Connecting Belgium and Germany using HVDC: A Preliminary Study

1 Connecting Belgium and Germany using HVDC: A Preliminary Study S. Cole, Member, IEEE, D. Van Hertem, Member, IEEE, R. Belmans, Fellow, IEEE Abstract Although Belgium and Germany are neighbouring countries, there is no direct high-voltage line between them. A study that assessed the opportunity of an ac high-voltage line between the two countries was performed at request of Belgian, German, and Luxembourgian governments. The result of the technical study was that no substantial import capacity can be achieved. In this paper, the conclusions of a new study are presented. This study considers the connection of Belgium and Germany via an HVDC link. It is shown that HVDC can increase import capacity significantly. It would be the first HVDC land cable connection in the UCTE grid. Index Terms--HVDC Transmission, VSC HVDC, Power transmission. T I. INTRODUCTION HE European electricity grid is highly meshed. In this meshed grid, the Belgian transmission system forms one of the most important electrical crossroads in Europe, with significant international power flows taking a large part of the available transmission capacity. The European community already identified the congestion problems on the France- Belgium-Netherlands-Germany borders to be one of the axes for priority projects [1]. The congestion limits the internal market operation within Belgium. In view of increasing import capacity, existing international connections are upgraded. Recently the second circuit of the Avelgem-Avelin connection with France was installed. A second connection with France, Aubange-Moulaine, will be upgraded in the future. International East-West connections are currently non-existent: although Germany and Belgium are neighbours, no direct connection exists between the two countries (Fig. 1). It is therefore requested by Belgian, German, and Luxembourgian governments to assess the opportunity of an interconnection between Belgium and Germany. From the point of view of the Belgian government, import Manuscript received March 9, 2007. The research performed at the KU Leuven is financially supported by the Belgian 'Fonds voor Wetenschappelijk Onderzoek (F.W.O.)-Vlaanderen'. Dirk Van Hertem is a doctoral research assistant of the F.W.O.-Vlaanderen. S. Cole (e-mail: stijn.cole@esat.kuleuven.be), D. Van Hertem (e-mail: dirk.vanhertem@esat.kuleuven.be), and R. Belmans (e-mail: ronnie.belmans@esat.kuleuven.be) are with the Electrical Engineering Department of University of Leuven, Kasteelpark Arenberg 10, 3001 Leuven, Belgium. R. Belmans is chairman of the board of directors of Elia, the Belgian TSO. capacity is of utmost importance. The Belgium government is interested whether the connection would substantially increase import capacity. This would also enhance competition in the Belgian market. II. PREVIOUS STUDY A joint study has been conducted by Elia, the Belgian transmission system operator, and RWE Transportnetz Strom, the German transmission system operator in the area adjacent to the Belgian grid, to assess the opportunity of the requested interconnection between Belgium and Germany. Only ac overhead 380 kv double circuits were considered in this study. Two trajectories were investigated. The first one from Brume (B) to Dalhem (D), 65 km, and the second one from Lixhe (B) to Oberzier (D), 60 km (Fig. 1). The result of the technical study was that no substantial import capacity can be achieved, regardless of the chosen trajectory. The capacity found was in the order of magnitude of 200 MW. Optimization of the flows by phase shifting transformers that will be installed elsewhere in the Belgian power system, was taken into account in the study. The reason for this very limited increase in import capacity is that the power in Europe primarily flows along North-South paths. Moreover, from environmental, planning and regulatory point of view, major difficulties can be expected. The variant Brume Dalhem crosses the national park "Hohes Venn-Eifel", as well as many cities and villages. The Lixhe Oberzier variant crosses the densely populated area around Aachen. For these reasons this investment was not considered beneficial, and was abandoned. Summarizing, it may be said that the construction of a connection Belgium Germany was rejected for three reasons: limited import capacity; high environmental impact, e.g. crossing a national park; expected permitting problems. Nevertheless, the connection as such is very important for the Belgian government. III. PRESENT STUDY Having the conclusions of the previous study in mind,

2 other means of connecting Belgium and Germany are discussed. At present, several alternatives to overhead lines exist. In [2], an overview of technical developments for the future transmission grid is given. The use of underground ac cable instead of overhead lines would reduce the environmental impact. Obtaining permits is considered easier, but can nevertheless take a lot of time. The major drawback of underground ac cables is their substantially higher cost. Although reducing the high environmental impact and thus the permitting problems, a connection by underground ac cable solves in no way the third problem: import capacity will not increase by using underground ac cable instead of overhead ac lines. A connection with underground ac cable would therefore not justify a new study. The use of a High Voltage DC (HVDC) connection however would warrant a new investigation as neither the technical problem nor the problem of obtaining permits exists to the same extent when using dc cable connections. The import level can be higher: while the flow through an ac line, underground or overhead, cannot be controlled, unless expensive power flow controlling devices are installed, the flow through a dc link can be fully controlled. IV. HVDC Power electronic ac-dc converters can be divided in two main categories: Current Source Converter (CSC) and Voltage Source Converter (VSC), related to the operation of the valves and the dc bus voltage or current. For HVDC applications, the CSC converter is most widely spread, using the thyristor as its fundamental switching device. Installations using this topology are usually referred to as Line Commutated Converter HVDC (LCC HVDC). VSC converter technology is known for quite a long time from the variable speed drive technology, however the first VSC HVDC was built in the end of the 1990s, after development of self-commutating power electronic components such as the GTO and the IGBT for sufficiently high power ratings, and the increasing computational power of digital signal processors (DSPs) [3]. The distinct advantages of VSC (Voltage Sourced Converter) over LCC (Line Commutated Converter) HVDC are not of particular importance in a connection between Germany and Belgium. Black start capability for instance is not an issue. On the other hand, as price differences between the two technologies are converging, choosing VSC HVDC is an option. Fig. 1. Belgian grid (source: UCTE) V. TECHNICAL ANALYSIS A. Scenarios The goal of this chapter is to describe the various scenarios used throughout the study. 2012 and 2019 are selected as time horizons for the scenarios because data on these time horizons is available in the capacity plan [4] and the indicative program for generation [5]. For every time horizon, a network model was built and different load and generation patterns were selected, as well as offshore wind generation predictions. Combining the scenarios and retaining the most extreme ones resulted in four base scenarios (Table I). In each of the four scenarios, different import-export variants were tested. Only

3 the Belgian side of the grid was considered. TABLE I FOUR SCENARIOS Low gen, high load High gen, low load 2012 2012 Low 2012 High 2019 2019 Low 2019 High Increasing import capacity by a HVDC connection in the south of Belgium proved most difficult in the 2012 High and 2019 High cases. The structure of the Belgian grid is not suited for these scenarios. The 2019 High case is the worst as it is not N-1 secure for the base case. Therefore it is not considered in the remainder of the study. The network model used throughout the study consists of a reduced network of the neighbouring countries, France, The Netherlands, Germany, wherein a detailed Belgian network, comprising all 150, 220 and 380 kv network elements, is inserted. Planned upgrades are incorporated so as to represent the network as it is expected to be in 2012 and beyond. Three 380 kv nodes are selected for a possible HVDC connection with Germany: Reppel, Brume and Lixhe (Fig. 2). Although there is no 380 kv substation in Reppel, this node was retained due to the loss of generation in the neighbourhood that is expected in future. The other converter is placed in Oberzier, Germany. For each node it is calculated how much power can be imported and exported through the HVDC line in each of the scenarios without compromising N-1 and N-2 security in Belgium. security analysis as well as some elements in the neighbouring countries grids that have a large impact on the Belgian grid. It is not allowed to reduce the consequences of a single failure by switching operations. If, after a single failure, no element exceeds its limits, the network is considered N-1 secure. Every scenario is also checked on certain double failures. A double failure is defined as the loss of two generation units at the same time or the loss of a single generation unit combined with the loss of a network element at the same time. Elia does not considerer the simultaneous loss of two network elements because it is rare. In contrast to N-1 security analysis, one single switching operation is allowed when checking N-2 security. C. Results 1) Lixhe Lixhe is only connected to the rest of the 380 kv network by a single circuit line, Lixhe-Herderen (Fig. 3). In Lixhe, power is transformed down to 220 kv and 150 kv levels. When Lixhe-Herderen is not available, the power has to find its way through the weaker 220 kv lines between Lixhe and Rimière. The description makes it clear that the 220 kv network between Lixhe and Rimière is the weakest link. The worst case failure is a fault on Herderen-Lixhe when exporting to Germany. All other single failures would still allow importing or exporting 1.3 GW in the base case without repercussions on the safety of the grid. Fig. 2. Selected nodes B. Contingency analysis N-1 is a widely accepted criterion to provide an acceptable level of network security for a given cost. It is defined as follows: a single failure should not compromise network security. There are however numerous ways to implement this definition. This study uses the criteria as found in the capacity plan of Elia. A single failure is defined as the loss of a single generation unit or a single network element. The loss of a double-circuit overhead line is not regarded as a single failure because each circuit is considered as a single network element. All elements of the Belgian grid are included in the Fig. 3. The 220 kv network south of Lixhe A strengthening of the 220 kv network is obviously a solution but may be impractical. Extra transformers are needed in Lixhe, and also in Rimière when higher power exchange levels are envisaged. Another solution is building redundancy in the 380 kv paths by doubling the Herderen Lixhe circuit. In this way, a connection to the rest of the 380 kv network is guaranteed under N-1 and N-2 conditions, so that no high stress is placed on the 220 kv network and the transformers in Lixhe.

4 For all scenarios, it is assumed that the Herderen-Lixhe circuit is upgraded by adding the second circuit on the existing pylons, and that Aubange-Moulaine is upgraded as well, as this international connection to France is a limiting factor. Table II depicts the amount of power that can be imported to or exported from Lixhe under the different scenarios. The 2019 High scenario is not calculated because this scenario is not N-1 secure, even when no HVDC line is installed. TABLE II LIXHE Imp Exp Imp Exp 2012 L 700 1000 700 700 2012 H 0 1500 0 1300 2019 L 200 1300 0 200 When selecting Lixhe as connection point for an HVDC connection, two upgrades have to be accomplished: Herderen- Lixhe and Aubange-Moulaine. Even with these two upgrades, there is limited possibility for import to Lixhe. There is more room for export. In the 2019 High case, additional grid investments would be needed, regardless of the installation of a HVDC cable. 2) Brume A connection from Brume to Germany would create a path from France to Germany via Moulaine-Aubange-Brume with a capacity of 1700 MW in the scenario provided, thereby alleviating other flow paths from France to Germany via Belgium. An upgrade of Aubange-Moulaine is very beneficial for import capacity. Other reinforcements of the 380 kv grid are not necessary. Attention has to be paid to the 1100 MW pumped storage plant of Coo, which is also connected to Brume. At all times, the worst case scenario is assumed. This means that generators at Coo are operating at maximum power when importing from Germany through the HVDC link and that the turbines are pumping at maximum power when exporting. It can be seen in Table III that a connection to Brume would allow a good import level, especially when south-north flows are not too high. TABLE III BRUME Imp Exp Imp Exp 2012 L 900 850 900 800 2012 H 0 750 350 600 2019 L 200 800 900 300 Similar to the Lixhe case, Aubange-Moulaine has to be upgraded to allow import and export without problems. However, in the Brume case, more import is possible compared to the Lixhe case, and there is no need to upgrade Herderen-Lixhe. When abstracting from the 2012 High case, whose problems are explained, a reasonable import level can be achieved. 3) Reppel The rationale behind a connection to Reppel (figure 4) is the pending closure of the coal power plant in Mol. There is no other power plant that can supply power in the area. In that case the 150 kv line Beringen-Mol becomes overloaded, because generation has to be imported from other areas. The solution proposed by Elia is the construction of a substation in Reppel and a new 150 kv connection from Reppel to Overpelt. This new connection will relieve the Beringen-Mol connection by creating a second path to the region. Fig. 4. Reppel Besides these upgrades, additional measures have to be taken at the 70 kv level to prevent overloads in the 150 kv network. In the present study this could not be verified and therefore abstraction is made of all eventual local faults on 150 kv and below. Clearly this link will only be used for import as it has to replace a generation unit. Nevertheless under certain situations, limited export is possible, notably when northsouth flows are high. The results are depicted in table IV. TABLE IV REPPEL Import Export Import Export 2012 L 1100 0 1800 600 2012 H 1600 200 500 1300 2019 L 1600 200 500 1300 As can be expected, export from Reppel would be nearly impossible, except when large imports from The Netherlands feed power to Reppel. Import would be beneficial. Note that the figures in this section and in table IV are only valid when no power is injected in Mol. A disadvantage is the location of Reppel: a cable to Germany would most likely have to cross Dutch territory. VI. CONCLUSIONS In this study, possibilities for HVDC connections from Belgium to Germany are investigated. In general, the volume of import or export that can be achieved depends on future demand and generation of electricity within the Belgian

5 control area. A scenario with high generation and low demand in Belgium causes high flows in the grid. In this situation, only export is possible. In the other scenarios, both export and import is possible. When connecting HVDC to Lixhe or Brume, the international connection Aubange-Moulaine has to be reinforced in order to assure import or export. For a connection to Lixhe, the connection Lixhe-Herderen has to be upgraded as well. In general, more power can be imported to Brume than to Lixhe and more power can be exported from Lixhe than from Brume. The closing of the Mol power plant would justify an HVDC connection to Reppel. However, this would require serious investments in the 150 kv and 70 kv network. A new 150 kv connection Beringen-Mol has to be erected. Which investments exactly are needed at lower voltage levels can not be examined with the available data. Exporting to Germany would not be possible, except when large north-south flows occur, or when generation in Belgium would rise drastically. Ronnie Belmans (S'77-M'84-SM'89-F'05) received the M.S. degree in electrical engineering in 1979 and the Ph.D. degree in 1984, both from the K.U.Leuven, Belgium, the Special Doctorate in 1989 and the Habilitierung in 1993, both from the RWTH, Aachen, Germany. Currently, he is a full professor with the K.U.Leuven, teaching electric power and energy systems. His research interests include technoeconomic aspects of power systems, power quality and distributed generation. He is also guest professor at Imperial College of Science, Medicine and Technology, London-UK. Since June 2002 he is chairman of the board of directors of ELIA, the Belgian transmission grid operator. VII. REFERENCES [1] European Commission: Trans-European Energy Networks: TEN-E Priority Projects, 2005. Available online: http://ec.europa.eu/ten/energy/documentation/index_en.htm. [2] Cole S., Van Hertem D., Meeus L., Belmans R.: Technical Developments for the Future Transmission Grid, International conference on future power systems, ISBN 90-78205-01-6, Amsterdam, the Netherlands, 16-18 November, 2005; 6 pages. [3] Van Hertem, D., Cole S., Belmans R.: High Voltage Direct Current (HVDC) Technology technical report, 2007. Available online: http://www.esat.kuleuven.be/electa/vsc-hvdc/. [4] Elia, Ontwikkelingsplan 2005 2012, 2006. (Available in Dutch and French). Available online: http://www.elia.be. [5] CREG, Indicatief programma van de productiemiddelen voor elektriciteit 2005 2014, 2005. (Available in Dutch and French), Available online: http://www.creg.be. VIII. BIOGRAPHIES Stijn Cole received the M.S. degree in electrical engineering in 2005 from the Katholieke Universiteit Leuven (K.U.Leuven), Belgium. Since 2005 he is working as a research assistant at K.U.Leuven. He is a member of ELECTA, the Electrical Energy research group, of the department of Electrical Engineering of the K.U.Leuven, where he is working towards a Ph.D. His fields of interest include power systems, grid of the future and HVDC. Dirk Van Hertem (S' 02) graduated as a master of engineering in 2001 from the K.H.K., Geel/Belgium and as a master of science in engineering from the K.U.Leuven/Belgium in 2003. From 2003 he has been working towards a Ph.D. in the ELECTA research group, department of Electrical Engineering of the K.U.L., Leuven/Belgium. From October 2004, he is a research assistant for the F.W.O.-Vl. In 2001, his masters thesis received the `VIK award' and in 2004, he received the `KBVE R&D award' for his second masters thesis. His special fields of interest include power system control and optimization and FACTS.