Third progress report

WP2: Techno-Economic Evaluation


ECN: Based on the remarks of the Advisory Group, the inventory of the wind farm locations and sizes in the North Sea (ECN-E–10-072) has been modified and improved. Next, the development of the EeFarm model of the NSTG has been started with the construction of a test model for the DC grid, consisting of converters and cables, connecting the countries and transporting part of the wind power. Since the transnational part of the DC grid is bidirectional, while EeFarm was originally intended for unidirectional power flow from a wind farm to shore, EeFarm component will (if necessary) be modified and tested for proper bidirectional operation.


WP3: Multi-Terminal Converter


ECN:  In the nonlinear dynamic wind farm model several improvements have been made concerning the control of the PM-generator. The generator control now uses an estimation of the rotor flux position instead of assuming to know the exact flux vector. Secondly, a drive-train damping controller has been implemented, which also involves a controlled DC-link braking resistor to compensate for the generator power variations during grid transients and thereby limits the DC-link voltage.


A description of the wind farm model including simulation results is reported in " Nonlinear dynamic wind farm model with full-scale converter turbines connected through HVDC-VSC, model description", which will be published in April 2011.


Next steps are to extend the wind farm model with different types of WF converters and to investigate different control options for wind farms connected to MTDC grids. Also the model will be linearised so that it can be applied for designing MTDC controllers.


TUD-EPP: Regardless of the voltage-source converter topology used for the DC transmission system, all the VSC-HVDC stations have similar layouts.


The design of a scaled-down VSC-HVDC terminal can be accomplish by obtaining the necessary information about the interconnected AC system and the DC voltage level, based on the modulation strategy. Then follows the selection of the VSC topology with further selection of the solid-state components and of the converter switching frequency based on the desired controller bandwidth.  After that it is possible to design the phase reactor, which serves 2 purposes: limit the converter current ripple and decouple the dynamics of the VSC from those of the AC-system by use of a closed-loop current controller. Finally, the design of AC-side filters is needed to guarantee power quality. All the aforementioned steps are thoroughly explained and detailed throughout the report. 

Conclusion: The building of one or more scaled-down VSC prototypes offer some advantages, such as reduced cost with regard to commercial converters, the possibility to test the prototypes in hardware-in-the-loop simulations and the possibility of building a scaled down multi-terminal DC network, which is one of the goals in the NSTG project.

WP4: Multi-Terminal Converter testing


TUD-EPP: The idea behind the scale-down model of a MTDC network using real-time digital simulator serves several purposes; for instance, validation of the different models developed for the NSTG project and test of MTDC control strategies, among others.


An initial simple setup that would allow getting start with PHIL (power hardware in the loop) type of simulations is a point-to-point VSC-HVDC transmission system, where a model – can be an average or switching model – of the VSC sits inside the simulator, while the second VSC is constituted by a real physical converter. It is desirable that the real physical converter be custom made so as to emulate the AC and DC dynamics of current large VSC-HVDC stations, a feature which could not be readily available with commercial off-the-shelf converters.  The design of the custom made converter that would mimic the behaviour of large VSC-HVDC has already been object of consideration in a previous report.


For the establishment of the scale-down setup with a real-time digital simulator in the frame of the NSTG project there are two options available, viz.:

     The RTDS of the HCPS group of TU Delft.

     The OPAL-RT of the Joint Research Center (JRC) of the EU.


Based on the technical evaluation of both available real-time digital simulators for the testing purposes inside the framework of the NSTG project, the author recommends that the OPAL-RT system be used, due to two aspects: its apparent higher performance for the establishment of PHIL simulations and also because of its intrinsic convergence with already existing Simulink models, both from TU Delft and ECN.


WP6: Grid


TUD-EPS: The efforts to include the dynamics of dc networks accurately into stability-type simulations have been continued in the past months. As a result, a model has been developed to include VSC-HVdc into stability simulations by means of an averaged model based on vector control. The model is developed in Matlab, where the response to short-circuits in the AC network can be simulated. The accuracy of the results of the hybrid method can be compared with both electromagnetic-transient-type simulation and traditional stability-type simulation methods. In the next period, the hybrid simulation method will be extended toward generalized network structures and arbitrary phenomena, such as unbalances, harmonic distortion, and dc offsets.


On March 11, TenneT has provided TU Delft with a steady-state and dynamic model of the former UCTE transmission network. It contains a detailed model of the Netherlands, Belgium, and Germany, and simplified grid representations of more remote countries. Input from the SAG is needed to extend this model with a more detailed representation of other North Sea countries such as UK and Norway. The model was built in the software package PSS/e. The goal is to extend this model by including offshore wind power plants connected through VSC-HVdc, for the steady state (WP 6.1) as well as for the dynamic model (WP 6.2). Subsequently, the newly-created model will be used to study the impact of several offshore grid topologies on the dynamic behaviour in the mainland system for future wind power penetration scenarios.


One M.Sc. student (Line Bergfjord) who started in Feb. 2011, is currently transforming the wind speed data (obtained from the TradeWind project, 250x250 km grid) to wind power series, based on the scenarios developed in WP2 and additional information obtained from public sources. We are also negotiating with a Swiss meteorological data provider for a 3-year data set on a 9x9 km grid, which would provide a more detailed description of the wind regimes in the various locations in the North Sea. These wind power time series are the necessary input for WP 6.1. The activities for WP 6.1, which deal with transmission planning and operation concepts, are based on an extension of the methods developed in the RealiseGrid project, to work for mixed AC-DC grids. One PhD student, Ana Ciupuliga, who has finished her work for RealiseGrid end of March 2011, is expected to start working on WP 6.1 from April 2011. Her work provides the initial steady-state operating points for the dynamic simulations performed in WP 6.2. Another M.Sc. student (Paulo Chainho) also started in Feb. 2011, and will work until Oct. 2011 together with Arjen to develop a dynamic model of a multi-terminal HVdc network that can be included into PSS/e.