor uses the real-time circuit breaker status within the substation to determine the system level topology. In the case of HVDC grid, the importance of such topology processor is bolder when it comes to detection of different islands. The islanding case should be recognized as real-time as possible in order to assign the DC slack bus, if not available in the island. In the literature, different algorithms and challenges have been investigated for topology analysis in the context of AC grid. This includes presentation artificial intelligence methods (e.g. artificial neural network) or integration of Phasor Measurement Unit (PMU) information in the process. Considering the future attentions toward overlay or interconnecting HVDC grid, the similar detailed methodology for HVDC grid with particular challenges seems to be vital. This work presents a multi-stage network processor as part of independent supervisory control in HVDC grid connecting or overlaying AC areas which can be extended to integrated supervisory control architecture. The proposed approach consists of two levels: 1) process of AC or DC substation topology locally and 2) the secondary process at HVDC grid central supervisory controller. At the substation level, the local topology processor determines the branch/bus model of the substation including the possible standalone HVDC converter using breaker status in the form of support matrix which is the combination of substation matrix and standing alone matrix. Substation Matrix holds information regarding this simplified substation model and the lines connected to it and number of logical buses/vertexes per substation. Standing alone matrix is created when the converter in the substation is not connected to DC grid and just standing alone. This can be useful for operator for starting the SVC functionality of the VSC stations. The same analysis is performed on all the substations locally and then the support matrices are sent to the network processor at SCADA level. At central level, the processor creates the general adjacent matrix based on the support matrix of the substations for whole HVDC grid to form the admittance matrix (Y). For the islanding analysis, the corresponding Laplacian Matrix is built from the adjacent matrix and clustering method is used to analysis the corresponding Eigen vectors of the Laplacian matrix. Besides, an extended version of current algorithm has been studied and presented for the integrated HVDC supervisory control architecture in which one AC area is responsible for the operation of HVDC grid.

The aim of the paper is to present the centralized architecture for power balancing management in an HVDC, High Voltage Direct Current, grid connecting different AC areas with high penetration of variable energy resources. Such a centralized high level DC Supervisory Control (DCSC) that functions in slower time scale compared to outer level controller has been evaluated in a real time co-simulation test-bed. The test platform includes OPAL-RT’s eMEGAsim real time simulator to model the power system, the ABB’s industrial HVDC controller (MACH), real time communication simulator OPNET to model the communication network and finally the DCSC application which is implemented on a Linux machine. The DCSC consists of a network topology manager to identify the grid configuration and employs an Optimal Power Flow (OPF) calculator based on interior point optimization method to determine the set-point values for all HVDC stations in a grid. The OPF calculator takes into account the DC voltage, converter and DC line constraints. The performance of the DC supervisory control has been tested for various test cases for a 7-terminal HVDC grid. Test cases include I) Variable power generation from wind farms, II) Station disconnection and III) DC grid islanding. Besides, the proper sampling rate has been chosen and justified to show the benefit of frequent updating of set-point compared to letting the DC droop control scheme take over the mismatch in the system. The results of different test cases show that a DCSC can improve the power extraction from wind farms by updating the set-points following any change in the system. Using a 3.2 GHz machine, it approximately takes 15 ms for the DCSC to converge to a proper solution and send the updated set-points.

This paper briefly introduces various features and challengesof HVDC grid development including topics such as voltageand power flow control schemes, operational responsibilitiesand multivendor interoperability problems. Second, it highlightsfactors to be considered in the design of the architecturefor the communication and control systems for the operationof HVDC grids. With these requirements as a basis, some examplesof control schemes, suitable for different operationalstrategies under development are presented.

The aim of the paper is to introduce an independentsupervisory architecture for optimal control of power flow inHVDC grids. Such an independent DC supervisory control isable to manage the optimal operation of connecting VSCs incooperation with other AC SCADA systems. Identifying the gridtopology, DC supervisory control employs an optimal power flow(OPF) calculator based on interior point optimization method todetermine the setpoint values for all HVDC stations in a grid.The performance of such an independent supervisory control hasbeen studied through various test case scenarios using a real-timeco-simulation platform.

This paper proposes a control scheme to coordinatethe DC voltage of converters in a HVDC grid using a Multi-agentSystem concept. Such interactions between agents, converters,and communication units are approached from a Cyber-PhysicalSystem (CPS) perspective. The Multi-agent based control schemetries to update the DC voltage set-points of the converters basedon local and neighboring information. The information exchangebetween the agents and communication delay tolerance of thecontrol scheme is studied by means of graph theory. The multiagentcontrol scheme has been implemented for a seven-terminalHVDC grid in a real time simulator and the impact ofcommunication delay on this distributed control approach hasbeen studied and is presented in the result.

Cyber Physical Energy Systems (CPES) development requires the combination of distributed intelligence to fulfill the future complex tasks and reach the increase the energy demands. Electrical Industrial Systems (EIS) are in continuous evolving integrating new technologies allowing to a better performance and increase the efficiency. This paper applies the consensus protocol for Multi-Agent Systems (MAS) to control the speed of multiple induction motors. In this paper, the behaviour of the system under different disturbances and scenarios has been simulated, thus, confirming the suitability and simplicity of this method for coordinating the control actions.

This paper presents work on the coordination of power sharing contribution of converters in an overlaid HVDC grid using a Multi-Agent System (MAS) approach. This approach is further implemented in a real-time co-simulation platform in order to study the proposed control scheme including the supporting information and communication Technology (ICT) systems. The platform consists of OPNET, a communication network simulator, connected to a real-time power system simulator through virtualized and real devices. Furthermore, the impact of different supporting system parameters such as bit-error rate has been studied using this real-time co-simulation platform.

Recent Information and Communication Technology (ICT) advances have enabled power system applications using measurement signals across the Wide Area Network (WAN). The application control performance relies on the quality of data delivery service. However, the characteristics of Quality of Service (QoS), such as latency, packet loss, and packet jitter, are unavoidable. It is a trend to take QoS metric requirement into the consideration of controller design. But, how to ensure the application receiving the data within the designed tolerant range is another challenge. This paper presents ongoing work on a novel Stateful Data Delivery Service (SDDS) for power system application to address the challenge from the side of the communication infrastructure. The SDDS monitors the QoS performance on-line and identifies the signals which satisfy the requirement for the application to use. As a proof of concept, a Power Oscillation Damping (POD) controller is connected to the SDDS. The result shows the improvement in robustness of the POD controller by application of the SDDS. The paper also shows the feasibility of applying SDDS to WAC applications.

PMU-based Wide Area Monitoring and Control(WAMC) system is introduced to improve the monitoring ofpower grid across large geographic areas and control the gridusing more efficient and smart applications. The performance ofWAMC applications in real power system scenarios and impactof their supporting Information and Communication Technology(ICT) on the data quality can be quantified and analyzed bypseudo-real co-simulation test beds. The purpose of this study isto propose and develop a WAMC testing platform to facilitate thereal-time simulation of dynamic power grid, the ICTinfrastructure that overlays the grid and WAMS applications.The platform consists of OPNET, a powerful communicationnetwork emulator, connected to a real-time power systemsimulator through virtualized PMU device. The end pointstations such as Phasor Data Concentrator or PMU-basedapplications are also linked to the platform through OPNET’sreal-simulation gateway called SITL (System-In-The-Loop). Toassess the performance of the platform architecture, a case studyhas been performed with five PMUs which collect the data from apower model and deliver to PMU-based mode-estimationapplication over a typical communication network. In this study,the results explicitly intend to quantify the effect of networkprotocols on data delay.

Due to the necessity of developing analysis tools inthe integrated ICT and power system domain, this paperdescribes and compares the architecture and configuration oftwo different co-simulation approaches called SITL and HLACo-simulation platforms. We provide several details on theirimplementation and present their features by means of theresults obtained for two test cases. The paper aims at clarifyingthe type of analysis that can be effectively carried out by usingthe two different platforms. Both platforms make use of thecommunication network simulator OPNET (Optimum NetworkPerformance). One of the two platforms operates in real time, inorder to perform system-in-the-loop (SITL) simulations. Theother platform is based on the interface between OPNET and thepower network simulation environment EMTP (ElectromagneticTransient Program). The interface is based on the application ofthe High-Level Architecture (HLA) standard.

This paper aims first to survey various aspects of DC grids from grid topologies to TSOs structure which has been carried out in several studies, and second try to highlight the factors to be considered in the design of the universal communication architecture for DC grid control and protection purposes. The result of the survey will be used as the first step to introduce the possible communication solutions such as network and data transfer protocols in order to ensure the reliable data delivery in operation of multi-national DC grids.

Synchrophasor technology is considered as a main enabler for the envisioned Smart Grid at the transmission level. Despite some preliminary field test results reported from China Southern Grid, the majority of the synchrophasor applications are still in the engineering or advanced laboratory tests phases. One possible impedance of any real-life adoption of such applicationsis a lack of test platforms that are able to verify the proposed application paradigms in cost efficient manners, and also to demonstrate the interactions between the power system and the supporting ICT systems with sufficient details. In this paper, we report on the continuing research efforts targeting at developing a test platform for the development of the synchrophasor based applications. The platform incorporates a virtual Phasor Measurement Unit (PMU), a real-time simulator for power systems, a real-time simulator for communication networks, and an implementation of an open-source Phasor Data Concentrator. Specifically, this article focuses on the implementation and verification of the virtual PMU, and on the extension of its functionality to incorporate communication over UDP.

Phasor -based Wide Area Monitoring and Control (WAMC) system promise to improve the monitoring of power grid across large geographic areas and accordingly controlling grid in more efficient and intelligent ways. In order to have confidence on the correctness, performance and reliability of applications used in WAMC system in real world scenarios, pseudo-real test beds where the requirements for the applications and their supporting Information and Communication Technology (ICT) can be quantified and analyzed. In this paper, a WAMC system testing platform is proposed to facilitate the real-time simulation of dynamic power grid, the ICT infrastructure that overlays the grid and WAMC applications. The proposed platform consists of OPNET as a powerful communication network simulator connected to real-time power system simulator on one side through virtualized Phasor Measurement Unit device, and linked to end point stations such as Phasor Data Concentrator or application on the other side through OPNET’s real-simulation gateway. A set of proof of concept scenarios using this framework is presented.

In order to restrain ferroresonance in coupling capacitor voltage transformers (CCVT), two kinds offerroresonance suppression circuits (FSCs) , are proposed in this paper: the resonant type, the fastsaturation reactor typewhich are connected to secondary side of CCVTs. Frequency domain analyses, for voltage ratio magnitude, are carried outfor CCVTs to obtain preferable FSC. Furthermore, using surge suppressors such as metal oxide varistors (MOV) toimprove the performance of the FSCs is investigated. Time domain simulations demonstrate that ferroresonance can becleared in few cycles in presence of FSCs. The Electromagnetic Transients Program (EMTP) is used for modeling and finetuningthe FSCs.

This paper proposes a distributed coordination of converters in an HVDC grid in order to provide close-to-real-time optimal DCvoltage and power profiles in the presence of wind uncertainty. This coordination is a part of HVDC grid secondary control actionthat carries out after the automatic primary control response by DC voltage droop controlling converters. It aims to optimize thegrid operation by finding the new DC power and voltage set-points that minimize the operational costs due to wind fluctuations andprediction error. Furthermore, the N-1 secure operation of HVDC grid is considered in this distributed coordination by adding a setof security constraints that tightens the upper and lower bounds of the problem variables. The coordination problem is formulatedas a convex optimization problem and therefore a modified version of alternating direction method of multipliers is adopted to solvethe problem in distributed manner. The performance of distributed coordination has been tested by a set of scenarios. The resultsimply that the distributed coordination is eligible in terms of time and complexity to be considered as either the main option forsecondary coordination or a complementary back-up solution for the central coordination.