direct current: a future under which conditions?

DIRECT CURRENT: A FUTURE UNDER WHICH CONDITIONS? Dominique Laousse; Cedric Brogard, Paris; Hervé Caron, Christian Courtois, Saint Denis

Zusammenfassung DE Zusammenfassung EN Nowadays, railway power supply systems use either direct or alternative current to provide trains with traction energy. Both systems families have their advantages and drawbacks, which we will detail afterwards. However, we cannot overlook the quite common opinion according to which the future is quite narrow for the existing direct current systems, and especially about the 1500V or 3000V DC (Direct Current) system in heavy railways. This paper aims at studying the possibilities of future evolutions as regards the direct current systems. Different points were under investigation during an innovative design workshop whose goal was to define an exploration strategy based on the opened questions that follow: Do the direct current systems have a future in heavy railways power supply? Even more stimulating, can direct current systems become the cutting edge power supply systems of tomorrow? And, if yes, under which conditions? This paper demonstrates that the disadvantages of DC traction power supply systems are due to the low voltage and not to DC itself. Future industrial improvements of equipment could give to DC a new future that could enhance dramatically electric railway system. Any work on research and development on the DC subject will anyway help to improve the power supply systems of numerous networks in the world electrified in DC systems Zusammenfassung FR Les réseaux de chemins de fer électrifiés utilisent de nos jours les systèmes à courant continu ou alternatif pour alimenter les trains. Les deux familles de systèmes ont leurs avantages et inconvénients respectifs qui seront détaillés dans cet article Néanmoins, on ne peut souscrire complétement à l’opinion commune selon laquelle le futur du courant continu serait très restreint , particulièrement pour le 1500 V et le 3000 V, dans le domaine du chemin de fer appelé lourd. Cet article a pour objectif de décrire les possibilités d’évolution des systèmes d’électrification à courant continu. Certains aspects ont été investigués lors de séances de créativité et d’innovation ayant pour but de définir une stratégie d’exploration basée sur la question ouverte suivante : le courant continu peut –il devenir le système d’alimentation à la pointe de demain ? Et si oui, dans quelles conditions ? Ce papier démontre que les inconvénients attribués à la traction à courant continu ont pour origine la tension basse et non le continu lui-même. Les développements industriels des matériels pourraient donner au continu un nouveau futur qui pourrait fortement améliorer le système d’alimentation électrique de la traction. Tout travail de recherche et développement sur le sujet continu permettra de toute façon d’améliorer les systèmes de traction en exploitation des nombreux réseaux électrifiés en courant continu dans le monde. .

1.

INTRODUCTION

In this introduction, an overview of the situation regarding the direct current in the railway power supply landscape will be drawn, with a special focus on the French network. The historical development of the railway power supply will also be described briefly, leading to the current state of the art. This introduction will end with a short presentation of the innovative design approach which was followed to structure this work.

DIRECT CURRENT : A FUTURE UNDER WHICH CONDITIONS ? SNCF - C.BROGARD / H.CARON / C.COURTOIS / D.LAOUSSE (WORKING VERSION)

1

1.1.

HISTORICAL CONTEXT

To begin with, it is essential to remind that the direct current was not the only solution adopted for the first th electrifications, in France or in the world. Indeed, from the very beginning of the 20 century, several embodiments were experimented, using either direct current, single-phase or three-phase alternative current [Reference 1]. It is interesting to notice that this issue took place only a decade after the “War of currents”, which opposed Edison, promoting the direct current, and Westinghouse promoting the alternative current with Tesla. It is also important to point out that, at this period, public three-phase grid was not implemented and the available technologies made it hard to use the industrial current frequency of 50Hz, especially to achieve the power supply of the traction motors. This led to the development of single-phase alternative current systems at special frequency, for example 16 2/3Hz or 25Hz in the US [Reference 1]. Thus, the French “Compagnie du Midi” designed, implemented and operated a 12 000V 16 2/3Hz power supply system to equip its lines. The system was commissioned on several sections of its major line from Toulouse to Bayonne in 1913 and 1914, on the Pyrénées branch lines or on the line from Perpignan to Villefranche in 1912 [Picture 1]. Meanwhile, 12 000V 25Hz was experimented by the “Compagnie des chemins de fer de Paris à Lyon et à la Méditerranée” (PLM) on its line from Cannes to Grasse between 1910 and 1914. Contrary to the 12 000V 16 2//3Hz, this experiment was not followed by industrialization. But, as a consequence of the First World War, the French government imposed the use of another system than the German one to equip the national network: the 1 500V DC. This put an end to a long lobbying opposing the national railway companies in favour of the direct current and the ones in favour of the alternative systems. This political choice was confirmed in 1920, one consequence being the conversion of the Midi lines fitted with 12 000V 162/3Hz to 1 500V DC between 1922 and 1923. Only the Perpignan to Villefranche branch line remained operated in 12 000V 16 2/3 Hz until its conversion in 1984. In the wake of this political decision, the 1500V DC system expanded massively in France between the 1920’s and the 1940’s. Although the “Compagnie du Midi” briefly considered the possibility of raising the voltage to 2 400V to lower the costs [Reference 2], their economical analyses did not confirm the relevance on the studied projects. In 1922, they asked for a derogation to resort to 3000V DC to equip their trans-Pyrenées lines, but their demand was politically refused [Reference 3]. As a consequence, the system remained unchanged while the main lines were progressively fitted with the 1 500V DC system. Meanwhile, different choices were made in other countries, for example 3000V DC in Belgium, three-phase 3000V 162/3Hz [Reference 1] in Italy, or single-phase 15 000V 162/3Hz in Germany. The turning point in favour of the single-phase AC at industrial frequency systems occurred in the 1940‘s. After the first tests in Hungary, a single-phase 20 000V 50Hz system was experimented on the specific Höllental line in the west of Germany [Pictures 2, 3]. Meanwhile in France, SNCF validated the choice of the single-phase AC at industrial frequency to electrify the lines “with average level of traffic where the 1 500V DC would not be justified” th [Reference 4]. This choice was made during a meeting on the 7 June 1944. The concept of designing a railways power supply system using direct current at high voltage (HVDC) had also been studied in parallel in the 1940’s, with the same goal of cost reduction [Reference 5]. However, the concept ended stillborn, especially because of the incapacity to overcome technical issues with the technologies then available, for example regarding power conversion or current breaking [Reference 5]. This, aside with other considerations, led to the emergence of the now widespread 25 000V 50Hz system. It is nevertheless important to notice that the USRR carried out experimentations on a 6000V DC power supply system, on the line from Gori to Tskhinvali (33km) in the 70’s [Reference 10]. But, again, the available technologies at this time did not enable a significant improvement of the performances.

Just after World War 2, the Höllental site went under French occupation and SNCF carried on the tests, sending engineers and organising very frequent visits [Picture 4]. The decision was finally made in the late 1940’s to equip a test line with such a system: the line from Aix-les-Bains to La Roche-sur-Foron in the Alps was chosen and fitted. The experimentation was successful and the single-phase 25 000V 50Hz adopted as the reference system, leading to a major slowdown in the DC electrifications. At that time, this new system was considered as having the capability to “supply the heaviest trains under the lightest Overhead lines”.


2


3

[Picture 1 (top): Perpignan-Villefranche prototype overhead lines for 12 000V 16.7Hz. – source : Wikipedia –free license] [Picture 2 (top left): Höllental prototype rolling stock in 1947. This EMU was then brought to France to be operated on the Savoy line. – source: Centre des archives historiques SNCF] [Picture 3 (bottom left): Höllental overhead line. – source: Centre des archives historiques SNCF] [Picture 4 (on the right): Höllental line visit program for SNCF in 1947. – source: Centre des archives historiques SNCF]

1.2.

CURRENT SITUATION

In France, one consequence of the previous description is the coexistence of two electrification systems: the 1500V DC, which mostly expended until the 1950’s, and the single-phase 25 000V 50Hz, used for most of the recent electrifications, including the High Speed Lines. The French network thus includes a total of 15 500 km of electrified lines, among which 5800 km are equipped with 1500V DC, and 9600 km with 25 000V AC. The map [Picture 10] illustrates this situation, and the [Picture 11] gives an overview of the most usual systems in Europe. In this paper, the reference DC system will be the 1 500V, its main characteristics being described just below. However, it must be clear that the same reasoning can be applied to the other low voltage DC systems, among them the 3 000V DC. The system architecture is briefly described in the picture [Picture 12], and a focus on substations is given in the picture [Picture 13]. In this system, the common substation has a power rated between 4 and 11MW. Substations are generally located every 10 or 20 km distance, and they are connected in parallel. Due to the quite high current strengths, a high eq. cupper section of overhead line is necessary, usually between 400 and 800mm². To achieve this, the usual SNCF overhead line has two contact wires and, if necessary, one or several feeder(s), as shown on the two pictures [Pictures 14, 15] of SNCF Réseau overhead lines. This requested section is obviously a drawback of this system, the overhead line being expensive both to install and maintain [Reference 6]. But we will focus on these drawbacks in the second part of this paper.


4

[Picture 10 (on the left): Map of electrified network in France. In red 25000V 50Hz, in green 1500V DC -source: SNCF Réseau)] [Picture 11 (on the right): Map of electrified network in Europe -source: web (free copyright)


5

[Picture 12 (top left): Architecture of the 1500V DC system. – source: SNCF Réseau] [Picture 13 (top right): Substation principle in 1500V DC. – source: SNCF Réseau] [Picture 14 (bottom left): Regular SNCF overhead line in 1500V DC -source: SNCF Réseau] [Picture 15 (bottom right): “Midi” overhead line on the Bordeaux to Hendaye line. This OHL was designed in the 1920’s by the “Compagnie du Midi” for its electrifications in 1500V DC. -source: SNCF photothèque]


6

1.3.

INNOVATIVE DESIGN APPROACH

The work presented in this paper is the result of a structured and guided innovation process, which was used to build and organise the exploration strategy about the future of DC systems in railways power supply. To achieve this, the DKCP method (Define, Knowledge, Concepts, Proposal) was used to organize innovative design workshop, based on the CK (Concept, Knowledge) theory. This innovative design theory is a world class theory for industrial innovation elaborated by a research team of Mines ParisTech graduate school, led by Pr A. Hatchuel [Reference 7]. This team won the outstanding award of the world design society in 2009 according to their contribution to disruptive innovation understanding and innovative design engineering method proposal. The DKCP approach is particularly efficient in technical domains which need much more than incremental innovation. Its baseline is that disruptive innovation is related to new knowledge inputs (∆K) and new concepts elaboration (∆C) which induces an exploration strategy to dig new innovative ways made of projects and R&D operations. Within SNCF, the projects following the DKCP method and its declinations are called “Lab” or “Minilab”. This method has been used on various subjects in the SNCF since 2010 to rise disruptive innovation, with more than 45 projects dealing for example with infrastructure maintenance in dense areas, energy efficiency management, services strategies, signalling systems, accessibility or unmanned aerial vehicles. In practice, a Lab is organized in 4 complementary phases (Define, Knowledge, Concept, Proposals). As said earlier, the main idea is that radical innovation originates from new knowledge intensive mutualisation by participants (phase Knowledge), which leads to new innovative ways emergence. These ways can then be structured (phase Concept) and eventually declined into sets of projects, both exploratory for research purpose and industrial ready for implementation (phase Proposals). The number of participants depends on the subject, but can reach approximately 80 participants for biggest Labs, either from SNCF departments or sometimes from other structures such as research facilities or different companies. As regards the future of the direct current systems in railways traction power supply, approximately 25 participants were involved in the DKCP workshop named “Very High DC Current (VHDC)”, led by the Power Supply Fixed Installations Engineering Department and the Innovation & Research direction. In addition, the following SNCF departments were involved: Rolling Stock Engineering, Signalling Engineering, Telecommunications Engineering and Maintenance Facilities operators. Group diversity was an asset to consider all technical dimensions under a systemic view of power supply avoiding deep focus on problem-solving which often drive to incremental innovation. Moreover, several external experts (IFSTTAR, RTE, Alstom, Microelettrica, and Poseico) were included not only to share advanced knowledge but to identify critical knowledge and new issues as inputs for design process. During Knowledge phase, participants became parts of DKCP collaborative community link by common technical language about HVDC. During conceptualization phase, participants became contributors to elaborate technical scenario and identify components, existing or missing ones to be build. To be more specific on workshop outputs, project set to be launched was completed by knowledge expansion on HVDC with PhD thesis and R&D work to be pursued. Another point to be considered is that radical innovation highlights competence moves to be done to master new technical ways. More, some bonuses have been made, namely new functional opportunities like patented “overhead line start & stop” made possible by DKCP method focus on technical breakthrough. Some results of this project are presented in the following parts of this paper.

2.

CURRENT LIMITS OF DC CURRENT

The limitations of the current DC systems are at the beginning of this whole reflexion. They are listed in this paragraph, with a focus on the 1 500V and 3 000V systems. In the next paragraph, these systems will be systematically compared to the reference AC systems to evaluate the relevance of an enhanced DC system. To begin with, the 1 500V DC system suffers from an expensive infrastructure, due to a heavy overhead line and a high number of substations. Indeed, the low voltage level implies high current strength levels, hence the high overhead line section, and high voltage drops between substations, hence the high number of substations. Both installation and maintenance costs are impacted by these characteristics, which were an incentive to the choice of 25 000V 50Hz, and later 2*25 000V 50Hz, as the reference system in France. Then, the energy losses are important; they can reach 14 to 16% of the global traction power in 1 500V DC, far above the 4 to 6% in 25 000V 50Hz. Nevertheless, it is important to stress the point that these losses are the consequence of the low voltage level, not of the DC itself.


7

The substations themselves are more complex in DC, due to the need of rectifiers to convert the power. The circuit breakers and switchgears are also more complex in DC, all the more as the voltage rises. The same conclusion applies to the fault detection. Moreover, the high levels of current strength in the return current path can also generate corrosion in the surrounding of the electrified lines. Finally, the high level of current induces a heavy wear of the overhead line contact wires, and also of the pantograph strips, both inducing more maintenance operations. However, in spite of all these drawbacks, the numerous advantages of the DC systems must not be overlooked. The case of the tramways infrastructures are at this point a significant example. They show that the DC can, in some cases, be the best solution. As far as the tramways are concerned, DC at low voltage offers at the same time low isolation distances, which is an advantage in dense urban areas, and a more spacious and lighter rolling stock (no need for transformer on board).

3.

WAYS OF IMPROVEMENT OF THE CURRENT DC SYSTEM? 3.1.

GENERAL

Now that the general drawbacks and advantages of the classical DC systems were described, we are going to focus on these points in order to try to define the relevance window of an improved DC system, and its essential characteristics to be competitive. To do so, each component and interface of the power supply system was analysed, in order to point out the advantages and drawbacks of the classical DC system for this component or interface, in comparison with the others electric traction systems. To draw this comparison, three widespread and representative systems were chosen: - 25 000V 50Hz. - 15 000V 16.7Hz. - 1 500V or 3 000V DC. The comparison between these three systems led to introduce an “enhanced direct current system”, which could be built so as to erase the main disadvantages of the current DC systems, and would be the result of a technological breakthrough. This system is detailed in the last part of this paragraph.

3.2.

ELECTRIC TRACTION: SYSTEM AND INTERFACES

To compare the previously introduced electrical traction systems, a component based analysis table was adopted. The components in the scope are reminded on the [Picture 31], on which some of the interfaces between the components themselves, or between the components and the environment, were represented. It shows the classical architecture of the electric traction system. The components are linked by many interactions, which are not going to be described here except for those directly influenced by the type of electrical traction system: - the pantograph and overhead line interaction depends on many parameters, directly or indirectly linked with the voltage level. It dictates the intensity level that needs to be absorbed, which influences the temperature conditions at the contact, the material to be used, and the overhead line section necessary. - the return current path is in interaction with the environment of the tracks through the corrosion phenomenon. Indeed, the return current runs from the train to the substation, through the rails of the tracks which don’t have a perfect electrical insulation. Some current then leaks towards the ground or other metal structures (close pipes, overhead line supports...) and induces a corrosion effect, all the more as the intensity at stake is high. - the return current path can also be in interaction with the signalling system through the track circuits when the line is equipped with such systems. The harmonics from the rolling stock traction systems, contained by the return current, must be different enough from them so as not to deceive the track circuit. - the substation is in interaction with the mains to which it is connected. This interaction can be bidirectional if the substation is equipped to send back energy to the mains, for example in case of regenerative braking


8

(energy sent back by the trains to the substations or other trains around). In this case, the current sent back must be “clean” enough to fit in the mains, that is to say with a proper waveform, without too much harmonic currents. Finally, we cannot overlook the interaction between the substations of a same line section, not represented on this figure. Depending on the traction system, they will either be in parallel, as shown on the previous picture [Picture 12: Architecture of the 1500V DC system. – source: SNCF], or disconnected from one another. This can have consequences on the operating modes and their flexibility.

[Picture 31: Electric traction principle: main components and interactions. -source SNCF Réseau]

3.3. COMPARISONS BETWEEN ELECTRIC TRACTION SYSTEMS The idea was to compare the 1 500 V DC (the same applies to the 3 000V DC) with the two most widespread AC systems: the 25 000V 50Hz and the 15 000V 16.7Hz. For each component and interaction, the positive and negative points were listed regarding all three compared systems. The items for which there were more positive than negative points were highlighted in green. The idea being to give an organised overview of the main advantages and drawbacks, we did not go further, for instance by weighting each point in regard of its economical or operating impacts. The results are synthesised in the following array [Picture 32]:

[Picture 32: Comparative analysis of several electrical traction systems. - source: SNCF] A quick view of the array shows that DC systems display some advantages, especially as concerns the rolling stock, its weight (no transformer) and complexity (no need of rectifiers). Among the advantages, the simple power supply diagram and its consequences in operational flexibility can be pointed out. Moreover, a key advantage lies in the fact that there are no inductive voltage drops due to the jLω term of the impedance. Nevertheless, the usual drawbacks of the 1 500V DC clearly stand out in this representation, may it be the high number of substations, or the overhead line and its weight, or the higher losses in the traction circuit. The high level of current return and its potential consequences on the external structures in terms of corrosion must also been underlined. All these drawbacks strongly penalise the economical relevance of the 1 500V DC system, for example regarding the heavy overhead line, inducing higher investments (more components and more metal) and higher operational costs (more components to maintain and higher wear). The same reasoning applies to the high number of substations, influencing the investments due to a higher number of costly connections to the main grids, and the operational costs because of the maintenance needed by these substations.


9

As regards the traffic capacity, the situation is not so clear, with a greater operational flexibility offered by the 1 500V DC system and its power supply diagram. The need of phase separation sections in the 25 000V 50Hz turns out to be penalising in dense areas, implying time losses and potential incidents. In this case, the 15 000V 16.7Hz has the same advantage of the 1 500V DC as far as the supply of substations is centralized like in Germany, Austria and Switzerland.


10

25 000 V 50Hz

COMPONENTS & INTERACTIONS

POWER SUPPLY / MAIN GRID

1 500 V DC / 3 000 V DC

+ No unbalance on the main grids. - Possible unbalance on the main grids. Possible connection to “weak” parts of - Strong electrical connections necessary. + the main grids. + Low number of connections to the main - High number of connections to the main grids. grids (expensive).

15 000 V 16,7Hz

-

Dedicated railways grid : need of power converters at the connections between this grid and the main grids. Low number of connections to the main grids.

+

-

High number of substations : higher investments, maintenance and operational Low number of substations : lower costs. investments, maintenance and operational Need of rectifiers : impact on investment, costs. maintenance, reliability. Simple circuit breakers and switch gears. Complex circuit breakers and switch Simple detection of faults in the traction gears. circuit. Complex detection of faults in the traction Big substations (need of land). circuit. Small substations (lower need of land).

+ SUBSTATION

SUBSTATIONS INTERACTIONS

CURRENT SENT BACK TO THE MAIN GRID

= Medium number of substations. + Simple circuit breakers and switch gears. + Simple detection of faults in the traction circuit. - Big substations (need of land).

-

+ +

-

Simple power supply diagram : no phase separation sections, interesting in dense areas of traffic. Substations in parallel : easier to operate in case of incident on one substation.

+

Simple power supply diagram : no phase separation sections as far as the supply of substations is centralized. Substations in parallel , as far as the supply of substations is centralized: easier to operate in case of incident on one substation.

+

+

-

+

- Higher isolation distances : difficult implementation in urban areas, tunnels. - Complex impedance, and therefore

+ Lower isolation distances : easier implementation in urban areas, tunnels. + No jLω part of the impedance, and therefore no inductive voltage drops. - More losses (14-16 %) in the traction circuit due to low voltage. - Heavy overhead lines due to high strength of current : higher investments. - Heavy wear of the contact wire : maintenance work. - Two contact wires. - Heavy wear of the pantograph contact

- Higher isolation distances : difficult implementation in urban areas, tunnels. - Complex impedance, and therefore inductive

+ + -

Complex power supply diagram : need of phase separation sections. Complex power supply diagram : less flexibility in case of incident on one substation. Use of basic transformers to send back current to overhead line.

OVERHEAD LINE & inductive voltage drops due to the jLω part of the impedance. FEEDERS Low losses (4-5 %) in the traction circuit, (CURRENT due to high voltage. TRANSPORTATION) Light overhead lines due to high strength of current : lower investments. Low wear of the contact wire. One contact wire.

+ + + + +

CURRENT COLLECTION

ROLLING STOCK

REGENERATIVE BRAKING (SENT BACK TO OVERHEAD LINE)

CURRENT RETURN LEAKS AND CORROSION

Better current collection in case of ice on the overhead line. Light overhead line enabling high speeds. Low wear of the pantograph contact strips due to low strength of current.

+ +

- Heavy and space consuming transformers on board. - Need of rectifiers on board. + Simple circuit breakers.

No use of basic transformers : voltage Use of basic transformers to send back inverters needed, with harmonics generated. current to overhead line.

strips due to high strength of current. Limited speed due to heavy overhead line. Risk of contact wire fusion due to high strength of current, especially at standstill.

-

+ No transformer on board, therefore lighter rolling stock. + No rectifier on board : inverter directly connected to the overhead line, therefore lighter and more reliable rolling stock. Complex circuit breakers.

-

voltage drops due to the jLω part of the impedance. Low losses in the traction circuit, due to high voltage. Light overhead lines due to high strength of current : lower investments. Low wear of the contact wire. One contact wire.

+ + + + + Better current collection in case of ice on the overhead line. + Light overhead line enabling high speeds. + Low wear of the pantograph contact strips due to low strength of current.

- Very heavy and space consuming transformers on board. - Need of rectifiers on board. + Simple circuit breakers.

-

+

-

+

-

+

+

-

+

=

=

=

+/=

+ No phase separation sections, interesting

Necessity to adjust the phase of the current with the overhead line current.

Low levels of current returning to the substations due to high voltage. Low risk of corrosion due to low strength of current leaks.

INTERACTIONS Interactions when power electronics is WITH SIGNALLING used. SYSTEM

CONCLUSION

+

Allows a more powerful traffic if generously dimensioned.

No adjustment of the phase of the sent back current with the overhead line current.

High levels of current returning to the substations due to low voltage. Higher risk of corrosion due to high strength of current leaks. Interactions when power electronics is used.

Necessity to adjust the phase of the current with the overhead line current.

Low levels of current returning to the substations due to high voltage. Low risk of corrosion due to low strength of current leaks. Interactions when power electronics is used.

in dense areas of traffic

Finally, it seems clear that most of the drawbacks of the 1 500V DC systems are consequences of the low voltage level. It therefore seems that a system with an entirely “green column” in the previous array would be imaginable, using at the same time DC and a high voltage level. It would combine the AC and DC advantages and erase, in


11

most cases, the drawbacks of those systems by using a higher voltage level. This system, which can be called an “enhanced direct current system”, will be described in the next paragraph.

3.4. A RELEVANCE WINDOW FOR AN ENHANCED DC SYSTEM As it was pointed out before, an “enhanced direct current system” could combine the advantages of the DC systems and most of the advantages of the AC systems, and even enable new features. The main characteristics of this system are described here. To begin, the “enhanced direct current system” has two main attributes: it uses direct current… … at a rather high voltage compared to the current systems, that is to say at least 5 000V. Several consequences come directly from those two main attributes: - fewer substations than for a regular DC system. The DC substations architecture would indeed allow taking more power on the mains without unbalancing the three-phase network. - the substations are in parallel as for current DC systems. - the current strength levels induced are reduced thanks to the higher voltage levels, enabling to resort to a lighter overhead line, reducing the investments, reducing the maintenance work, enabling higher performance or speeds. Another consequence is the reduction of the return current strength, limiting the corrosion. This being said, the “enhanced direct current system” can still be declined in a great variety of potential systems. This variety can, for example, be generated by discussing three key attributes of the power supply systems: - the voltage level in the overhead line : it is maybe the most important parameter, because it sets the global architecture of the systems, in terms of substations spacing, overhead line section, voltage drops or rolling stock compatibility. The planned range of voltages goes from 5 000V, to make a minimal leap compared to existing systems, to minimum 25 000V. The consequences of this choice on the system components in the next part will be detailed in the next part. - the architecture of energy transportation, for example the number of feeder(s), the polarity and the voltage level(s) in each feeder(s). Indeed, if the feeder voltage is the same as the overhead line voltage, the architecture will be similar to the current DC systems architectures. But, if the voltage in the feeder is different (for example, higher), it can lead to an hybrid system with high voltage in the feeder and regular voltage in the overhead line. The architecture can then be seen, to some extent, as the DC equivalent architecture of the 2*25000V 50Hz system. An advantage of such a hybrid enhanced system is the higher distance between substations, because the feeder limits the voltage drops along the overhead line. The compatibility with the existing rolling stock and with the existing overhead line, given that the overhead voltage remains unchanged, is another advantage. the implementation of new functions to the system : the development of such a new power supply system should lead to the development of new functions due to the possible use of power electronics and digital technologies. Such ideas are currently under investigation, some of them having led to patent issuances. Among the considered ideas, there is a Start & Stop system to adjust the number of operating substations to the power demand by trains, or the potential use of the overhead line to carry signalling or other data. Obviously, the previous attributes enable many systems possibilities. To summarise, it is possible to describe two characteristic systems of this “enhanced direct current systems” family, which can themselves be declined in other embodiments: - the Full HVDC (High Voltage Direct Current) System : a new system using a high DC voltage, for example between 15 000 and 25 000V. This system would use a new overhead line design or the one of AC systems, new components (for example circuit breakers) and therefore new operational procedures and regulations. - the Hybrid HVDC System : a system using the HVDC only in the feeders, among other things to limit the number of substations, the section of the overhead line. The voltage in the feeder can for instance be raised to 10 000V or more. In fact, a chosen level between 5 000 and 25 000V or more can be imagined, with converters to inject the 1 500V or 3 000V in the overhead line. This system could use the current overhead line, or a new design. New operational procedures would be necessary, but the rolling stock is already compatible.


12

As we can see after this first description, both Full HVDC and Hybrid HVDC systems display many advantages and are promising. But many issues exist as far as those potential systems are concerned, not only technical but also regulatory and economical. This is why a roadmap was constructed, in order to try to address the various locks which were identified. This roadmap is described in the next part.

4.

RESEARCH AND INNOVATION TOWARD “ENHANCED DIRECT CURRENT SYSTEMS” 4.1.

GENERAL

After an overview of the “enhanced direct current systems” in the previous paragraph, and a first description of their main attributes and possible architectures, the identified issues which arise when dealing with such systems are going to be mentioned. To do so, the same analysis array was kept, and each component and interface of the power supply system was once more analysed. Some of the components or interfaces have been removed from this new array, because no work is foreseen as far as they are concerns. This is for example the case of the connection to the mains. On the contrary, some transverse works that need to be undertaken, or have already been undertaken, has been added to the array: economic studies, operational impacts, or regulatory constraints.

4.2.

OVERVIEW OF THE WORK FIELDS

The array in [Picture 41] gives, for each component or interface, an outlook of the work that would be needed to go toward an industrialisation of an enhanced direct current system. It also gives the specific difficulties identified as regards the concerned component or interface. Finally, possible actors or partners are indicated, even if they sometimes are only ideas. For some items, there are no identified difficulties, which can be normal when this only refers to a study that needs to be carried out. As regards the actors and partners, the list is not necessarily complete, for this column aims, above all, at showing that the SNCF Réseau Power Supply Fixed Installations Engineering Department will not be working alone on this roadmap. The content of this array is simply an extraction of the roadmap which was built in 2012, at the end of the innovative design work described in the introduction (“Minilab”). This roadmap contains, in addition to the projects, identified difficulties, actors and partners, and a provisional timing and articulation between all of this projects, leading to the experimentation of enhanced direct current systems in the next years. [Picture 41: Innovation and research program and roadmap towards an enhanced DC system. -source: SNCF]

[Picture 41: Innovation and research program and roadmap towards an enhanced DC system. -source: SNCF]


13

COMPONENTS & INTERACTIONS

WORK PROGRAM

WORK STATUS

IDENTIFIED DIFFICULTIES

ACTORS AND PARTNERS

• Control-command of substations and of a global line. • Faults detection in High Voltage Direct Current (HVDC) and treatment procedure of these defaults. • Current strength and voltage measurement in HVDC. • HVDC circuit breakers in substations.

• Early stages • Faults detection of studies (for in real time. all the points) • Circuit breakers for HVDC in substations.

• Main grids operators: RTE. • Industry: controlcommand.

• Flexible power supply diagram: matching the number of operating substations or traction groups and the effective power needs from trains in a line section: Start&Stop system, patent in 2014 by SNCF.

• Early stages of studies

• Rolling stock engineering.

• Reversible substation in direct current.

• Prototype in service

• Main grids operators: RTE. • Industry : power converters.

• Overhead line design compatible with a direct current, high voltage system: possible adaptation of the new DC overhead line designed by SNCF Réseau (project name CSRR). • New overhead line designed especially for an enhanced direct current system: new cantilever, new geometry. • Evolutionary overhead line: overhead line design capable of evolutions between several systems, for example: 25 000 V 50Hz towards 15 000 V DC.

•Industrializati • Insulating on materials.

• Research institutes: IFSTTAR, Laplace. • Main grids operators: RTE.

• Impacts of high voltage direct current on the current collection. This aspect should not be an issue for this is already managed in 25 000 V, even though at 50Hz.

• Not started

• Rolling stock engineering.

• Modifications on the rolling stock to be compatible with an enhanced direct current system using HVDC: new power converters, for example using electronic transformers? New traction architecture? Compatibility with existing rolling stock? • HVDC circuit breakers on board.

• Early stages • Circuit breakers of studies in HVDC in rolling stock. • Early stages of studies

• Rolling stock engineering. • Research institutes: IFSTTAR, Laplace. • Industry : power converters.

• Impacts of enhanced direct current systems on the current return.


LEAKS AND CORROSION

• Impacts of enhanced direct current systems on the current return.


INTERACTIONS WITH SIGNALLING SYSTEM

• Impacts of enhanced direct current systems on the current return and its compatibility with the existing signalling systems.


• Signalling engineering department.

• Operational impacts of enhanced direct current systems: reliability, available power, rolling stock performances. • Redaction of the new operating principles of an enhanced direct current system: protection, breaking procedures, maintenance procedures.

• Launched

• Rolling stock engineering. • Trains operators. • Infrastructure maintenance department.

•Industrializati on

ECONOMIC STUDIES

• Economic impact, at a global system scale, of enhanced direct current systems: evaluation of the investments, maintenance and operational costs, energy consumption, rolling stock costs (…) and aggregation in a few global indicators to help decision-making between several systems on a given project. A prototype work was carried out on an electrification project in 2014. The industrialization of the method to build a dedicated tool began in 2015. • Possible modifications of the TSI (Technical Specifications for Interoperability) to allow the implementation of a new power supply system in addition to the existing 25kV at 50Hz and 15 000 V 16,7Hz? • Impact of regulation concerning the voltage level used in the system.

• Not started

REGULATORY CONSTRAINTS

• Potential use of the overhead line to carry information and data for multiple uses: signalling information (from signalling component to component or from component to trains, or from trains to components), internet (to provide internet on board or in locations close to the line).

• Early stages • Data integrity of studies and recollection.

SUBSTATION

SUBSTATIONS INTERACTIONS

CURRENT SENT BACK TO THE MAIN GRID

OVERHEAD LINE & FEEDERS

CURRENT COLLECTION

ROLLING STOCK

CURRENT RETURN

OPERATIONAL IMPACTS

NEW SYSTEM FUNCTIONS


• Not started • Launched

• Launched

• Infrastructure maintenance department.

• TSI modifications.

• European networks. • UIC. • European commission.

• Not started • Signalling engineering department. • Rolling stock engineering. • Industry: power converters, communications.

14

4.3.

FOCUS ON SEVERAL POINTS OF THE ROADMAP

4.3.1.

Current breaking

As shown in the overview array, there are many issues that need to be studied in order to develop and implement an enhanced DC system. Obviously, depending on the characteristics of the concerned system in terms of voltage or energy transportation architecture, each point will be more or less critical. We can nevertheless highlight some of the points that are inevitable issues. One of the main technical issues concerns the circuit breaking function. With a system architecture similar to the current one, this function is necessary both in the fixed installations and on board. In both situations, the requirements imposes a big step in performance, for example concerning the response time, the reliability, the amount of energy to be cut, or the available space (especially on board). At this point, it seems that the technologies of DC circuit breakers do not enable to go further than approximately 8 000V or 9 000V for railways applications, without additional features. Nevertheless, options exist to overcome this limitation and raise the operational voltage. One of these solutions is two realise the circuit breaking function in an alternative current part of the system, where this is easier when the voltage comes to zero. This possibility is obviously adapted to situations where the structure of the system allows realising the operation on the alternative part of the grid, for example in substations connected to the mains. Another possibility is to resort to “hybrid” circuit breakers [Reference 9], that is to say a circuit breaker in parallel with a semiconductor device whose goal is to temporarily lower the current running through the circuit breaker. This device is synchronised with the circuit breaker and therefore enables the latter to operate within acceptable levels of voltage and current strength. Alstom Belgium developed some years ago the so called “DHR” which is a very High Speed circuit breaker as stated in the European Standard EN 50123 series. These two solutions can be considered as “bypass” solutions, for their principle is to artificially lower the voltage level before cutting the current. Yet they are promising solutions, in addition to the more “classical” ways of improving the performances of the component in itself to allow higher voltage levels, for example by modifying the insulating materials used in the circuit breakers, or by cutting the current in vacuum.

4.3.2.

Insulators fatigue

As far as the insulating materials are concerned, an issue that needs to be considered is the long term resistance of the overhead line insulators, undergoing a permanent solicitation from a high voltage DC in the overhead line or in the feeder. Even if the voltage levels would not be at first sight higher in DC than in the classical AC systems, the response of the materials may be different under DC conditions. As regards those two previous points, it is interesting to stress the point that some HVDC high voltage lines are already operated to transport electricity, for example in Canada or between France and Spain. In those cases, one advantage of this technology is to enable an interconnection between unsynchronised grids, or between AC grids using different frequencies. Those installations have overcome the previous technical difficulties and can be used as examples to a certain extent, even if the operational environments are different. This is why RTE (main grid operator in France) was part of the innovative workshops described earlier.

4.3.3.

Impact on rolling stock

Another issue that needs to be raised is the global electrical power architecture of the rolling stock designed to operate on an enhanced DC system. In case of high voltage in DC in the overhead line, some structural questions will arise, as for example whether to remove the classical transformers and switch to an electronic transformer [Reference 8]? In this case the same component could process both AC and DC at high voltage, and provide with an elegant solution for electrical interoperability. The weight would also be reduced subsequently by removing the traction transformers. Nevertheless, such architecture would provide no galvanic isolation. That would lead to a need of modification of the international regulations (the Technical Specifications for Interoperability) to certify alternative protections. This technology is currently under operational test, since a power-electronic traction transformer (or PETT) from ABB was installed on a SBB locomotive in 2011, replacing the existing traction transformer and GTO rectifiers [Picture 47].


15

As regards the general architecture of the rolling stock, several solutions are possible, raising many questions. For example: Which characteristics for the electric motors to be used? What type of electronic converters to work with them? What about the compatibility with the signalling systems (harmonics…) ? Should the regenerative braking be necessary? And if yes, with which components should it be realized? Is an on board circuit breaker necessary or is it possible to rely only on ground circuit-breakers? Which solution for an HVDC pantograph?

[Picture 47: SBB and ABB prototype equipped with power-electronic traction transformer. -source: ABB]

4.3.4.

Operating principle and integration

An enhanced DC system would require the construction of a dedicated set of operating principles, which would then be declined in operating procedures dedicated to the new system. From an even more global point of view, studies need to be undertaken to evaluate the implications of the implementation of a new electrification system in the landscape of existing systems. Obviously, a Hybrid HVDC System would offer an easy compatibility with the existing rolling stock and most of the existing infrastructure, and also some of the operational and maintenance procedures. On the contrary, a Full HVDC System would be more disruptive on all of these aspects and the others described in the array and consequently more difficult to implement on a large scale. The relevance window of such a system would therefore be different, but not inexistent. To give an example, the problem of cohabitation between this new system and the existing electrification systems could be overlooked in cases of lines disconnected from the network, or operated without, or with only few interactions with their adjacent lines. Still regarding a Full HVDC system, the choice of the voltage level could widen the relevance window by facilitating the compatibility with the rolling stock, for instance by setting a voltage of 15 000 or 25 000V in DC. Such a system, combined with the use of electronic transformers in rolling stock as mentioned before, would facilitate the cohabitation with the existing AC systems. Last but not least, it would also allow an easier migration from the existing AC system to the DC system with the same voltage in the overhead line.


16

4.3.5.

Zoom on 2x1500 V architecture

Of course, some of the highlighted issues have already been studied to various extents at SNCF. For example, the reflexions about the new operating procedures have already been initiated concerning a form of Hybrid HVDC system called 2*1 500V DC (see picture [Picture 42]), for an electrification project in the South East of France on the line from Marseille to Aix-en-Provence. A global economical analysis was carried out to compare several electrification possibilities on this project, in particular the 2*1 500V DC with the classical 1 500V DC system. In parallel, an evaluation of the operational performances of the 2*1 500V system was carried out on the MarseilleAix project, based on the dimensioning work, to compare the expected performances levels, mainly in terms of travel time. However, these kinds of studies need to be extrapolated to the other potentials forms of enhanced direct current systems. [Picture 42: Principle of the 2*1500V system. -source: SNCF]

Finally, other studies still need to be carried out, for instance as regards the influence of new power converters on the current return, Anyways, all of these issues involve various fields of competences and responsibilities, from SNCF engineering departments to regulation authorities, industrial manufacturers or research facilities, as indicated in the previous work program array. SNCF Réseau, acting as infra manager asks for a reliable, available system, with an economic view over a fifty year life cycle of its asset.

4.4.

IMPACT ON DIMENSIONING: FIRST EXPERIMENTATION

In addition to the work fields previously exposed, a last experimentation was carried out to validate the relevance of a Full HVDC System. The idea was to simulate an existing line operated in Full HVDC System, so as to compare the simulation with the current situation, in terms of energy efficiency and operational performances. A portion of the Paris-Lyon High Speed Line was therefore chosen. This section is currently equipped with a 2*25000V 50Hz system, with two substations in Sarry (at kilometer 164) and Commune (at kilometer 257), almost 100km away from each other. The quality of the voltage profile in this area is medium, because of the long distance between the substations combined with a very high traffic density such as a 20MVA train in a four minute headway on each track. The voltage drops in this area limit the power available for the high speed trains, lowering the operational performances. The power supply will therefore be reinforced in order to limit the voltage drops between the two substations, improving the performances of the high speed trains. When this is implemented, it will provide the reference voltage profiles shown on the picture 44. This picture actually shows a superposition of all the voltage profiles seen by the trains on this area during a period of 240 minutes. Even if the profiles are better than the current situation (not shown here), the voltage drops between the two substations are still quite important. This is the voltage profiles set that will be used to draw the comparison with the DC system.


17

[Picture 44: Voltage profiles superposition on Paris Lyon line 25000 V 50Hz – reference situation -source: Etienne Sourdille, SNCF Réseau]

For this reference situation, the average active power consumption, seen from the main grid, was calculated for a representative period of 240 minutes. The consumption is of 133,6 MW. Then, the voltage parameters of the power supply were switched from the previous 2*25000V 50Hz system to a new 25000V DC system. All the other parameters remained unchanged: the traffic density, the speed profiles, rolling stock characteristics… The Sarry and Commune substations remained at the same locations, but were set in parallel using several paralleling points as shown on the picture 45.

[Picture 45: substations and paralleling points for the 25000V DC situation -source: Etienne Sourdille, SNCF Réseau]

The picture 46 gives the superposition of the voltage profiles obtained with this new situation. They are clearly better than the ones of the reference situation in AC, with a significant reduction of the voltage drops between the two substations. Another interesting result comes from the average active power consumption, also seen from the main grid, and calculated for the same representative period of 240 minutes, with the same traffic hypothesis : the consumption for this 25000V DC situation is of 128,9 MW.


18

This gives a power consumption reduction of about 3.5% between the two situations, with the same operational performances and rolling stock characteristics. This reduction is therefore due to the use of DC instead of AC, reducing the inductive voltage drops (no jLω part of the impedance) and enabling the substations to be set in parallel. [Picture 46: simulation of voltage profiles superposition on Paris Lyon line – 25000V DC situation -source: Etienne Sourdille, SNCF Réseau]

Key to picture46: line to km point 164: 25000 V 50Hz From km 164 to 257: 25000 V DC From km 257 : 25000 V 50 Hz

POWER SUPPLY SYSTEM Mean voltage at the pantographs (same traffic)

Minimum voltage on the sector (same traffic)

Average active power consumption on 240mn (MW) (same traffic) Specific equipment

2*25 000 V 50Hz

25 000 V DC

23 000 V

25 000 V

18 000 V

25 400 V

133,6 MW

128,9 MW (-3,5%)

Feeders, autotransformers and paralleling points

Feeders and paralleling points

Distance between substations (same traffic)

Possible increase of traffic (same substations)

+30% (*)

+25% (*)

(*) in DC while keeping the same voltage profiles quality than in AC

To conclude with this experiment, it is important to point out that the simulation only deals with the electrical behaviour of the system, not taking into account the other advantages brought by a DC power supply, for example


19

the absence of phase separation sections, the impact of a lighter rolling stock, the absence of negative sequence voltage on the High Voltage mains (no unbalance)…

5.

CONCLUSION

As it was pointed out through this paper, there appears to be a space of technical relevance for an enhanced direct current power supply system in heavy railway systems. Such a system should benefit from the advantages from both alternative and direct current systems by using a higher level of voltage in direct current. Nowadays, even though technical difficulties still exist as regards the use of high voltage direct current in railways power supply, we can assume that such a system could be within technical reach, especially thanks to many technological leaps during the last decades, in power conversion or HVDC circuit breaking for example. Moreover, several high voltage lines are already operated in HVDC by main grids operators. As developed before, several projects dealing with these systems are already launched, or about to be. And even if there are still some uncertainties, we can already put forward two general outcomes: - there is a relevance window for an Enhanced Direct Current system, not only technical, but also operational and economic. - the easier the compatibility with the current systems, in terms of voltage levels, architecture of energy transportation, regulations frameworks, operational procedures, or rolling stock compatibility, the wider the relevance space. To go further, an interesting step could be, for instance, to build a full scale demonstrator of an Enhanced Direct Current power supply system, in the wake of the Höllental line we described at the beginning. To achieve this, a collaborative ecosystem gathering at the same time some European network train operators, infrastructure managers, manufacturers, research and innovation bodies could be a solution for long term implementation allaround railway operators. European authorities such as European Rail Agency should also take part to support innovative research works and see the link with European interoperability regulation.

BIBLIOGRAPHY [1]

JM.Allenbach, P.Chapas, M.Comte, R.Kaller, “Traction électrique, volume 1”, PPUR presses polytechniques, 2008, p.12-15

[2]

A.Ferrand, “Traction électrique par courant continu haute tension. Tension à adopter”. Revue Générale des Chemins de Fer, décembre 1920, p.313-352

[3]

C.Bouneau, “Modernisation et territoire: l'éléctrification du grand sud-ouest de la fin du XIXe siècle à 1946”, Fédération historique du Sud-Ouest, 1997, p.237

[4]

Jean-François Picard, « Fernand Nouvion, génie de la traction électrique », Revue d’histoire des chemins de fer, 2003, n°26, p.229-249

[5]

M.Garreau, “Aperçu sur le problème de la traction électrique par courant monophasé à la fréquence de 50Hz”, Revue Générale de l’Electricité, 1948, Tome 57, n°7, p.271-275

[6]

C.Courtois, “Does DC power supply have a future? », ESARS 2015, Aachen, Germany, March 2015.

[7]

A.Hatchuel, B.Weil, “A new approach of innovative design: an introduction to C-K theory”, ICED’03, Stockholm, Sweden, August 2003.

[8]

M.Claessens, D.Dujic, F.Canales, JK. Steinke, P.Stefanutti, C.Vetterli, “Traction transformation. A powerelectronic traction transformer”, ABB review 1|12.

[9]

J.-M.Meyer, A.Rufer, “A DC Hybrid Circuit Breaker With Ultra-Fast Contact Opening and Integrated Gate-Commutated Thyristors (IGCTs)”, IEEE Transactions on Power Delivery, vol. 21, num. 2, p. 646651, 2006

[10]

Wikipedia, “Электроподвижной состав на напряжение 6000 В” (consulted on 15 December 2015) https://ru.wikipedia.org/wiki/%D0%AD%D0%BB%D0%B5%D0%BA%D1%82%D1%80%D0%BE%D0%B F%D0%BE%D0%B4%D0%B2%D0%B8%D0%B6%D0%BD%D0%BE%D0%B9_%D1%81%D0%BE%D


th

20

1%81%D1%82%D0%B0%D0%B2_%D0%BD%D0%B0_%D0%BD%D0%B0%D0%BF%D1%80%D1%8 F%D0%B6%D0%B5%D0%BD%D0%B8%D0%B5_6000_%D0%92

[9]

these Andréa Verdicchio

Authors

Dominique Laousse Dominique LAOUSSE, (57), MBA, is head of "Innovation & prospective group" at SNCF . He is “Synapse” Expert (SNCF scientific and technical expert network). Responsible of innovative design and prospective/ foresight on mobility, transport and mutations of society influencing both. He works on disruptive innovation based on CK design theory and KCP Workshops (LAB) elaborated in close partnership with Mines ParisTech. Address: SNCF PARIS 40 AV DES TERROIRS DE FRANCE, DIRECTION SYSTEME & TECHNO FERROVIAIRE Phone: +33621244388 Mail: [email protected]

Cedric Brogard Cedric Brogard (28), civil engineer Mines de Paris; SNCF Innovation & Research department from 2012 and 2015; SNCF Network Maintenance since 2015. Address : Infrapole Paris Rive Gauche, 6 rue de Juvisy, 91200 Athis Mons, France Phone : +33 (0)6 14 39 50 98 Mail : [email protected]

Hervé Caron EB hat die Information

Christian Courtois EB hat die Information


21