Epameinondas Kontos

Impact of HVDC Transmission System Topology on Multiterminal DC Network Faults

This paper compares how a dc fault affects a multi-terminal dc (MTdc) network depending on the HVDC transmission system topology. To this end, a six-step methodology is proposed for the selection of the necessary dc fault protection measures. The network consists of four voltage-source converters converters radially connected. The converters natural fault response to a dc fault for the different topologies is studied using dynamic simulation models. For clearing of the dc faults, four different dc breaker technologies are compared based on their fault interruption time, together with a current direction fault detection method. If necessary, the converters are reinforced with limiting reactors to decrease the peak value and rate of rise of the fault currents providing sufficient time for the breakers to isolate the fault without interrupting the MTdc network operation. The study shows that the symmetric monopolar topology is least affected by dc contingencies. Considering bipolar topologies, the bipolar with metallic return exhibits better fault response compared to the one with ground return. Topologies with ground or metallic return require full semiconductor or hybrid breakers with reactors to successfully isolate a dc fault.

Providing dc fault ride-through capability to H-bridge MMC-based HVDC networks

This paper proposes a framework to achieve dc fault ride-through capability in multi-terminal dc networks (MTdc), when H-bridge multilevel modular converters (MMC) are used. The studied network consists of four voltage-source converters (VSC) for high voltage direct current (HVdc) transmission. Two of these VSC converters connect two offshore wind farms (OWF) to the main HVdc link between two asynchronous onshore grids, in a radial configuration. In case of a dc fault, H-bridge MMCs are able to block the fast developing currents and drive them to zero, allowing for fast mechanical disconnectors to isolate the faulty cable segment and reconfigure the grid layout. In this paper, the effect of the dc fault location to the grid behavior is analysed both at the fault isolation phase, as well as at the grid restoration phase. Moreover, the worst-case dc fault scenario for the studied network is identified. Finally, the total fault recovery time of the MTdc network is estimated. The study showed that H-bridge MMCs are unable to isolate the faulty part of the network without de-energizing the MTdc grid. However, the proposed framework allows for fast grid restoration within 3.6 s without the need for expensive dc breakers.

Optimization of limiting reactors design for DC fault protection of multi-terminal HVDC networks

In multi-terminal dc networks (MTdc) reactors are required to limit the rate of rise and the peak values of the fast developing currents in case of a dc fault. In this way, dc breakers have more time to isolate a fault and the system can restore its post-fault operation. This paper proposes a methodology to optimize the design of limiting reactors used for the protection of voltage-source converter (VSC) based MTdc networks. The limiting reactors were optimized using the covariance matrix adaptation evolution strategy (CMA-ES) optimization algorithm with two design objectives: first, the minimization of the reactor inductance value at the output of each VSC station to achieve N-1 security and second, the minimization of the reactors cost and mass and the peak dc fault current. Following the methodology steps, the effect of the dc fault location and the pre-fault power level of the converters on the dc fault network response are investigated in a four-terminal radially-connected grid. It resulted that the VSC stations controlling the dc voltage of the MTdc network are the first to respond to a dc fault and thus, limiting reactors higher than 97 mH are required in their dc output, in combination with dc breakers faster than 5 ms, to successfully protect the grid of the present case study.

Steady-State Loss Model of Half-Bridge Modular Multilevel Converters

Lesnicar and Marquardt introduced a modular multilevel converter (MMC) topology back in 2003. Although this topology has received a great deal of attention in recent years by both the research community and industry, hitherto no steady-state model has been developed which accurately captured all the relevant power losses while being computationally light. Hence, the aim of this paper is to introduce a fast MMC loss model which captures the key sources of power losses in steady-state operation. The proposed model was compared to a loss model developed by Marquardt. Both models presented similar results under the same assumptions. However, the proposed model captures the switching losses more realistically and considers the temperature of operation of the electronics, as well as the losses of the inductors and cooling system, in the overall efficiency of the MMC. To validate these new additions, the proposed steady-state model was compared to a dynamic model. Once again, the proposed model was able to capture the different sources of power losses. Nonetheless, results demonstrated that the balancing strategy greatly influences the efficiency of the MMC. Therefore, information regarding the envisioned control strategy is necessary to accurately calculate the efficiency curve of the MMC.

Fast DC fault recovery technique for H-bridge MMC-based HVDC networks

This paper proposes a post-fault control technique for H-bridge multilevel modular converters (MMC), which in combination with a restoration framework, can achieve fast operation recovery in multi-terminal direct current (MTdc) networks under different dc fault types. The studied network consists of four voltage-source converters (VSC) for high voltage dc (HVdc) transmission. A meshed MTdc grid topology is used for the connection of two asynchronous grids with two offshore wind farms (OWFs). The effect of different dc fault types on the grid restoration time is evaluated. Following the proposed restoration steps, the grid is able to restore its operation within 7.4 s in case of a pole-to-ground fault. This time is reduced to only 34 ms, if a pole-to-pole fault occurs. To overcome the main time constraint of the restoration framework, a control technique is proposed that allows momentarily the connection of the two dc pole cables through the MMC-VSC switch valves. In this way, the dc grid is discharged faster and thus, the operation can be faster restored. The study showed that, with the proposed strategy, an H-bridge MMC-based network can restore its operation after a pole-to-ground dc fault within 158 ms without the need for expensive dc breakers.

Multi-terminal network options for the interconnection of offshore wind farms: A case study between Britain and The Netherlands

The increasing global energy needs have led to the realization of interconnectors to strengthen regional electricity systems. As interconnectors are often placed offshore, combining those with wind farm grids can increase profitability compared to separate infrastructure and improve the grid integration, offering higher security of supply and controllability. However, a systematic approach for selecting the best transmission options for offshore multi-terminal networks is still missing in literature. This paper investigates options for interconnecting two offshore wind farms (OWFs) to two onshore asynchronous grids in a multi-terminal network. The options are presented for a case study comprising the Dutch and the British grid along with two OWFs in the coastal zone of each country. A review of the available technologies regarding high-voltage ac (HVac) and high-voltage dc (HVdc) transmission systems is given and, on this basis, six technical scenarios are identified and evaluated. For each scenario, the technical specifications are elaborated and a rough cost calculation is made. Based on the design possibilities for a four-terminal network, the technical bottlenecks and the most important research challenges for its realization are presented.

A fast steady-state loss model of a modular multilevel converter for optimization purposes

Lesnicar and Marquardt introduced a Modular Multilevel Converter (MMC) topology back in 2003. Although this topology has received a great deal of attention in recent years by both the research community and industry, hitherto no steady-state model has been developed which accurately captured all the relevant power losses while being computationally light. Hence, the aim of this paper is to introduce a fast MMC loss model which captures the key sources of power losses in steady-state operation. The model only requires information which is known a priori, e.g. datasheet information of the components. The proposed model was compared to a loss model developed by Marquardt. Both models presented similar results under the same assumptions. However, the proposed model captures the switching losses more realistically and considers the temperature of operation of the electronics, as well as the losses of the inductors and cooling system in the overall efficiency of the MMC. To validate these new additions, the proposed steady-state model was compared to a dynamic model. Once again the proposed model was able to capture the different sources of power losses. Nonetheless, results demonstrated that the balancing strategy greatly influences the efficiency of the MMC. Therefore, information regarding the envisioned control strategy is necessary to accurately calculate the efficiency curve of the MMC.

Control of a Modular Multilevel Converter STATCOM under internal and external unbalances

This paper develops a complete control scheme for the Modular Multilevel Converter (MMC) in a STATCOM application. The focus is laid on the internal balancing of the converter and its reliable operation under internal mismatches as well as under external unbalanced conditions. The leg energy controller ensures that the sum of the capacitor voltages of each phase are equal. An intuitive arm energy controller is also implemented which allows the decoupled control of the energy levels of every arm. The control performance of the STATCOM was evaluated under asymmetrical grid faults and the simulation results verified the capability to provide maximum reactive power to the AC grid while maintaining internal balance through the individual control of each arm energy level.

Analytical model of MMC-based multi-terminal HVDC grid for normal and DC fault operation

This paper proposes a new modeling approach for multi-terminal high voltage direct current (MTdc) grids which use the modular multilevel voltage source converter (MMC-VSC) technology. In this paper, primary focus is given to the mathematical analysis of the MMC operation and control, which coupled with the state-space model that describes the dc grid, can accurately simulate the normal operation and dc fault response of a dc grid with low computational requirements. A 3-terminal MTdc network with half-bridge MMC converter stations in radial configuration is used as a case study for the evaluation of the model and the obtained results are compared to a dynamic switching model implemented in Matlab/Simulink both for normal operation as well as in case of a positive pole-to-ground fault. The study showed that the model can accurately estimate the most important parameters needed for the design of an MTdc grid and can simulate fast dynamics without the need for computationally heavy software implementation.

Refurbishing existing mvac distribution cables to operate under dc conditions

Refurbishing existing ac distribution cables to operate under dc conditions can offer several advantages in terms of capacity enhancement, efficiency and flexibility in power and voltage control, among others. In this paper, technical aspects such as insulation ageing, capacity and efficiency enhancement are explored. A novel idea of dynamic cable voltage rating based on the temperature dependent electric field is developed. The proposed algorithm can be incorporated in the dc link converters to obtain additional efficiency and capacity gains. Finally, challenges in the presented concept are highlighted.