Output : Transnational Manual on Advanced Energy Storage Systems. Part 2 Stationary energy storage systems for trolleybus systems

Transcription

1 Promoting Electric Public Transport TROLLEY Project Output : Transnational Manual on Advanced Energy Storage Systems Part 2 Stationary energy storage systems for trolleybus systems as of September 2013 Prepared by: Status: Dissemination level: Barnim Bus Company mbh (external expert: Fraunhofer IVI) City of Gdynia (external expert: PKT Gdnyia) Final Version Public Document The TROLLEY project is implemented through the CENTRAL EUROPE Programme co financed by the ERDF

2 This document has been prepared by the authors in the framework of the TROLLEY project. PART A: Calculation of the Electrical Network and Sketch Planning for the Energy-Efficient Operation of Trolley Buses of the Barnimer Busgesellschaft mbh Author: Barnimer Busgesellschaft mbh PART B: Analysis of Gdynia trolleybus network areas predisposed for the application of energy storage devices eg. "Supercapacitors". Author: City of Gdnyia Any liability for the content of this publication lies with the authors. The European Commission is not responsible for any use that may be made of the information contained herein. 2 of 72

3 Table of Contents PART A (BBG) 1 Vehicle simulation and Energy Consumption 1.1 Boundary conditions and assumptions of vehicle simulation Vehicle database Description of the simulation calculation vehicle simulation 1.2 Boundary conditions and assumptions Database of energy supply Description of the simulation calculation 1.3 Results Vehicle operation Use case 1 MAN NGE Use case 2 Solaris 2. Dimensioning of wayside energy storage systems on the basis of super capacitors use case MAN NGE Database 2.2 Presetting / Determination of an energy storage cycle 2.3 Description of the simulation calculation 2.4 Results 2.5 Energy storage costs and amortization 3 References 3 of 72

4 PART B (City of Gdynia) 1. Introduction 2. Traction network load 2.1 Analysis of traction network performance from a stationary point of view a classic power supply system 2.2 Analysis of traction network performance from a stationary point of view application of supercapacitors 2.3 Initial determination of supercapacitor location 2.4 Estimation of energy savings opportunities 2.5 Conclusions 3. Analysis of traction network performance: fleet perspective 3.1 Operation of registration and the contents of data collected by the fleet 3.2 Analysis of data collected by the fleet and pre-selection of areas with greatest potential for storing braking energy in a supercapacitor 3.3 Analysis of fleet data and pre-selection of sections with the highest voltage changes and needs for conceivable point support for the traction network power supply from the supercapacitor 4. Analysis of traction network performance from a stationary point of view a classic power supply system 4.1 Cause and nature of braking energy excess dissipation 4.2 Selection of the estimation method for braking energy possible to be stored in a supercapacitor 5. Determination of the optimal supercapacitor power and capacity 6. Summary 4 of 72

5 Introduction and Background The INTERREG Central Europe project TROLLEY Promoting electric public transport - contributes to an improved accessibility of, and within, Central European cities, focusing on urban transport. By taking an integrated approach the project has one main aim: the promotion of trolleybuses as the cleanest and most economical transport mode for sustainable cities and regions in Central Europe. The Central Europe project TROLLEY ( is one consortium of 7 European cities: Salzburg in Austria, Gdynia in Poland, Leipzig and Eberswalde in Germany, Brno in the Czech Republic, Szeged in Hungary and Parma in Italy. Horizontal support for research and communication tasks is given by the University of Gdansk, Poland, and the international action group to promote ebus systems with zero emission: trolley:motion. The project TROLLEY promotes trolleybus systems as a ready-to-use, electric urban transport solution for European cities, because trolleybuses are efficient, sustainable, safe, and taking into account external costs much more competitive than diesel buses. The project directly responds to the fact that congestion and climate change come hand in hand with rising costs and that air and noise pollution are resulting in growing health costs. Trolleybus systems are assisting with the on-going transition from our current reliance on dieselpowered buses to highly efficient, green means of transportation. Therefore, the TROLLEY project seeks to capitalise on existing trolleybus knowledge, which is truly rich in central Europe, where trolleybus systems are more widespread. The following document Transnational Manual on Advanced Energy Storage Systems presents the results of feasibility and simulation studies as well as real-life evaluation reports of TROLLEY s pilot studies and pilot investment in the area of advanced energy storage systems for trolleybus systems. The document exists of three parts: Part 0 presents a general introduction to different energy storage systems available in the market. Part I describes on-board energy storage systems and shows the evaluation results of TROLLEY s investment pilots installation of supercaps on trolleybuses in Parma and the installation of a lithium-ion battery on a trolleybus in Eberswalde. Part II illustrates the results of TROLLEY s feasibility studies for network based energy storage systems. It describes the dimensioning of a network-based energy storage system on the basis of wayside installed super capacitors in the networks of TROLLEY s partner cities Eberswalde and Gdynia. 5 of 72

6 PART A 1. Vehicle simulation and Energy Consumption The simulations presented in this report are intended to assess the potentials for the increase of energy efficiency of trolley bus operation of the Barnimer Busgesellschaft mbh (BBG) by saving hitherto unused braking energy by means of wayside energy storages and energy storages on the vehicles. This assessment is the basis for a selection of appropriate energy saving technologies, including their dimensioning for the exploitation of the available energy saving potentials. The conventional propulsion concept of trolley bus operation is based on allelectric buses and wayside infrastructure for the continuous energy supply of these buses. The concept consists of the following components within the vehicle: an electrical drive, a dc link, an internal power supply of the vehicle, a brake chopper and a pantograph. The wayside infrastructure is made up of catenary systems for energy distribution, which are installed all along the routes. They consist of: supply and return conductor to supply points, cross connections, section disconnectors to isolate adjacent sections of the trolley system and mast circuit breaker at supply points as well as facilities for energy supply, consisting of substations (SUB) for energy transformation from the medium-voltage power grid and feed-in into the direct current grid in the form of diode rectifiers and 6 of 72

7 underground supply conductors (supply and return conductor) to the mast circuit breakers. 1.1 Boundary conditions and assumptions of vehicle simulation Vehicle database The simulations have been carried out for two different trolley bus versions described in table 1 and table 2. Assumptions and estimations have been used in case of unavailable information and are clearly marked as such. Table 1: Used data for trolley bus version 1 - MAN NGE 152 NGE 152 type trolley buses were not equipped with onboard energy storages. In opposite to that Solaris type trolley buses are equipped with a super capacitor energy storage from LS Cable (table 2). In order to assess the energy saving potentials of the onboard energy storage both simulations with and without them have been carried out. 7 of 72

8 Table 2: Used data for trolley bus version 2 - Solaris Driving cycles Trolley buses in Eberswalde are operated on two routes (861/862) which have varying terminal stops (861-Nordend, 862-Ostend) and a joint terminal stop (Kleiner Stern, Schönholzer Straße). Between the stops»am Markt«and»Eisenspalterei«, vehicles from both routes run on an identical section of the network. The outward and return journeys of both routes are not identical, as route 861 (in the sections»eisenspalterei Schönholzer Straße«and»Ackerstrasse Nordend«) and route 862 (in the section»eisenspalterei Kleiner Stern«) both run in reverse direction on return journeys. Route lengths are resulting as shown in Table 3. 8 of 72

9 Table 3: Bus routes of the BBG For the design of electric vehicles in public transport, there is currently no standardized driving cycle. Therefore, the load profiles necessary for the simulation works were derived from measurement data, which had been collected using a MAN NGE 152 in regular service on the routes 861 and 862 from March 18 May 4, From the database developed thereby, the operational profiles for the routes 861 and 862 and the respective specification of the routes, which are necessary for the simulation calculation, have been derived (fig. 1). Figure 1: Operational profiles along the routes of BBG bus lines 861 and of 72

10 Efficiency from drive wheel to voltage link The distance of power transmission between the powered wheels of the vehicles und their electrical dc link contains several lossy transfer elements, whose degree of efficiency is either subject to the transferred power (propulsion converter and differential) or is additionally determined by the driving speed and the direction of the power flow (propulsion engine). As it was not possible to derive an efficiency map for all transfer elements depending on drive power and vehicle speed, distinguished between motor and generator operation, a respective efficiency map of a comparable, all-electric vehicle was used Description of the simulation calculation vehicle simulation Based on vehicle and route data, the resulting power demand of the driven axle of each vehicle was determined by means of conventional calculation algorithms for vehicle dynamics. Using an efficiency map for the power transmission between the driven axle and the vehicle voltage link (cf. chapter 1.1.1), the calculation of the resulting electrical propulsion and braking power was carried out on the dc link of the vehicles. The calculated results contain constant average requirements of electrical auxiliaries according to table 1 and table Boundary conditions and assumptions Database of energy supply Chainage for simulation Due to calculation reasons, it is necessary to clearly distinguish the respective route sections regarding the chainage (figure 2). Chainage means the presetting of route lengths that a vehicle has to run, depending on the bus route. In case of bidirectional laying of catenaries and their use on all bus routes on outward and return journeys, it is possible to carry out the chainage of the section only once. For the route network of the BBG, this can be done between the distinctive waypoints»viertelmeilenstein«(855/31770 m) and»eisenspalterei«(5745/10000 m). For the sections between»nordend«(0/29950 m) and»viertelmeilenstein«,»ostend«(19950 m) and»markt«(1945/22120 m) as well as the 10 of 72

11 section between»kleiner Stern«(7930/13970 m) and»eisenspalterei «, the chainage must be done separately for outward and return journey. Figure 2: Route sections in Eberswalde with chosen chainage By presetting the sections to be served in a defined order, the assignation of a bus route is realized within the simulation program. Substations and sections of trolley system The modeling of the electrical network was done on the basis of network documents as of 12/2005. The catenary network of the BBG consists of 10 sections, which are supplied by 3 rectifier substations via 13 supply points. The rated voltage of all substations was assumed at 680 V and the internal points of the substations, which were considered during network modeling. They are marked as mast circuit breakers (MS). Furthermore, table 4 describes the assumed catenary lengths of the connecting cables between the substations and the supply points. 11 of 72

12 Table 4: Supply points and catenary lengths of the connectors of the substations Disconnectors For the appropriate modeling of an electrical trolley system, it is necessary to know about the supply of the substation and the position of the disconnectors. In table 5, the disconnectors considered during network modeling are listed with the respective chainage. Table 5: Disconnectors in the network of the BBG Cross connections The precise positions of the available cross connections had not been modeled. An average clearance of 500 meters between the cross connections was assumed to be precise enough. Resulting electrical grid To illustrate the grid modeling, figure 3 exemplifies the part of the electrical network which is supplied by substation West. The modeled part consists of the substation, 5 supply points S1 S5 (mast circuit breaker MS9 13), supply conductors and 4 sections of the catenary system, which are electrically isolated to each other by 5 disconnectors. 12 of 72

13 Figure 3: Detail of the modeled electrical network (substation section West) 13 of 72

14 Electrical resistances The supply conductors between the substations and the supply points are running underground and have a wire cross section of 2*500 mm2. The material of the wires is aluminium. The specific electrical resistance (20 C) and a linear resistance temperature coefficient of K -1 lead to a specific electrical resistance of for the underground wire (10 C). With the lengths of the supply conductors according to table 4, the respective cable resistances (cf. Rzfx and Rzsx in figure 3) can be calculated. Thereby the same value for forward and return conductor is assumed. For the specific electrical resistance of a catenary, a considerable amount of data, partly also deviating from each other, can be found in the literature. Apart from the applied profile and the cross section of the catenary, parameters such as depreciation and ambient temperature may influence this value enormously. For the calculation of the electrical network, it was known that catenaries, type BRI 100, are used. According to [KIE 98], a specific electrical or the sections of the trolley system (cf. Rf+ and Rfin figure 3). Differing from the calculation of the light rail system operation, the same value for supply and return conductors is assumed Description of the simulation calculation Timetable scenarios A realistic description of the operation and the figures to be determined cannot be made on the basis of just one part of the annual timetable. However, to reduce the simulation efforts to an acceptable level, two scenarios have been selected after a detailed analysis of the timetable. By means of a respective weighting of the calculated energy consumption figures, an estimation of the figures to be expected when looking at the annual average is carried out within the evaluation of the simulation results. The two selected timetable scenarios are in the following referred to as»working day traffic«(scenario 1) and»weekend traffic«(scenario 2). Table 6 contains a description of both selected parts of the timetable. 14 of 72

15 Table 6: Selected parts of the timetable (scenarios) Thus, both selected scenarios do not differ in succession of buses, but by the interval of two successive departures. To illustrate the selected scenarios, the route network with the 7 defined route sections is shown in a satellite picture in figure 4 (source: Google Earth). In the scenarios»working day traffic«and»weekend traffic«, 3 vehicles each are running with departure times according to table 6 on the sections: 861 forward: E-A-C-G 861 return: H-C-A-F 862 forward: B-C-H 862 return: G-C-B Figure 4: Defined sections on the routes of bus lines 861 and of 72

16 For the averaging of the calculated energy consumption figures, shares of the overall timetable of 65% for the scenario»working day traffic«and 35% for the scenario»weekend traffic«are implied. In order to model the overall operation of one year, the increasing demand for auxiliary energy, depending on the ambient temperature, for the heating of the passenger compartment must be taken into consideration. For this purpose, both scenarios are calculated with different continuous auxiliary power demand according to table 1 and table 2 and averaged with the implied weighting of 42 weeks of summer operation and 10 weeks of winter operation. The weighting was derived from the analysis of the lowest daytime temperatures in 2010 in the city of Eberswalde (source: Lowest daytime temperatures of 0 C were hereby allocated to winter operation. At first, all calculations were carried out without the use of wayside energy storages. The detected amounts of energy consumption show the actual state within the network of the BBG. In a second step, energy storages were allocated to each of the three substations as parallel connections, completely storing the spare braking energy and resupplying it into vehicle operation. Only breaking energy that was not used to operate auxiliaries of the same braking vehicle, stored into the vehicle storage (if available) or used by other vehicles in the same substation section was stored into the wayside energy storages. 1.3 Results The results are listed in the following, both for the three substation sectors and the overall network Vehicle operation For an assessment of the calculation results, particularly regarding the estimation of utility of the use of wayside energy storages, characteristics of vehicle operation resulting from route and timetable design must be taken into consideration. Table 7 shows the respective figures 16 of 72

17 for both investigated timetable scenarios for the three substation sectors and the overall network. Table 7: Figures of vehicle operation (substation sections and overall network) It is noticeable that in the area of the substation West, a high share of the total mileage is produced. This can be explained by the high proportion of catenaries laid in this section compared to the overall network. Striking are also the much higher average speed which the vehicles reach in this section of the network. This results primarily from the comparatively wide distances between stops in this section. In figure 1, the speed profile of the vehicles in the area of substation West starting from 4700 m for route 861 out and from approx m for route 862 out are shown. The higher average speed in the scenario»weekend traffic«results from significantly lower times for passenger change compared to the scenario»working day traffic«. This fact is also considered in the timetable Use case 1 MAN NGE 152 Use case 1 considers the operation of MAN NGE 152 type trolley buses without energy storages on the vehicle. The according vehicle data is listed in table 1. Specific energy consumption The energy consumption of a vehicle comprises the traction energy and the energy for the operation of the auxiliaries. While during summer operation, the vehicles without air conditioning of passenger compartment investigated in this case have a very low energy 17 of 72

18 consumption for auxiliaries (hydraulic system, pneumatic system, internal power supply), this part of overall consumption can reach significant dimensions if the heating must be switched on during winter operation. Figure 5: Specific energy consumption; summer/winter (substation sections and overall network) Figure 5 shows the calculated energy consumption figures, distinguished between summer and winter operation. Results: While the section which is supplied with electrical energy by the substation»west«is characterized by a traction energy demand below average, vehicles within the section of substation»central«require comparatively more energy for traction. The reason for this is primarily the varying number of planned and traffic-related stops, which is significantly higher in the inner-city area of the substation»central«compared to the section of the substation»west«which features rather suburban traffic conditions. The differences regarding the average speed within the three substation sections (cf. table 7) also cause different specific energy consumptions for the supply of the auxiliaries. 18 of 72

19 The electrical heating of the passenger compartment during winter operation causes increased energy requirements of the vehicle. If the share of the auxiliaries within the overall energy requirements in summer is lower than 5 %, it reaches in average 43 % at low ambient temperatures. Energy consumption (without wayside energy storages) The energy demand of vehicles is primarily provided by substations. Thereby transfer losses occur due to cable resistances in the substations, their underground supply conductors to the supply points and in the catenaries. While the losses in the substations and in the supply conductors depend on the current, the ohmic losses in the catenaries are additionally influenced by the respective distance between a vehicle and the supply point. The described losses, in addition to the vehicle energy demands, must be supplied by the substations into the catenary network, which increases the energy consumption from the substations. A smaller share of the overall energy demand of vehicles is provided by energy exchange between braking vehicles and energy-consuming vehicles. The precondition for this use of braking energy is that at least one energy-providing vehicle (during braking) and one energyconsuming vehicle (e.g. during acceleration) are in the same substation section at the same time. Without knowledge of the actual power control during braking it was implied for the calculations that the braking energy of the vehicle is used at first for the supply of auxiliaries on the vehicle. Then power demand of other vehicles in the same substation section is satisfied. Finally still excessive braking energy must be transformed into heat on the brake resistor of the vehicle. Table 8 shows the calculated energy consumption figures for the investigated scenarios, subdivided into»summer«(su) and»winter«(wi) operation. Thereby the calculated transmission losses were allocated to the energy consumption. 19 of 72

20 Table 8: Specific energy consumptions without wayside energy storages Results: The higher average speed reached in scenario 2 causes lower energy consumptions by auxiliaries. Simultaneously, the specific value for braking energy used to operate auxiliaries is decreasing, too. Less vehicles in scenario 2 cause a reduced potential for braking energy used by an energy-consuming vehicle within the same substation section. For scenario 2»weekend traffic«, a higher specific energy supply from substations has been calculated for both summer and winter, compared to scenario 1»working day traffic«. 20 of 72

21 Using the calculated figures and taking into consideration the described weighting factors between both timetable scenarios as well as summer and winter operation, the specific gross energy consumption, the specific use of braking energy by other vehicles and the specific net energy consumption as average for the operation of vehicles throughout the year in the overall network can be calculated. Thus, a conclusion can be drawn regarding the accuracy of the simulation results as well as the assumptions and boundary conditions. Basis of comparison were the respective values from a vehicle of the BBG, which had been obtained in vehicle operation on approx. 780 Tkm. Table 9 contrasts the simulation results with the values from use in practice. Table 9: Energy consumptios (simulation and measurement data) The average energy requirements of the vehicles were determined to be approx. 2% higher than the empirical value from use in practice. This is considered to be a satisfactory convergence. For comparison: with the measured average speed of approx. 20 km/h, an increase of auxiliary energy demand by 1 kw leads to an increase of the specific energy demand by 50 Wh/km. Energy consumption (with wayside energy storages) The simulations were to determine possible reductions of energy supply from substations. This shall be achieved by storing previously unused braking energy, using wayside energy storages. Simultaneously, energy shall be fed into the vehicle operation in phases with power demand within a substation section. For the simulations, a wayside energy storage as parallel connection at the busbar of the substation was allocated to each substation. The wayside energy storage is controlled by analyzing the voltage of the substation. If this increases above the off-load voltage due to the braking of one or more vehicles, the excess braking energy of a vehicle is stored entirely, if possible, but reduced by the transfer losses in the grid. During phases of vehicle operation, in which due to acceleration or auxiliaries of 21 of 72

22 one or more vehicles a power demand exists within a substation section, the voltage of the substation sinks below its offload voltage. The allocated storage then provides at least a part of the electric power demand, which decreases the amount of energy provided by the substation, compared to vehicle operation without wayside energy storages. Vehicle operation is supplied with as much energy from the energy storage as is taken in during phases of power excess, reduced by the amount of internal losses of the energy storage. These internal losses will be discussed in detail within the dimensioning of the wayside energy storage system. Table 10: Calculated energy savings by use of wayside energy storages (ES) Table 10 shows the values for the specific energy supply by each substation and for the overall network with and without wayside energy storages. 22 of 72

23 Results: By using wayside energy storages, the overall energy consumption (sum of energy supplied by wayside storages and substations) is decreasing slightly. This is put down to reduced losses without further investigation. The use of wayside energy storages causes a much higher reduction of substation supply during summer operation. The reason for this is the smaller share of unused braking energy of vehicles during winter operation due to the increased energy demand of auxiliaries. Within the scenario weekend traffic, there is a higher potential for recuperation of unused braking energy, compared to the scenario working day traffic, primarily due to lower vehicle density and the reduced potential for using braking energy directly. Energy savings by using wayside energy storages The decision criterion for the operator of trolley buses is the reduction of energy supply by a substation. Table 11 summarizes the calculated energy savings, which can be expected by using wayside energy storages in addition to the existing supply by the substations, as annual average. The same weighting factors have been applied to consider both timetable scenarios as well as summer and winter operation. Table 11: Energy savings by using wayside energy storages (ES) It should be noted that during the simulations, only small excerpts of the timetable (cf. table 6) and only the driving of the vehicles have been considered. Changes in the specific supply by substations, resulting during vehicle operation from, e.g., 23 of 72

24 energy consumption of vehicle auxiliaries while staying in the depot, at terminal stops or during preparation or wrap-up of services, the influence of the varying altitude profiles on outward and return journeys, which can be eliminated during simulations only by choosing an unreasonably extended assessment period, the influence of the ambient temperature on the specific electrical resistance of the supply cables and catenaries and the energy consumption of auxiliary systems of the energy supply network, particularly in the substations, have not been recorded during the simulations. With the boundary conditions of an average annual vehicle operation of 48 weeks on each route, the recording of complete cycles of the timetable only, an averaging of 70% / 30% between school year and holiday operation (Mon Fri) and an energy price of 14 ct/kwh, monetary savings as listed in table 12 have been calculated. Table 12: Potentials for saving energy (monetary) by the use of wayside energy storages (annual average) Use case 2 Solaris Use case 2 considers the operation of Solaris type trolley buses with energy storages on the vehicle. The according vehicle data is listed in table of 72

25 Specific energy consumption Figure 6 shows the calculated energy consumption values, subdivided into summer and winter operation. Figure 6: Specific energy consumption; summer/winter (substations sections and overall network) The air conditioning during summer operation and the heating of the passenger compartment during winter operation cause a significant share of the total energy consumption. Auxiliaries amount for approx. 35 % of the total energy consumption during summer operation and approx. 43 % during winter operation. Energy consumption (without vehicle or wayside energy storages) Without knowledge of the actual power control during braking it was implied for the calculations that the brake power of the vehicle is used at first for the supply of auxiliaries on the vehicle. Then power demand of other vehicles in the same substation section is satisfied. Finally still excessive braking energy is transformed into heat on the brake resistor of the vehicle. Table 13 lists the calculated energy consumption figures for the investigated scenarios, subdivided into»summer«(su) and»winter«(wi) operation. The calculated transmission losses were again allocated to the energy consumptions. 25 of 72

26 Table 13: Specific energy consumption without wayside energy storages Using the calculated figures and taking into consideration the described weighting factors between both timetable scenarios and summer and winter operation, the specific gross energy consumption, the specific use of braking energy by other vehicles and the specific net energy consumption as average for the operation of vehicles throughout the year in the overall network can be calculated (table 14). 26 of 72

27 Table 14: Energy consumptions Energy consumption (with vehicle energy storages) The simulations were to determine possible reductions of energy supply from substations. This shall be achieved by storing previously unused braking energy into vehicle energy storages and by using it in case of energy demand. For the simulations, each bus had been equipped with an energy storage according to table 2 thus increasing the vehicle kerb weight. This led to an increase of traction energy demand by 1.3 % and to 0.8 % more total energy consumption. The energy management of the vehicle was modeled as follows: Supply of vehicle auxiliaries with braking energy Storing of as much as possible braking energy (amount depending on free energy storage) Transmission of not storable energy into the catenary if an energyconsuming vehicle is in the same substation section Still excessive braking energy must be transformed into heat on the brake resistor of the vehicle. Supply of energy demand by the vehicle storage until its lower state of charge limit is reached. Therefore the vehicle energy storage provides as much energy to the vehicle as it stores during braking, only reduced by internal charging and discharging losses. 27 of 72

28 Table 15: Calculated energy savings by use of vehicle energy storages (ES) Table 15 shows the specific values for the energy supply by each substation and for the overall network with and without vehicle energy storages. Results: The use of vehicle energy storages causes again a much higher reduction of substation supply during summer operation. The reason for this is the smaller share of unused braking energy of vehicles during winter operation due to the increased auxiliary energy demand. Within the scenario weekend traffic, there is also a higher potential for recuperation of unused braking energy, compared to the scenario working day traffic, primarily due to lower vehicle density and the reduced potential for using braking energy directly. Energy savings by using vehicle energy storages Table 16 lists the calculated energy savings, which can be expected by using vehicle energy storages. Again, the same weighting factors have been applied to consider both timetable scenarios as well as summer and winter operation. 28 of 72

29 Table 16: Energy savings by using vehicle energy storages (ES) Assuming an average annual vehicle operation of 48 weeks on each route, the recording of complete cycles of the timetable only, an averaging of 70% / 30% between school year and holiday operation (Mon Fri) and an energy price of 14 ct/kwh, monetary savings as listed in table 17 have been calculated. Table 17: Potentials for saving energy (monetary) by the use of vehicle energy storages (annual average) In total the use of energy storages on all BBG trolley buses would lead to annual savings of approx. 47 T. Additional savings can be obtained by reducing power peaks at the catenary network which reduces the demand charge. Excess braking energy The objective of the simulations was the assessment of both vehicle and wayside energy storages. Braking energy that cannot 29 of 72

30 be used by vehicle auxiliaries, be stored into the vehicle energy storages or be consumed by other vehicles in the same substation section could still be stored in wayside energy storages. However, the remaining amount of energy to be stored in wayside energy storages is very low and does not suggest the installation of such storages if the vehicles are already equipped with energy storages (cf. table 18). Table 18: Remaining energy saving potentials for wayside energy storages 2. Dimensioning of wayside energy storage systems on the basis of super capacitors use case MAN NGE 152 The wayside energy storage system consisting of super capacitors should be capable of storing excessive braking energy entirely, which can be supplied into the catenary network at subsequent power covering. 2.1 Database Table 19 contains data of the super capacitor cell type which is the basis for the configuration of the energy storage system examined. The configuration of an energy storage consists of a number of parallel phases consisting of serial single cells or modules. 30 of 72

31 Table 19: Data of the available super capacitor cell type The DC/DC converter for the super capacitor shall be designed as bidirectional boost converter to reduce the necessary power electronic effort. This means that the maximum voltage of the energy storage system must be lower than the respective minimal electrical voltage of the vehicle. Assuming the minimal terminal voltage of energy storage observed during simulation to be 620 V and considering a reserve of 20 V for the direct current chopper, a maximum operating voltage of 600 V for the entire storage system is calculated. Due to lifetime considerations, the maximum cell voltage is set at 2.5 V, which leads to the series connection of 240 single cells (phase). Assuming a minimal depth of discharge of 1.0 V, approx. 72 % of the theoretical energy content of a single cell can be used (2.5V² - 1.0V²) / (2.7V²). In this case, the series connection of 240 single cells (phase) has a voltage range of V, a usable energy content of kwh and a total weight of 132 kg. 2.2 Presetting / Determination of an energy storage cycle In order to estimate the necessary configuration of an energy storage, an energy storage cycle is necessary which should represent the maximum requirements within the substation section throughout a year. Average power output of energy storages (energy supply) can be drawn from the figures of vehicle operation in table 7 and the specific shares of stationary energy storage which have been calculated, cf. table 10. Table 20 summarizes the respective values, subdivided into the scenarios as well as summer and winter operation. 31 of 72

32 Table 20: Average power of the wayside energy storages (ES) Regarding the maximum cycle of power transfer, the timetable scenario»weekend traffic«during summer operation for the wayside energy storage at the substation WEST is taken as an example. The respective energy storage cycle (power at the terminal) is shown in figure 7. Negative power means the charging of power into the storage and positive power means a power supply from the storage into the substation. Figure 7: Energy storage cycle at the terminal of the ES at the substation WEST for scenario 2 during summer operation 32 of 72

33 2.3 Description of the simulation calculation Based on the described energy storage cycle, the necessary energy storage system is designed regarding the number of super capacitor phases which must be placed in parallel arrangement, maintaining an overall efficiency high enough for the energy storage system and the design of the power section of the direct current chopper as a necessary power electronic device between the terminals of the substation and its super capacitors. As the modification of parameters of the energy storage system is directly affecting its overall behavior within the network simulation, the design process must be iterative. The internal losses of the energy storage are taken into consideration, which are made up of different power losses: ohmic drops in the super capacitor cells, ohmic drops in the connectors between the cells and between the parallel phases, switching and transmission losses of the main elements (power switches, diodes, converter inductors, balancing capacitors) in the power section of the direct current chopper and power consumption of auxiliaries (cooling system, control electronics). The individual power losses depend on numerous parameters, varying with the arrangement of the super capacitors and the selected power electronic connection. The individual parameters shall not be discussed in detail here, but the overall losses in the super capacitors and in the power section of the direct current chopper are also included in the presentation of the results below. 2.4 Results The number of single phases in parallel arrangement increases the overall capacity of the energy storage system. The average voltage level of the energy storage increases when the power cycle is constant, causing the current load of the energy storage system to be reduced. This leads to reduced losses during the process of charge and discharge both in the super capacitor as well as in the power section of the direct current chopper. Table of 72

34 summarizes the determined dimensions of the energy storages (number of single phases) as well as the calculated power losses and resulting storage efficiencies. Table 21: Power losses and efficiencies for the selected energy storages In order to calculate the storage efficiency, the overall losses, which occur during charging and discharging of the energy storage, are put in relation to the amount of energy discharged from the energy storage system. The configuration of an energy storage with 6 super capacitor phases, arranged in parallel connection and consisting of 240 cells each, is regarded as a sufficient dimensioning for the energy storage of substation WEST. A further enlargement does not lead to a significant increase in efficiency for the overall system, but increases cost in procurement and maintenance. Figure 8: Voltage cycle at the terminal of the ES for the selected configuration Figure 8 shows the voltage cycle at the terminal of the energy storage of substation West. Besides an adequate storage efficiency, the size of an energy storage also determines the current load cycle of the super capacitors at a given load profile. This aspect must be considered when designing the energy storage system, particularly due to reasons of 34 of 72

35 lifetime. Figure 9 exemplifies the resulting current load of the energy storage system at substation WEST. The average current load is regarded as being noncritical in terms of premature aging of the super capacitor cells. Figure 9: Current load of the energy storage (terminal current and RMS value of a single phase) 2.5 Energy storage costs and amortization In order to assess the economic efficiency of the procurement and operation of a wayside energy storage system, the monetary gains of energy saving to be expected must be put in relation to the procurement and maintenance costs of the system. Based on data provided by Fraunhofer IVI, the procurement costs for a super capacitor energy storage system as discussed above can be estimated, with sufficient precision, at 150 % of the price of the cells. The annual maintenance costs were calculated at 2 % of the procurement costs. Table 22 compares the calculated monetary savings (cf. table 12) to the estimated procurement and maintenance costs of the energy storage. Furthermore, the payback periods, assuming an interest rate of 4 %, are listed. Table 22: Estimation of the economic amortization 35 of 72

36 3. References KIE 98 KIESSLING, Friedrich; PUSCHMANN, Rainer; SCHMIEDER, Axel; SCHMIDT, Peter: Fahrleitungen elektrischer Bahnen : Planung Berechnung Ausführung. 2nd revised edition; Stuttgart: B.G. Teubner, ISBN X 36 of 72

37 PART B 1. Introduction Trolleybus transport in Gdynia was first launched in Three traction substations Grabówek, Dworzec and Redłowo were built to power the first line. After the end of the Second World War there followed an intensive development of this means of transport. In there were put to start two traction substations Sopot and Sopot II designed to supply power to a suburban line Gdynia Sopot, existing objects being also expanded. In the beginning of the 1970s the first makeshift substation Cisowa has been put in operation. 1980s brought the next period of trolleybus network intensive expansion, at the time of which there were built further two substations Chwaszczyńska and Północna (to replace the makeshift substation Cisowa ). In 2010 there began a modernization of the traction network power supply system, in the framework of which there were built five new traction substations: Kielecka, Wielkopolska, Sopot Reja, Plac Konstytucji and Wendy, as well as existing objects were modernized. This investment was to increase reliability of existing system and create a basis for further development of trolleybus transport in Gdynia substations Kielecka, Sopot, Sopot Reja, Redłowo and Wielkopolska are predicted as to power new planned trolleybus routes. 2. Traction network load 2.1. Analysis of traction network performance from a stationary point of view a classic power supply system The traction network together with the other elements of the power supply system is a medium for electric energy transmission for trolleybus power supply, both that derived from the traction substation rectifier units and vehicles braking in a regenerative way. As a result the characteristics and operation of the power supply system are crucial for the course of recuperation process in the trolleybus network. To enable the recuperation of energy generated during braking it is necessary to fulfil two main conditions: 37 of 72

38 the voltage generated by the vehicle during braking must be higher than the voltage of the traction network, there must be a recipient of the energy generated during braking. Fulfilment of the first condition requires on the part of the trolley drive system the creation of voltage on the collector higher than the nominal power supply voltage of the vehicle. The second condition involves the energy conservation principle, which implies the need to balance the energy given to and collected from the traction network. In the case of classic trolley substations, which are equipped with diode rectifiers, the only energy recipient may be another vehicle extracting electric energy. Another aspect related to the fulfilment of the second condition is the existence of possible flow of energy between the recuperation vehicle (braking) and vehicle collecting energy. Therefore, there must necessarily exist a galvanic connection between the vehicles, and consequently the vehicles must be on the same power supply section, or two different power supply sections yet powered from one traction substation. In case of bilateral power supply of the traction network vehicles can be found in the area of two different substations, however, bilateral power is rarely used for the trolley traction. Figure 1 shows the flow of energy from regenerative braking for vehicles that appear in one section of the power supply. In this case, energy is carried by the traction network directly, without the involvement of the traction substation. This is an optimal variant of recuperation, characterized by the smallest transmission losses. In real conditions, such recuperation may take place only in case of high traffic density power sections; where on average there are at least a few vehicles. 38 of 72

39 600 V Rectifier Traction substation busbars Traction substation Feeders H Traction network R Figure 1: Regenerative braking energy flow in case of vehicles on the same power supply section A more frequent way of regenerative braking energy flow is presented in figure 2, in which the braking vehicle and receiving vehicle are on different power supply sections of the same traction substation. Electric energy flows then through the feeders and traction substation busbars. This involves the occurrence of energy losses in the transmission of energy over a considerable distance. 600 V Rectifier Traction substation busbars Traction substation Feeders R H Traction network Figure 2: Regenerative braking energy flow in case of vehicles on different power supply sections of the same traction substation 39 of 72

40 600 V Rectifier Traction substation bus bars Traction substation Feeders R H Traction network Figure 3: Regenerative braking energy flow in case of vehicles on different power supply sections of the same traction substation with energy consumption from rectifier units. This way of regenerative braking is only possible in case of traction networks equipped with standard electromechanical fast track switches. In substations with security systems in the form of semiconductor safety devices it might be impossible due to unidirectional flow of current through such security. Distribution of energy presented in the figure 1 and 2, i.e. the flow of energy in the power supply system is limited solely to the way braking vehicle starting vehicle. Most frequently there is a bigger number of vehicles and, apart from the energy derived from a braking vehicle, energy is also taken from the substations rectifier units, as shown as an example in the figure 3. In numerous cases the situation of energy flow may be even more complicated. It occurs in case of dense traffic of many vehicles. Summing up the above states of power supply system performance during regenerative braking one can determine four possible states of its work: State I: The value of energy put into the traction network is lower than the value of energy consumer by all the vehicles. This corresponds to the situation presented in the figure 3. The busbars voltage is then lower than the voltage of substation s idle gear. State II: The value of regenerative braking energy is equal to the energy needs required by the vehicles, which corresponds to the situations presented in the figure 1 and 2. Then the 40 of 72

41 substation busbar voltage is between the idle rear value and the upper threshold of the work start-up by the regenerative braking resistor chopper. State III: In the area of substations power supply there are not enough numbers of collections for utilizing the braking energy generated during recovery (fig. 1 & 2). Then the work is begun by the braking resistor chopper separating recuperation voltage. The substations busbar voltage is then in the range of modulation of the braking resistance chopper, i.e V. State IV: In the substation traction area there is no reception of regenerative braking energy. In this case the whole energy is dissipated in the braking resistor and the voltage in the traction network is equal to the upper level voltage limit level during braking. State IV State I State II State III Substation busbar voltage 660 V 770 V 790 V Figure 4: Value of substations busbar voltage at various states of power supply system performance during regenerative braking The voltage level of the traction substation busbar can thus be a determinant of the extent of the utilization rate of energy generated during braking at a specific moment and the readiness for receiving recovery energy by the power supply system. A substation busbar voltage should be therefore regarded as an indicator of saturation of the power supply system with recuperation energy. A key issue is also the selection of voltage level of the substation busbar at idle work. As already indicated, a necessary condition for the occurence of energy recuperation is the production by the trolleybus the voltage higher than the traction network voltage. Higher traction substation busbar voltage means the necessity to produce higher voltage by a vehicle during braking, which limit value is reduced to 790 V. This means that too high voltage of traction substation operation can limit the opportunity for electric energy recovery during braking. 41 of 72

42 2.2 Analysis of traction network performance from a stationary point of view application of supercapacitors The aim of supercapacitors is collect recovery braking electric energy when it is not possible to absorb the electricity generated during recuperation by other vehicles moving in the area of the substations power supply. Energy stored in the supercapacitor can be utilized in the moment when there occurs a load on the traction substation. Supercapacitors may be placed in: vehicles traction substations at traction substations points, out of traction substations. Placing supercapacitors in the vehicles involves increased trolley weight. As a result of this a passenger volume declines. The weight of the vehicle tank is at a level of 500 kg, which means capacity reduction by 8 persons, i.e. in the case of the classic trolley length of 12 meters it is a capacity decrease of 10%. Furthermore, such increase in weight causes a rise in electric energy consumption, which reduces the efficiency of energy recovery. Stationary electricity storage tanks are not characterized by these defects. 600 V Rectifier Traction substation busbar Traction substation Supercap Feeders Traction substation H Figure 5: Placing a stationary supercapacitor in the traction substation From a technical point of view, the easiest solution is to place a supercapacitor on the traction substation, as schematically shown in figure 5. A supercap is then connected to the 42 of 72

43 substation busbar. Such a solution does not require costly investments in construction and it is not associated with the problem of limited access to the urban area. In numerous cases the best conditions for recovery occurrence are found only in certain areas of the traction network, which may be located at a considerable distance from the substation. Then it is preferable to locate the storage device at the site of recuperation occurrence, e.g. in the mountainous part of the traction network (fig. 6). An important advantage of such an approach is the reduction of transmission losses in the power supply feeders and the traction network due to a smaller distance between the recuperating vehicles and energy collections. Placing the energy storage also allows for the improvement of voltage stability in the traction network, which may be beneficial in the case of parts of the power supply area distant from the traction substation. 600 V Rectifier Traction substation busbar Traction substation H Traction network Supercap Figure 6: Placing a stationary supercapacitor out of the traction substation Placing the energy storage out of traction substation does not exclude the possibility of its cooperation with all other sections in the substation power supply area or sections galvanically connected to a specific traction network section. However, it involves an increase in transmission losses of energy from the remaining sections of the power supply. Then, energy transfer takes place via the substation busbar, analogically as in the case of vehicle a vehicle recuperation for vehicles located in various sections of power supply. 43 of 72

44 Rectifier 600 V Traction substation busbar Traction network Traction substation Supercap H Figure 7: Electric energy flow between power supply sections at placing a stationary supercapacitor out of the traction substation 2.3 Initial determination of supercapacitor location The main factor determining the applicability of supercapacitors installation are the potential electric energy savings resulting from the installation. In the case of storage installation energy generated during regenerative braking of the trolley can be absorbed in two ways: by other vehicles being in the substation s power supply area (energy recovery by vehicle a vehicle way), by a supercapacitor. From the point of view of energy efficiency energy recovery via vehicle - a vehicle is a better option, which is due to the lack of energy conversion losses. Nevertheless, to make such energy recovery possible, it is necessary for braking and start-up of vehicles to occur simultaneously. The likelihood of such a situation occurrence depends primarily on traffic conditions, and mainly on the traffic intensity of trolleybuses expressed by the average number of vehicles on the traction substation power supply area. The bigger this number, the more likely it is that at the moment of one vehicle braking another vehicle will be in a start-up phase. Thus, the greater traffic intensity the more effective energy recovery on a vehicle-a vehicle way and less justified it is to install supercaps. The second criterion connected with the volume of energy possible to be accumulated in the storage is a potential of generating recuperative braking energy in a 44 of 72

45 specific area. Moving of vehicles is related to the necessity to overcome motion resistance, which can be divided into: basic motion resistance, additional motion resistance. Basic motion resistance is related predominantly to the force of the friction of wheels and aerodynamic resistance. In the final energy balance energy collected to overcome them is converted into heat, and thus, it is not possible to recover it again. Additional motion resistance is associated with the movement of going up the hills, which entails the necessity to overcome the force of gravity. This force, in contrast to the force of friction, is a conservative one, and therefore, it is possible to recover it again. A vehicle going up the hill takes energy to overcome the force of gravity, and when it goes down this energy in turn is converted into kinetic energy. This creates the opportunity to re-convert the kinetic energy into electrical energy and to recover it. Therefore, the best conditions for regenerative braking occur on the routes of a mountainous character. Taking the above into consideration, it can thus be concluded that a potential volume of regenerative braking energy possible to be accumulated in a supercapacitor storage device depends on two factors: traffic intensity expressed by an average number of vehicles in the power supply area, this relationship being of a decreasing value, route profile expressed as a difference in altitude in the substations power supply area, this relationship being of an increasing value. Table 1 presents a summary of traction substations powering the trolleybus network in Gdynia from the point of view of the above criteria. The factors influencing costeffectiveness of applying energy storage devices, i.e. traffic intensity and route profile were specified in the scale of 1 5, with 1 being the lowest value of a given factor and 5 the highest one. The final result, which is determining the profitability of a storage device application, is defined as the ratio of the route profile to traffic intensity. 45 of 72

46 Table 1: Analysis of profitability of storage devices installation on individual traction substations Substation Traffic intensity Route profile Storage device installation profitability Północna 5 1 0,2 Grabówek 4 1 0,25 Dworzec 3 1 0,33 Wendy 3 1 0,33 Kielecka 2 1 0,5 Redłowo Sopot I Sopot II Wielkopolska 2 5 2,5 Chwaszczyńska 4 3 0,75 The conducted analysis shows that the most predestined substation for energy storage devices installation is Wielkopolska substation. 2.4 Estimation of energy savings opportunities The value of substations busbar voltage can be a measure of the potential of regenerative braking energy usage. This parameter was used to assess each individual traction substation operated by Przedsiebiorstwo Komunikacji Trolejbusowej in terms of electrical energy storage devices installation. There were made the measurements of one-second values of substations busbar voltage. Their results are the basis for further analysis. Figures 8 and 9 illustrate a histogram of the busbar voltage on two trolleybus substations: Wielkopolska and Chwaszczyńska, with a division to working days, Saturdays and Sundays. 46 of 72

47 Figure 8: The histogram of Wielkopolska substation busbar voltage Wielkopolska and Północna substations are characterized by a different kind of the traction network load. Północna substation has 6 feeders, and its power supply area is flat. Wielkopolska substation supplies power to one section of a mountainous route profile. Figure 9: The histogram of Północna substation busbar voltage This difference is reflected in the results of substations busbar voltage measurements. In the case of Wielkopolska substation the histogram shows the area corresponding to a frequent occurrence of substation busbar voltage at the level of 770 V V, which involves 47 of 72

48 generation of a large amount of energy created by the recovery, which cannot be consumed by other vehicles and is dissipated in braking resistors. A flat profile and a large size of Północna substation power supply area entails relatively poor conditions for recuperation as well as the possibility of recovery energy consumption on a vehicle a vehicle way, so values of the traction network voltage higher than 750 V are rare. Figure 10 shows a comparison of busbar voltage histograms for all traction substations. The highest level of non-use of regenerative braking energy, reflected by the occurence of busbar voltage values above 750 V, is visible for Sopot and Wielkopolska substations. Figure 10: The histogram of busbar voltage of all substations for a working day The shape of the busbar voltage histogram allows for a qualitative assessment of regenerative braking energy usage level on individual traction substations. For the purpose of quantitative evaluation one may use the value of relative occurrence time of busbar voltage at the level of 770V - 790V, which means the presence of resistor braking. It equals not using the energy generated during braking. A higher value of this ratio indicates a lower level of recuperation energy consumption and hence greater potential energy savings connected with the storage device installation. This coefficient can be defined as: 48 of 72

49 where: tp time of exceeding 770 V voltage on the busbar, tc time of substation operation. Figure 11 presents the values of relative time of exceeding 770 V voltage for individual substations. Figure 11: Relative time of exceeding 770 V voltage Relative time of exceeding 770 V voltage defines relative potential of regenerative braking energy non-use level related to the trolleybus traffic time. This information, however, is not conclusive enough to assess the potential of savings in electrical energy consumption. It should be referred to the total electricity consumption by the substation. For this purpose, one may define the coefficient of relative potential for energy savings: This coefficient can be defined as: where: tp time of exceeding 770 V voltage on the busbar, tc time of substation operation, P average daily substation load. 49 of 72

50 Figure 12: Relative energy savings potential Figures 13 and 14 present the values of coefficients made average for all weekdays. Figure 13: Relative time of exceeding 770 V voltage made average for the week 50 of 72

51 Figure 14: Relative energy savings potential made average for the week 2.5 Conclusions The greatest relative idle time of non-using regenerative braking energy is present in Sopot and Wielkopolska substations, these two substations being predestined for supercapacitor storage devices installation. However, due to a little average load of Sopot substation, potential electrical energy saving is considerably smaller than in Wielkopolska substation, thus a supercap shall be installed in this substation. 3. Analysis of traction network performance: fleet perspective 3.1 Operation of registration and the contents of data collected by the fleet In the last quarter of 2009 Przedsiębiorstwo Komunikacji Trolejbusowej Company purchased two new and modern trolleybuses Solaris Trollino 12M. One special, extremely useful and attractive feature of these vehicles at that time was the ability to run without any contact with the traction network. These trolleybuses differed from the other by asynchronous drives with braking energy recovery to the traction network produced by MEDCOM and full air conditioning. At the time PKT already had almost a ten years positive experience with traction drives effectively recovering braking energy, which were DC drives with adjustable pulse regulators of IEL production. The previous experience with asynchronous drives manufactured by CEGELEC, although they had very good traction parameters, was not optimistic in terms of energy because they rarely undertook regenerative braking despite 51 of 72

52 favourable network conditions, and if they actually recovered energy it was via series recuperation resistor, which obviously considerably deteriorated energy efficiency. It should be noted that at the same time as Solaris Trollino 12M, what made its debut was a bus converted into a trolleybus with a competitive traction equipment by ENIKA, which had similar characteristics. During the preparation of public procurement, the result of which the Solaris Trollino 12M trolleybuses were purchased, having taken into account previous experience and new developments, a decision was made to order also the system to record and upload many vehicle parameters. Along with the trolleybuses there was provided a system by PIXEL, which collects data from the whole trolley i.e. electric measurement of equipment by MEDCOM, the data on the state of the body's performance of Solaris and some measurements from PIXEL devices. This system was also purchased with further Solaris trolleybuses Trollino 12M and has been extended to certain vehicles with ENIKA equipment. In addition, in the version without the data from the drive the system also comprises the trolleybuses Solaris Trollino 12AC and Solaris Trollino 12T. A set of recorded data includes, among others: energy taken by the trolley; energy given by the trolley; energy taken by the trolley drive; energy given by the trolley drive; measurement time; trolley drive current; the whole trolley current; traction network voltage; state of traction battery load; state of linear contactor; heating switch; trolley power supply system (network / traction battery / lack of power supply); 52 of 72

53 power supply voltage KPP i.e. floor installation 24 V; GPS position; indicator switch; opening of each door Data recorded by PIXEL KPP in trolleybuses after driving in to the depot are automatically and without employees engagement transmitted by the radio to PIXEL database, and then via the net to CVS server. The task of the server is to generate csv files where each file contains data from one trolleybus for one day. These files are available in the company's Ethernet network. The data format in the file is standard, which allowed for automation of data processing and analysis by means of a standard spreadsheet. Originally, the system supplier has provided a computer program to analyze the data recorded in the trolley and collected in the base station without the usage of any other devices. This form probably met with the satisfaction of users in bus depots, however, in the case of PKT it appeared to be too limited and slow at such extensive records. In response to a real demand a CSV server has been launched. Currently csv files are analyzed by means of MS Excel 2010, where the folder ANALIZATOR.xlsx was created originally serving only as an overlay on the csv files with a small amount of calculation. During the use the folder has repeatedly been modified by adding new useful functions and modifying the already existing ones in order to adapt it to current needs. For example, for the purposes of the study hereby there were added formulas processing the data broken down to power supply areas by individual traction substations with the data prepared for copying into another folder. Due to a large size of the ANALIZATOR.xlsx file containing over 65,000 lines, about 100 columns, often quite complicated formulas in the cells and a few diagrams with a great number of input data, processing of the data from one trolleybus takes a few minutes depending on the computer. One should also mention time-consuming opening of MS EXCEL files. However, this method of data processing has very important advantages: shortening the access to the data in comparison with the previous one; 53 of 72

54 full operation flexibility by relative ease of modification to the current needs of the users; possibility to create several spreadsheets suited to individual user s needs; easiness to copy data. 3.2 Analysis of data collected by the fleet and pre-selection of areas with greatest potential for storing braking energy in a supercapacitor In order to make a pre-selection of the areas with the greatest opportunities for braking energy storage in the energy storage device, the data collected from the trolleys were divided into those registered during individual traction substations power supply. Separated data were processed through several stages, which produced daily values for each trolley needed for further data processing in all the trolleys simultaneously. Taking into account the possibilities connected with the type of recorded data and their relevance to the pre-selection of the areas of greatest potential for braking energy storage in the supercap, it has been assumed that the key evaluation criteria will be the following: taken in energies EPN and Eon energies given away by the drives in areas supplied with power by each substation. In order to avoid undeliberate copying of any potential error in the calculation of taken and given energy and the error resulting from rounding the values of the energy between individual records, which most often differ by a change arising in time 1s, energies were calculated individually by numerical integration of the product of drive and voltage current at the time. In the daily calculation this method gave consistent, though slightly different, values with those recorded by the apparatus in the trolleys, which is a positive verification of the correctness of the calculation method. In addition, there were introduced formulas representing filters of steady state errors of drive current measurement transducers. Tnr[h] time in which there is no possibility of full recovery of trolleys braking energy to the traction network in areas powered by each substation. This time is calculated when the voltage in the traction network exceeds 770 V. This parameter is only an indicator, and it is difficult to apply to other calculations. In can only be talked here about a strong correlation between this time and the fact of braking energy excess dissipation in the trolley resistors. Surely the same moments when it 54 of 72

55 Weekdays was not possible to return power to the overhead line braking were recorded by various trolleys that were at the same time in the areas supplied with power by the same traction substations. Registration of this time by a given trolley does not mean that in this very trolleybus there is just taking place electrodynamic braking with partial or total power dissipation in the braking resistor. In only means that it is happening in one of the trolleys powered at the same time from the same traction substation. Ts[h] time of supplying power to trolleybuses in areas powered by particular substations. This value was recorded for the reference to the time in which there is no possibility of full recovery of braking energy to the traction network. S[km] mileage of trolleybuses in areas powered by particular substations. This value was recorded for one s own view and support. The tables and graphs below show the results of calculations made for the pre-selection of the areas with the greatest potential for braking energy storage in the supercap. Table Summary of the calculations results in order to pre-select the power supply area. Substation Ts[h] Tnr[h] Enp [kwh] Eno [kwh] S[km] Tnr/Ts Eno/S [kwh/km] Eno/S [kwh/km] Eno/Enp Północna 175,76 0, ,9 1243,8 2069,9 0, , , ,2786 Grabówek 256,91 0, ,4 1403,9 2652,3 0, , , ,9937 Kielecka 115,78 0, ,1 750, , , , ,67 Plac Konstytucji 110,78 0, ,9 580, , , , ,8509 Wendy 133,54 0, ,9 1106, , , , ,7039 Redłowo 137,89 1, ,2 1865,1 2937,5 0, , , ,5491 Wielkopolska 34,331 2, ,4 399,66 924,49 6, , , ,6392 Chwaszczyńska 231,73 0, ,2 1798,6 3207,7 0, , , ,8886 Sopot 1 12,441 1, ,72 72, ,9 11,0211 1, , ,7342 Sopot 2 18,236 0, ,33 26,951 82,349 2,8272 1, , ,5207 Total 1227,4 8, , , , , , of 72

56 Saturdays Sundays and holidays No division 1227,7 8, , , , , ,7961 Północna 106,12 0, , ,8 0, , , ,7714 Grabówek 191,64 0, ,6 846, ,1 0, , , ,9753 Kielecka 74,75 1, ,8 486,14 830,19 1,5745 1, , ,4131 Plac Konstytucji 69,984 1, ,49 284,75 597,16 1, , , ,1715 Wendy 71,408 1, ,2 671,2 1037,5 1, , , ,6492 Redłowo 90,072 1, ,3 1128,6 1876,2 1, , , ,8537 Wielkopolska 20,245 1, ,01 233,92 578,44 9, , , ,9503 Chwaszczyńska 141,55 0, ,6 1187, , , , ,6138 Sopot 1 10,383 0, ,52 64, ,05 8, , , ,913 Sopot 2 16,748 0,37 102,29 20,988 67,57 2, , , ,5175 Total 792,89 9, , , , , ,8484 No division 721,13 9, , , , , ,8726 Północna 130,84 0, ,1 989, ,9 0, , , ,0801 Grabówek 125,23 1, ,3 955, ,5 1, , , ,9182 Kielecka 72,506 2, ,7 349,7 660,54 2, , , ,7429 Plac Konstytucji 47,373 1, ,89 218,94 474,02 2, , , ,1959 Wendy 55,193 1, ,3 549,34 940,54 2,761 1, , ,06 Redłowo 60,204 2, , ,2 3, , , ,2011 Wielkopolska 12,443 1, ,66 118,64 340,82 10,202 1, , ,2077 Chwaszczyńska 78,799 1, , ,6 1, , , ,8323 Sopot 1 6,7283 0, ,65 22, ,37 11,5143 1, , ,42316 Sopot 2 12,127 0, ,629 10,277 44,232 2, , , ,4244 Total 601,45 12, ,4 8803,7 2, , , ,8828 No division 577,65 12, ,9 8813,3 2, , , , of 72

57 57 of 72

58 58 of 72

59 59 of 72

60 60 of 72

61 The measurements analysis results both presented by diagrams and tables indicate that the largest braking energy storage in the supercap is possible in areas supplied with power from Wielkopolska and Sopot 1 traction substations. It is best illustrated by the graphs , and but on the other , , , and is also visible. The graph shows that at the Sunday schedule and traffic level in the areas fed from Wielkopolska and Sopot 1 substations the advantage of braking energy storage options in the supercaps on the substation is not so clear in reference to the other. It should be noted, however, that the data from the weekdays are most crucial. To sum up, Wielkopolska and Sopot 1, which are supplied with power from substations, are the result of the pre-selection of the areas with the greatest opportunities for braking energy storage in a supecap. Taking into consideration, among other things, Tnr times it can be concluded that it is best to install the braking energy storage device on Wielkopolska traction substation in the first place and in Sopot 1 in the second place. One should also take into account the progressive replacement of the fleet operated by PKT, which may cause that in the future it will be worth installing energy storage devices on more traction substations or the points in the network. 61 of 72

62 3.3. Analysis of fleet data and pre-selection of sections with the highest voltage changes and needs for conceivable point support for the traction network power supply from the supercapacitor Electric vehicles powered from the traction network, similarly to other electrical energy receivers tolerate voltage supply with values falling within a well-defined range. The permissible range of voltage supply to traction vehicles is defined by a European standard EN The standard implies that in Gdynia trolleybus traction network with rated voltage of 600 V DC the value of voltage Us can be: 400 V Us 720 V which is continuous thus without a time limit. The value of 720 V according to the standard is at the same time the maximum permissible voltage of the traction network working in an idle state when all the trolleybuses are separated from it by descending the collectors. One can also refer to an idle state to the time when the trolleys are connected to the network yet they do withdraw any current, eg. at standstill with a static inverter and other auxiliary systems turned off, but the standard does not define voltage in this state. Trolleys often have EMC filters installed, which contain considerable capacities that can cause an increase in the average voltage value; 720 V < Us 800 V with a time limit to 5 minutes. These values are most frequently achieved as a result of a trolley regenerative braking; 800 V < Us 1016 V permissible non-recurring clasp with duration time of 20 ms tp 1 s dependent on the voltage Us. The relationship between the duration time and the voltage value in this respect is defined by the ratio: Us(t) [V] = 1016 V * t^(-0,0611) Voltage in the DC traction also has a variable component resulting from the performance of rectifier units, static inverters, and power electronics drives. This is often a superposition of deformation variables from different devices, which is clearly seen in the oscillograms. Variability of the component variable nature of makes its consistent determination difficult; however, one must take into account its amplitude of reaching sometimes up to 20% of rated voltage of the overhead 62 of 72

63 line. RMS, average and peak values for pure DC voltage are the same, but in the case of a component variable content they can vary greatly. Unfortunately, the standard does not define clearly what values of voltage are concerned - peak, average or RMS, which is often the cause of misunderstandings and the compatibility behaviour of different types of fleet requires a firm stance of the carrier towards power electronics drives and static converters suppliers, restricting the requirements set by the standard. In the subchapter hereby one should consider the issue of changes in voltage in the traction network brought about by voltage drops caused by the flow of currents through the resistance of wires, leading network and power supply sources i.e. traction substations. Depending on the direction of the current flow voltage drops on the abovementioned resistances may cause point drops (in the sense of a decline) or increases in the traction network voltage. The point voltage increases in the traction network can be caused by regenerative braking of a trolley at one point with its reception at another, most often a distant point or points of the network. These points must, of course, be galvanically connected to each other. According to Ohm's law, a regenerative braking current, i.e. with a minus sign, causes a minus voltage drop on resistances and a rise in voltage on the trolleybus collectors. In extreme cases this can result in dissipation of a part of braking energy in the trolleybus resistor in order to avoid an excessive increase in the voltage at collectors, despite the fact that in the same area supplied with power by the same traction substation there are collections that are able to take advantage of the full braking power (trolleybuses and / or energy storage device). This situation is a double disadvantage since it leads to a necessity disperse a part of the braking energy, and additionally there are power losses due to the current flow through the network resistance. The above-described situation is distinct from the one when a trolley braking in a regenerative way (it feeds only itself) increases the voltage in the traction network without returning the current due to the absence of any other vehicle and / or energy storage device in the power supply area. Then we can talk about an increase in the voltage across the whole power supply area by a particular traction substation. 63 of 72

64 Provided that the voltage on traction substations busbars usually exceeds 650 V, the point voltage drops (in terms of a decline) can sometimes even cause a disruption of the drives and trolleybus auxiliary systems performance. Moreover, their presence is much detrimental to the overall energy efficiency of the trolleybus traction system. Similarly as in the case of increases, the voltage drops (in terms of a decline), according to Ohm's law, the current collected by a trolley (positive; with a plus sign), causes a voltage drop on resistances and thus a voltage drop (in terms of a decline) on the collectors. Regenerative braking of another trolley that is nearby in the same sector of the traction network, may partly or fully offset and even produce the opposite effect to the voltage drop i.e. a rise. This is due to the resultant circuit formed at a particular time by a combination of current, voltage, and resistance sources, all in a specific and often unique configuration. This situation naturally improves the energy efficiency of the system but it cannot be an effective and reliable means of avoiding trolleybuses operation interference. Since both point voltage drops (in terms of a decline) and voltage increases arise from Ohm's law, in order to determine the traction network sections with the largest changes in voltage an analysis was conducted of only point voltage drops (in terms of a decline). The biggest changes in voltage might be expected at points with the highest replacement internal resistance (Thevenin), which is best to find to seeking the situation when the voltage drops below a certain value. So as to avoid registration eg. section insulators or a total voltage loss, an indication of the point with known coordinates took place when both the single sample was in the range of 250 V <U <500 V, and the average of five (examined, the two previous and the two following ones) reached a value of below 500 V. The measurement samples were recorded when a change occurred, but not less frequently than every 1s which in practice usually was at every 1s. The diagram contains plotted points that are and /or indicate geographical positions: the trolleys travel routes as a merged network (designated data series: the route, black colour); the trolleys loops. Shown as arrows (separate series for each point); boundary points of power supply areas connected with a broken line (blue line); 64 of 72

65 indications of the points recorded by various trolleys and meeting all the established and above-described criteria (bold purple lines highlighting the points and / or sections, going to a one central point). The diagram shows that the sections with the biggest changes in voltage are: practically the whole network in Pustki Cisowskie distric area; Morska street on both sides of a boundary point Grabówek/Plac Konstytucji; street on both sides of a boundary point Kielecka/Plac Konstytucji; Wielkopolska street at the end of the power supply area from Redłowo substation at a boundary point Redłowo/Wielkopolska; Chwaszczyńska street at the end of the power supply area from Chwaszczyńska substation at a boundary point Chwaszczyńska/Wielkopolska; several other points (streets: Chwaszczyńska, Kacze Buki, Świętojańska, Morska in the area of the power supply from Północna substation). 65 of 72