Electric Power Systems Research 81 (2011) 21272138

Contents lists available at ScienceDirect

Electric Power Systems Research

journal homepage: www.elsevier.com/locate/epsr

Energy management in metro-transit systems: An innovative proposal toward an integrated and sustainable urban mobility system including plug-in electric vehicles
Maria Carmen Falvo a , Regina Lamedica a, , Richard Bartoni b , Gennaro Maranzano c
a

DIAEE Electrical Engineering, University of Rome Sapienza, via delle Sette Sale 12b, 00199 Rome, Italy Maire Tecnimont, Italy c Atac Roma S.p.A. (Rome Urban Mobility Company), Italy
b

a r t i c l e

i n f o

a b s t r a c t
The energy consumptions growth, the upward interest for environmental sustainability and the technological evolution carry to the necessity to review the planning criteria of urban mobility systems in large cities and metropolitan areas. With this aim, new studies and projects are in progress, especially dealing with the power systems for metro-transit lines and surface electric vehicles. In this framework, the authors perform a study that, from an overview about the main energy management issues connected to the city transport, provides an innovative proposal for the design of sustainable urban mobility system: the integration of the metro-lines with surface plug-in electric vehicles. The present paper includes the energy analysis results, obtained by an application on a real case study of an home-made simulation software, describes the proposal in terms of power systems architecture and business models, pointing out the potential advantages that its implementation could give in terms of energy saving, environmental sustainability and reduced economic impact, as a result of the maximum exploitation of existing electric power plant. 2011 Elsevier B.V. All rights reserved.

Article history: Received 10 March 2011 Received in revised form 7 June 2011 Accepted 4 August 2011 Available online 9 September 2011 Keywords: Plug-in electric vehicles Energy saving Environmental sustainability Metro-transit system Regenerative braking Urban transport

1. Introduction Nowadays the energy and environmental sustainability in the mobility systems design is one of the most important objectives. The achievement of this requirement is a nearly necessary condition for guaranteeing also the economic efciency of the proposed solutions [14]. In fact it is demonstrated that an energy-efcient and environmental sustainable transport system is essential to the high-quality performance of modern economic systems: the gross domestic product (GDP) and the transport activity of a Country are closely coupled. As example Fig. 1 shows such coupling in the case of the European Union (EU) and its future trend under a businessas-usual scenario. In other words, transport demand will increase as a result of economic growth, but the latter is only possible if

Abbreviations: AC, alternative current; BEV, battery electric vehicle; DC, direct current; DSO, distribution system operator; ESS, electric sub station; EU, European Union; EV, electric vehicle; GDP, gross domestic product; GHG, green house gas; G2V, grid to vehicle; IEMS, Intelligent Energy Management System; LV, low voltage; MV, medium voltage; PHEV, plug-in hybrid electric vehicle; TS, trafc scenario; V2G, vehicle to grid. Corresponding author. Tel.: +39 06 44 585 808; fax: +39 06 48 83 235. E-mail addresses: mariacarmen.falvo@uniroma1.it (M.C. Falvo), regina.lamedica@uniroma1.it (R. Lamedica). 0378-7796/$ see front matter 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.epsr.2011.08.004

transport activity is able to develop consequently. On the other hand, it is well known that the transport sector in EU is responsible for: approximately 30% of total energy consumptions, 27% of total greenhouse gas (GHG) emissions, 72% of which are attributable to road transport alone. Figs. 2 and 3 respectively show up these two aspects. Recent reports by Intergovernmental Panel on Climate Change point at the need to reduce global GHG emissions by 60% at least by 2050. In this context developed countries will have to cut emissions even further (around 80%), considering that developing countries would nd harder to combine strong emission reductions with fast economic growth. Nowadays a reduction of GHG emissions only from transport around 80% is not compatible with the use of oil-derived fuels and internal combustion engines, on which transport is almost entirely dependent. Moreover, oil price volatility and its potentially insecure supplies have serious effects on global economies and the picture is likely to get worse in the future: global oil production expected to peak and start declining in the near future, while demand continues to grow steadily at the same time. In this worldwide context, the introduction of more energy-efcient and environmental sustainable alternative transport solutions becomes essential. For these reasons the automotive sector alone is spending 20 billion D per year on research and technological development in Europe. At the same time the EU

2128

M.C. Falvo et al. / Electric Power Systems Research 81 (2011) 21272138

- the metro-transit transport system, that is the most energyimpacting, considering the large quantity of power in use; - the on-road public and private vehicles park for the surface transport, that is the most environmental-impacting, considering the large number of units in use. About the metro-transit systems, the result of this renewed interest entailed studies and projects, dealing with proposals on upgrading actions including new technologies, and on management techniques, derived from new simulation models and software [513]. In this framework, in the last years, the authors implemented a multi-stage program for performing high-quality energy studies on metro-transit systems [1423]. Recently a reviewed and updated version of the software has been realized in order to achieve other evaluations, principally regarding the assessment of the impact of the braking energy recovering on the power consumptions. This analysis has also allowed revaluating the choices on the power system design for the management of the braking energy, according to the actual available technologic solutions [24,25]. About the surface urban transport, the strong request of reducing the environmental impact made necessary the introduction of zero-emission vehicles and in particular of electric vehicles (EV) of different types. The authors performed an analysis on the penetration of this type of transport in urban context and on the barriers for its spread, mainly linked to the electric infrastructure for their recharge. Different technological solutions and business models have been got in literature [2633]. At the end of the analysis on the two transport systems, matching the results obtained by the simulation studies and the actual state of the art on the involved technologies, the authors suggest an innovative proposal for clearing the energy problems connected to them, that would allow energy saving with a reduced economic impact, as a result of the maximum exploitation of the already existing electric power systems. The present paper is organized in 7 sections. In particular: - Section 2 reports the overview on metro-transit energy management techniques for saving, proposed in literature and real implemented in worldwide metro-lines, in particular dealing with the braking energy recovering; - Section 3 shows the main features of a simulation software, made by the authors, with the specic task of evaluating energy questions about metro-transit systems; - Section 4 reports the software application on a real metro-line, giving the details of the case study and the main simulation results, in order to get a real assessment of the energy saving associable to the braking energy recovering;

Fig. 1. Trend of the transport activity in EU in function of the GDP growth.

Fig. 2. European energy consumptions for sectors [MTEP].

is supporting under the 7th Framework Program all the research ideas proposing greener, safer and smarter transport systems with a research budget of over 4 billion D in 7 years. Focusing on the urban mobility problems in large cities and metropolitan areas, this general soaring interest for energy saving and environmental sustainability, with the increasing of the consumptions and the technological evolution, have renewed in the last years more than ever, the attention of the researchers at issues about:

(a)
Services, etc. 6.3%

Other 0.6% Energy Industries 34.3%

(b)
Railways 0.6%

Other Transport 0.8% Road Transport 71.9%

Navigation Households 10.5% Industry 20.9% Transport 27.4% Civil Aviation 12.0% 14.7%

Fig. 3. (a) Total GHG emissions by sector in the EU in 2005. (b) Transport-related GHG emissions in the EU in 2005.

M.C. Falvo et al. / Electric Power Systems Research 81 (2011) 21272138

2129

- Section 5 includes an overview on the actual state of the art on electric vehicles and the feasible business models for the management of their recharge infrastructure; - Section 6 describes the innovative proposal for the energy management in metro-transit, taking into account the use of EV, as consequence of matching the results obtained from the analysis performed on the single transport systems; - Section 7 reports the conclusions. 2. An overview on metro-transit braking energy management techniques Since 1990s, several studies have been proposed dealing with the maximization of the energy efciency in urban metro transit systems. A literature references review on this subject, extended on the last fteen years, has pointed out the following main approaches to energy efciency optimization and to the environmental impact reduction in these systems: - optimization in the design and sizing of the electric supplying system (electric sub-station layout, traction line conguration, etc.); - trafc control both at the individual train level (e.g. train performance control in terms of speed and acceleration proles) and at eet level (timetables, coordination of individual cinematic proles). The former point mainly affects traction drives, whereas the latter is software-heavy, implying signaling and SCADA-related issues [518].In this context many simulation studies and experimental surveys have been performed focusing on already existing lines, where only upgrading actions and management techniques can improve their energy performance. Among the possible measures, many proposals regard the way to manage the trains energy yielded by regenerative braking, that can be a substantial part of the overall energy by the supplying network, especially for the lines with consistent slopes and curves. Depending on the electric supplying system layout and on the train drives, the recovered energy can be used in three different ways: 1. feeding other trains running on the same line; 2. storing in onboard or stationary storage system; 3. feeding back to the primary supply network, if bidirectional conversion stations are available. Taking into account the existing metro-transit systems, the lack of bidirectional electric sub-stations (with exception of lines with relevant slopes) and of onboard storage (that implies anyway problems of space and weight) rules out the last solution and one possibility of the second solution, for technicaleconomical aspects regarding the project of electric infrastructure and trains. So the only practicable options are thus represented by the real-time use of the recovered energy for supplying other trains on the same line and/or the use of stationary storage system. The rst option aroused the interest of many researchers; studies have shown that the effective ability of the network to receive the recovered energy cannot be taken for granted, not least because of the signicant randomness of relevant parameters (i.e. train and eet dynamics). As a consequence, a substantial share of the braking energy has to be dissipated in heat by means of on-board rheostats. In this way the rst consequence is the high reduction of the energy system efciency, but it is worth to point out also the problem of heat production in an underground environment, such as the metro transit system one: with the aim of guaranteeing a good level of comfort, especially to metro-workers and then to passengers, it

involves an over-sizing of the tunnel and station fan plants with a consequent further growth in power consumptions. Therefore the introduction of stationary storages in metro-transit for the braking energy management, is becoming roughly a necessity both for energy saving and for comfort issues. In favor of this proposal, there is also the technological evolution of the batteries that nowadays guarantee, in comparison with the rst generation of batteries of 1990s: higher values of specic energy (1001000 Wh/kg); higher values of specic power (1000 kW/kg); longer life cycle; reduced environmental impact and cost; better dynamic performances. In addition, stationary storage insertion could provide further benets, such as: to increase in the system security, being an external source in emergency that could supply trains, for reaching the nearest station, in case of failure of the main power supply; to shave the power peaks with consequent savings in investment costs on the power system; to get better voltage prole on power line, with a consequence decreasing in losses and an improvement in terms of power quality. Then, the recovering of trains braking energy in stationary storages would allow a global energy saving, as better as it is a good total consumptions percentage, with an additional improvement in case of optimization of the same storage systems design. Obviously this result depends on the main characteristic of the metro-line: power system layout, metro-line path, type of train, trafc scenario, etc. In order to have an assessment of the real energy saving guaranteed by the train braking recovering, it is necessary to perform specic analysis with dedicated software on the single cases. With this aim the authors have implemented a specialized multi-stage program and applied it at a real case study.

3. Simulations software for energy analysis on metro-transit systems A multi-stage program has been implemented in Visual Basic and Fortran language. The software is called TRAINSIM and includes: - an electro-mechanical simulator for the evaluation of the metrotrains power consumption for dened trafc scenarios; - an electric and thermal simulator to carry out all the variables necessary to evaluate the energy state of the traction system. The software includes deterministicprobabilistic models necessary for performing high-quality studies on the energy point of view, as literature species [1418]. An earlier version of the software [1922] has been already used for some planning studies on transport systems of Rome, in order to evaluate if their power systems were able or not to guarantee the right operation, with dened quality and continuity standards, in different trafc congurations. In particular the application regarded two studies on: 1. the residual carrying capacity of Line A in Rome metro-transit system in consequence of the introduction of new trains, characterized by a higher nominal power ratings and by a higher transport frequency [22]; 2. the upgrading actions necessary for the railway line, which joins Rome to Ostia, in consequence of the high development of this area: a great growth of passengers number made necessary to change the line service from railway to metro-transit [23].

2130

M.C. Falvo et al. / Electric Power Systems Research 81 (2011) 21272138

Fig. 4. Graphical input interface of the software.

The results obtained by the rst study allowed dening upgrading actions on the traction line and electric sub-stations (ESS) of the metro-transit line necessary to guarantee the right operation at full trafc frequency. In fact the new trains employment involved an increasing of the power consumptions and so the necessity of upgrading the power systems in ESS and traction line size, under the inevitable environmental and mechanical constraints of the existing system. Recently a reviewed and updated version of the software has been realized in order to perform more highquality energy analyses to the same metro-line, referring to the post upgrading asset of the traction system [24]. The main upgrading of the software regards the realization of a graphical interface, shown in Fig. 4, with which it is possible to set many parameters such as: trains departure time on the two ways; specic trains sequence; trains stop time period in each station; trains shift time on the two ways. A ner capability of the software has been got also with an upgrading regarding some output options. In particular the new version lets to calculate and to draw: - for each train, referring to a simulation time step: the recoverable and the recovered braking power referred to the pantograph voltage; - for the global system, referring to a simulation time step: the total required power by the trains; the total supplied power by the ESS; the total potential recoverable and effective recovered braking power. The calculation of these parameters is made by two off-line tools, whose graphical interface is shown in Figs. 5 and 6. Therefore, by means of all the input/output options, it is possible to analyze

different trafc congurations, and to get specic information on the energy performance of the system. 4. An application to a real case: Rome metro-Line A An application of the software, described in Section 3, has been performed for the assessment of the impact of the braking energy recovering on the energy consumption in Rome metro-transit system. The analysis has allowed to get a comparison with a similar experimental study performed on the same line, many years ago and reported in [20], and so to revaluate the choices on the design of the power system for the management of the braking energy, according to the actual available technologic solutions. 4.1. Metro-line description Rome metro-transit system includes two lines: Line A and Line B. Line A, object of the study, is on two ways, 18.76 km long, with 27 stations and a layout including many slopes and curves [24]. The power system is composed by 6 electric sub-stations (ESS), as shown in Fig. 7. Each ESS includes conversion groups, whose voltage output is 1.5 kV DC for the traction line supplying, and equipped with AC/DC full diode bridge converters with a nominal power ratings of 3.5 MW. By 2002 an expansion project of the metro-line had scheduled the change of the trains eet, with a progressive introduction of new vehicles with superior energy consumption. In particular the new vehicles are MA300 trains, made by CAF Espania and including Bombardier Drivers. MA300 allows braking energy saving, but it has higher energy consumption, due to the superior nominal power

M.C. Falvo et al. / Electric Power Systems Research 81 (2011) 21272138

2131

Fig. 5. Graphical input/output interface of the tool for system power parameters calculation.

ratings of drives and to the presence of onboard conditioning plants. The main features of MA300 trains are described by the traction curve and the braking curve, reported in Figs. 8 and 9. MA300 trains braking curve (the upper curve in Fig. 9) is made of two components: the rst one (the intermediate curve in Fig. 9) dening the electric braking effort, that decreases with the speed and it is null under a speed of 10 km/h; the second one (the lower curve in Fig. 9) dening the pneumatic braking effort. It is worth to stress that the total braking effort is constant with the speed for every value of train mass, so that it is possible to have a constant deceleration in m/s2 . The balance between the two components is regulated by an onboard computer, whereas the driver can operate only on the minimum speed (even superior to 10 km/h) under which the electric braking effort is null. This powering of traction loads implied the need to investigate upgrading actions on the traction line and ESS of the metro-line. A rst study [22], performed by some of the authors, gave as results some upgrading actions to guarantee the right operation at full trafc scenario, such as the addition of new transformation/conversion groups in some ESS, and the upgrading of the traction line section, by the introduction of feeders on some stretches of the line.
Table 1 ESS main features. ESS Battistini Cipro Flaminio V. Emanuele Travertino Cinecitt Progressive (km) 1.37 5.16 9.26 12.46 15.06 18.36

So the nal asset of the traction system is that one summarized in Table 1. About the traction line it is composed by two 100 mm2 contact wires, two 120 mm2 carrying cables, and one 120 mm2 feeder. The upgrading actions on the traction line, obtained by the abovesaid study [22], included the addition of 2 150 mm2 feeders from Cinecitt ESS to Flaminio ESS. 4.2. Simulations hypotheses: electric-thermal parameters and trafc scenarios setting Referring to the nal asset of the metro-line of Rome, some simulations have been carried out for the system energy performance evaluation, considering the full trafc scenario (all MA300 trains running). A wide analysis has been performed considering different trains frequencies and different values of the minimum speed under which the train electric braking effort is null. At the basis of all the simulations there are some hypotheses about electric and thermal parameters values. The values considered for electric parameters are summarized in Table 2, whereas, referring to the model for the traction line in [8], the thermal parameters values used are reported in Table 3. The simulations are referred to trafc scenarios including the

Number of transformation/conversion groups 2 + 2 (+1 as reserve) 2 + 2 (+1 as reserve) 3 + 3 (+1 as reserve) 2 + 3 (+1 as reserve) 2 + 3 (+1 as reserve) 5 + 3 (+1 as reserve)

Nominal power ratings (MVA) 3.86 3.86 4.00 4.00 4.00 4.00

Voltage input and output (kV) 20/0.59/0.59 20/0.59/0.59 20/0.64/0.64 20/0.64/0.64 20/0.64/0.64 20/0.64/0.64

2132

M.C. Falvo et al. / Electric Power Systems Research 81 (2011) 21272138

Fig. 6. Graphical input/output interface of the tool for trains power parameters calculation.

Fig. 7. Layout of the metro-line power system. Table 2 Main gures for electrical simulation. Thevenin no-load voltage for Battistini and Cipro ESS transformation/conversion groups Equivalent resistance for Battistini and Cipro ESS transformation/conversion groups Thevenin no-load voltage for Flaminio, Vittorio, Travertino and Cinecitt ESS transformation/conversion groups Equivalent resistance for Flaminio, Vittorio, Travertino and Cinecitt ESS transformation/conversion groups Minimum voltage that allows to train drives to work with a constant power Rail resistance Traction line and rail total resistance Electric simulation software time step 1.578 V 0.48 1.660 V 0.069 1.050 V 0.017 /km for one way 0.0537 /km for one way 5s

M.C. Falvo et al. / Electric Power Systems Research 81 (2011) 21272138 Table 3 Main gures for thermal simulation. Time step for thermal simulation Environmental temperature Initial traction line temperature Radiation emissivity Solar radiation Electrical resistance temperature coefcient Radiation coefcient Material density Air speed Specic heat capacity 5s 25 C 25 C 0.6 0 W/m2 0.0042 0.95 8890 kg/m3 0.5 m/s 385.3 J/kg C

2133

50,00 45,00 40,00 35,00 30,00 Total required energy by the trains Total supplied energy by ESS Total potenal recoverable braking energy Total eecve recovered braking energy Energy losses

[MW]

25,00 20,00 15,00 10,00 5,00 0,00 120" 150" 180" 240" 300" 360" 600"

Trains frequency [s]

Fig. 10. System energy values in MW for 1 h simulation for different TS in CASE-10.

the required power by the trains; the supplied power by the ESS; the effective recovered power by the trains braking; the potential recoverable power by the trains braking.

Fig. 8. MA300 traction curve.

presence of only MA300 trains. A constant stop time period in station, equal to 20 s, has been chosen as medium value. Simulations are then performed referring to different values of trains frequency, for each one is dened a trafc scenario (TS): 120 s (TSA), 150 s (TS-B), 180 s (TS-C), 240 s (TS-D), 300 s (TS-E), 360 s and 600 s (TS-F). Besides six trains are hypnotized stopped on turnout tracks. For each TS, the minimum speed under which the electric braking effort is null, has been changed in order to have an evaluation of the impact that the driver guide style has on the energy consumption. In particular, after a sensitivity analysis, it has been possible to point out that a minimum variation of 10 km/h is necessary to get a difference in power consumptions. Therefore for each TS, two values of minimum speed have been considered: - 10 km/h (simulation named TS-X and CASE-10); - 20 km/h (simulation named TS-X and CASE-20).

By the software results, it is possible to calculate: the total required energy by the trains; the total supplied energy by the ESS; the total potential recoverable and effective recovered braking energy. So some important percentages can be evaluated, such as: the energy saving percentage (ES% ), dened as: ES% =
W REC EESS

EESS

W/O REC

100

(1)

W REC is the supplied energy by ESS in case of recovering where EESS

of the trains braking energy; EESS is the supplied energy by ESS without the recovering of the trains braking energy; the effective recovered braking energy percentage (ER% ) in respect of the recoverable braking energy, dened as: ER% = ETR REC, ED 100 ETR REC, BLE (2)

W/O REC

4.3. Simulations results on the whole system For each TS-X and CASE-YY, the system output variables, for each time step in the total simulation time, are:
350 325 300 275 250 225 200 175 150 125 100 75 50 25 0 0 10 20 30

where ETR REC, ED is the effective recovered braking energy; ETR REC, BLE is the potential recoverable braking energy; the effective recovered braking energy percentage (ER% ) in respect of the total required energy by the trains, dened as: ER% = ETR REC, ED 100 ETR REQ (3)

Electrical Braking Effort Pneumatic Braking Effort Total Braking Effort

40

50

60

70

80

Speed [km/h]
Fig. 9. MA300 braking curve.

where ETR REC, ED is the effective recovered braking energy; ETR REQ is the requested energy by the trains. ES% values give an assessment of the impact of the braking energy recovering on the consumption; ER% values give an assessment of the capacity of the line to receive the braking energy by the train; ER% values give an assessment of the contribution of the recovered braking energy at the traction load. In order to summarize the simulation results, some curves and histograms have been drawn. In Figs. 10 and 11 a comparison between some system energy values, in the simulation time of 3600 s, is reported referring to different TS-X of the CASE-10. From those gures, it is possible to point out that the total required energy by the trains decreases for an increasing of trains frequency. Besides, in 3600 s simulation time, for an increasing of 30 of the train frequency:

Braking Effort [kN]

2134

M.C. Falvo et al. / Electric Power Systems Research 81 (2011) 21272138

98% 92% 89% 90% 78% 77%

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

95%

ary systems. For a rst study, to know the daily value of this energy it is possible to consider a real diagram of the trains frequency in 24 h service that includes: 6 h without service and 18 h with service; 4 h of service with a trains frequency at 150 ; 4 h of service with a trains frequency at 360 ; 10 h of service with a trains frequency at 240 .

38% 34%

35%37%

33%35%

31%33%

36% 33%

28%30%

30% 27%

120"

150"

180"

240"

300"

360"

600"

Trains frequency [s]

Perceptual eecve recovered braking energy (ER%) in respect of the recoverable braking energy Perceptual eecve recovered braking energy (ER%) in respect of the total required energy by the trains Perceptual energy saving (ES%)

The daily value of energy dissipated in heat becomes 14.5 MWh. Therefore a stationary storage would manage large quantity of energy that could be stored in batteries located in each ESS. For a given trains frequency and for a 10 km/h increasing of the minimum speed under which the electric braking effort is null, in 3600 s simulation time: - the total supplied energy by the ESS increases of 7%; - the potential recoverable braking energy decreases of 12%; - the effective recovered braking energy decreases of 10%. As consequences: - ES% decreases about 5%; - ER% decreases about 2%; - ER% decreases about 4%. In other words, it is better to exploit the braking energy available also at low speeds. 4.4. Simulations results on the single train For each TS-X and CASE-YY, the train energy output variables, for each time step in the total simulation time, are: - the required power by the single train; - the potential recoverable and the effective recovered braking power by the single train; - the pantograph voltage of the single train. By the software results, it is possible to calculate: the total required energy by the single train; the effective recoverable and the potential recovered braking energy by the single train. As example, the diagram of potential recoverable and effective recovered braking power and of the pantograph voltage, for each time step in the total simulation time, is reported in Fig. 13, for the TS-F (600 trains frequency) CASE-10 that is the worst. In Fig. 13 it is also shown that, when the pantograph voltage is higher of the limit (1.7 kV), the effective recovered braking power

Fig. 11. System energy percentage values in (1), (2), (3) for different TS in CASE-10.

- the total energy supplied by the ESS decreases of 8%; - the potential recoverable braking energy decreases of 4%; - the effective recovered braking energy decreases of 4%. As consequences, for a total trains frequency variation from 120 to 600 : - ES% decreases from 38% to 30%; - ER% decreases from 95% to 77%; - ER% decreases from 34% to 27%. In other words, when the trains number is little, the braking energy potential available to be recovered is small, the line capacity to receive this energy is reduced, the energy saving is reduced. Besides, comparing the value of ES% with that one obtained on the same line 12 years ago [20,21], all the energy values are atly higher (in present system about 110 MWh, in the past system about 10100 kWh). In particular, the energy saving associated to the braking energy recovering is higher about 10% points. This result stresses the renewed importance of a right management of this energy in the metro-transit system. The difference between the potential recoverable and the effective recovered braking energy is shown in Fig. 12, where it is evident that the value changes for each trains frequency with a non-linear function, and it is included in a range from 0.2 to 1.2 MWh. This differential energy is usually dissipated in heat by means of on-board rheostats. In alternative it could be recovered using storage station18,00

Total potenal recoverable braking energy

16,00 14,00 12,00

Total eecve recovered braking energy

[MW]

10,00 8,00 6,00 4,00 2,00 0,00 120" 150"

Dierence between total potenal recoverable braking energy and total eecve recovered braking energy

180"

240"

300"

360"

600"

Trains frequency [s]

Fig. 12. Comparison between system potential recoverable and effective recovered braking energy for each TS in CASE-10. Fig. 13. Comparison between potential recoverable and effective recovered braking power of a train in TS-F CASE 10.

M.C. Falvo et al. / Electric Power Systems Research 81 (2011) 21272138 Table 4 Some energy gures for BEV and PHEV. Vehicle type BEV (with weight < 1.000 kg) BEV (with weight 1.000 2.000 kg) PHEV (with weight 1.000 2.000 kg) Battery capacity (kWh) 16 30 16 Specic average energy consumption (Wh/km) 125 180 0.04 (l/km)

2135

Autonomy in km for a standard driving cycle (km) 130 170 60

is less than potential recoverable braking power. Their difference is the power dissipated in heat by means of on-board rheostats and it is about 2 MW. In terms of energy, by integration, it is possible to nd an effective recoverable braking energy of 251 kWh and a potential recovered braking energy of 205 kWh for the single train that means 50 kWh could be recovered using storage systems, in 1 h of service. Referring to TS-B (150 trains frequency) CASE 10, it is found an equal effective recoverable braking energy of 251 kWh and a potential recovered braking energy 249 kWh for a single train, that means only 2 kWh could be recovered using storage systems in 1 h of service. Those results point out that, if the storage system is located onboard the train, the value of energy that it has really to manage in 1 h is very variable in function of the trafc condition. In any case for a single train there are large powers (about 1 MW) and small energies (110 kWh), that means the most appropriate system to install onboard for storage are ultra-capacitors and not batteries. Anyway the introduction of storage systems onboard train is a solution that would require an upgrading of the vehicle drives design and so it could not be operated by a metro-transit utility in brief time.

demand on the distribution network use for their recharge. Obviously the power demand depends on the storage technology aboard the vehicles and on the type of recharge. In order to have a sensibility about the power consumption of the single EV in function of its technology, some data are summarized in Table 4, where the battery capacity, the average specic power consumption for km and the autonomy in km for a standard driving cycle are reported. Another aspect regards the choice of the suitable recharge typology on the basis of the service that the EV has to perform. Actually it is possible to select among the types summarized in Table 5. Each type of recharge can be directly associated to a Service Provider and so to a business model. Among the possibilities reported in Table 5, considering the present state of technologies and worldwide regulatory context, the only three feasible solutions are actually: slow-power recharge, adopted in Europe; fast-power recharge, adopted in Japan; battery swap, adopted in USA and Israel. Each choice obviously involves different electric infrastructures and it is characterized not only by different voltage and power level, but also by diverse impact on the distribution network used in terms of power ow and quality. The introduction of recharging points on a distribution network entails that: - the network must provide for the charging of EV in an intelligent way, according to the load needs and the network status. These functions are called grid to vehicle (G2V); - EV can perform many functions to support and to benet the network, improving its performance. These functions are called vehicle to grid (V2G). Some V2G functions are easy to be implemented and do not require a direct involvement of EV, such as: interruption of charge for controlling the power ows on the network; modulation of the charge for load shedding or demand side management, etc. Other V2G functions require a two-way energy transfer between network and vehicles, and they increase the number of charge/discharge cycles of the batteries, such as compensation for short breaks, compensation of voltage dips, etc. These second order functions could support the network management, with many advantages such as: reducing the overloads, improving the power quality, simplifying the voltage regulation, etc. With this aim it is clear that an Intelligent Energy Management System (IEMS), including different

5. Background and state of the art on electric vehicles (EV) spread The strong requirement of environmental sustainability and energy saving in urban transport got, as rst result, the appearance of a rainbow of solutions for on-road vehicles, characterized by low or zero emissions. Nowadays hybrid, electric, hydrogen or bio-fuels vehicles are available on the market. Disregarding hydrogen or bio-fuels vehicles, whose development is strongly linked to the availability of the primary resource, the main technological options matching the above-mentioned requirements and, at the same time, compatible with the present state of technology are: battery electric vehicles (BEV) and plug-in hybrid electric vehicles (PHEV). Considering the more recent literature [2633], a rst analysis on the market penetration of BEV and PHEV from 2010 to 2030 leads to the results: BEV will represent approximately 7%, whereas PHEV will do approximately 15% of the market of new vehicle sales in 2030. Considering that both the categories include plug-in EV, this consistent penetration will involve a consistent growth of power
Table 5 Types of recharge for BEV and PHEV and possible service provider. Recharge type Slow power Fast power Super Fast power Extra Fast power Battery swap Nominal power range for the recharging point <3.3 to 3.6 kW <20 kW <43 kW >43 kW power < limits of battery technology

Nominal voltage level Single Phase AC (compatible with LV grid) Three-phase AC (compatible with LV grid) Three-phase AC (charging stations connected to MV grid) Charging DC (charging stations connected to MV grid) Charging station connected to MV grid (Need space and equipment for handling batteries)

Possible service provider LV Distribution Utility (private recharging point) LV Distribution Utility or External Service Provider (public recharging point) MV Distribution Utility or External Service Provider (public recharging point) MV Distribution Utility or External Service Provider (public recharging point) External Service Provider (public recharging point)

2136

M.C. Falvo et al. / Electric Power Systems Research 81 (2011) 21272138

Fig. 14. Layout of the power system for the proposed integrated mobility system.

protocols, has to be developed: it has to require to EV different information (i.e. level of state of charge, minimum end time; characteristics of the battery or the vehicle, etc.) and then to select the possible protocol to obtain the desired level of charge, within a specic time, compatible with the electric network characteristics. At the same time, the IEMS can use EV to bring benets to both end users and distribution network: EV can become a resource for the distribution utilities for improving the performance of network in stability, security and power quality. Considering all those aspects, it is evident that the development of vehicles recharge infrastructures and the related standardization of the supply systems are the necessary condition for the electric propulsion diffusion. The development of a wide-scale infrastructure has to follow the increase of the number of electric vehicles otherwise the high requested investment cannot be justiable. In this sense an important role is expected to be played by public transport in urban areas: public procurement of BEV and PHEV could provide sufcient demand to allow building the recharging infrastructure that can later on be made available to private customers as well.

vide. Matching these evaluations with the statements on the EV reported in Section 5, the last step of this research is an innovative proposal: to exploit plug-in EV batteries, as storage system of the metro-train braking energy or, in other words, to use the recoverable train braking energy as the source for the recharge of BEV or PHEV batteries. This proposal could be dened as a project of energy integration of two urban mobility systems for the resolution of their single problems. The power system architecture for realizing this integration is shown in Fig. 14. Considering the evaluation and business models reported in Section 5, Table 5 shows that, in case of Fast Power and Battery Swap recharge protocol, the Service Provider could be or a MV Distribution Utility (DSO) or an External Service Provider. Besides the two types of recharge correspond to different business models: - the Service Provider could be the holder of the only recharge points, connected to the distribution network and so it sells the energy for the EV battery recharge; - the Service Provider could be the direct holder of the batteries that could give in leasing to the nal users: it sales stored energy in battery by a temporary rent of the batteries (payment for the good + service). These two business models are applicable for all the companies that are not DSO, but that have the properties of a MV network that can be used for install EV recharge points. A clear example is represented by the metro-transit system electric companies: if they work as External Service Provider, they do not have to ask for a new connection at LV or MV DSO, but they can use their own network for this scope, according to the above reported business models. In this way two transport systems, metro-transit and EV, would be integrated by the energy point of view. The conrmation of the feasibility of this energy integration is given by the evaluations, reported in the previous sections:

6. A new proposal for urban transport: metro-transit and EV integration From the overview on the actual solutions for the metro-transit braking energy management, reported in Section 2, and from the results of the analysis on a real metro-transit line, described in Section 4, the right conclusion has been that the introduction of stationary storages in metro-transit for the braking energy management, is becoming roughly a necessity for energy saving (with a total recovering, possible energy saving from 30% to 38%) and comfort issues. In favor of this proposal, there are the batteries technological evolution and further benets, in terms of security, quality and costs that the storage system insertion in ESS can pro-

M.C. Falvo et al. / Electric Power Systems Research 81 (2011) 21272138

2137

- the daily value of energy that could be storage in stationary system located in ESS is demonstrated to be about 15 MWh; - the daily energy consumption for an EV is about 15 kWh for guaranteeing an autonomy of 5060 km on a standard city cycle, considering a recharge at 3 kW for 6 h. Matching those data, it is clear that theoretically the only energy saving of the metro-transit trains braking could guarantee the recharge of about 1.000 EV. Obviously the power system for integration would be realized by an IEMS, implementing the right coupling at least to guarantee: - on the metro-transit side, the maximum recovering in EV batteries of the energy made available from the trains during the breaking phase (V2G function); - on the EV side, the right recharge of batteries (G2V function). Both two functions have to be performed maximizing the energy efciency of the whole transport system. To those basic purposes of the energy integration, other functions could be added improving the IEMS capability. An example of advanced V2G function could be the control of the EV batteries power recharge for leveling the ESS load during the day hours when the metro-transit is in service, with a clear benet in terms of power quality improvement and reduction of operational cost. An example of advanced G2V function could be a recharging service in daytime (i.e. in battery swap if metro-transit Company has a system for batteries replacement in a few minutes) for: - EV for the private transport, parked in the working day-hours; in case of battery swap the nal costumer has to pay for good + service, because the battery is leased and he pays rent with the energy present accumulator; - its own EV that, if recharged during the day when the metrotransit is in service and its braking energy is available, can be then used for the public transport during the night on the same path of the metro-transit that is out of service. In this way also a service integration would guarantee between the two transport systems: a 24-h non-stop public transport service is delivered. The same infrastructure could be used for the recharge of other EV (i.e. vehicles for goods distribution, little electric buses, etc.) also during the night, when the metro trains load is quite zero, because of the service interruption. In this way the maximum exploitation of an infrastructure already existing would guarantee. Obviously strategic points for the construction of this kind of infrastructure could be the interchange parking areas located near the main metro-transit stations, where right space is available also for the realization of the battery swap system. In conclusion, the proposed architecture allows achieving two project purposes: - a signicant energy saving of the whole integrated transport system, thanks to the re-using of the recovered braking energy in the metro transit system; - a considerable reduction of environmental impact of the whole integrated transport system, both in the local urban area, thanks to the lower level of polluting emissions, guaranteed by the use of the EV, and in global environment, thanks to the lower volume of green house gases (GHG) emissions and a more efcient use of the primary power in the energy chain. Moreover the proposed mobility system would be realizable with a reduced economic impact thanks to the optimal exploitation of the existing electric plant, avoiding new investments on the power system. The only necessary deal would regard the IEMS that

Table 6 E-Mobility project main gures: objectives to 2020. Companies involved Italian pilot cities Electrical vehicles number (lithium ion fed) Parking with recharging points 3 (2 automotive companies 1 electrical company) 3 (Rome, Pisa and Milan) 100 (at least) 32% indoor private 34% outdoor private 34% public parking 400 total 300 public 100 private

Number of recharge points

would make over the metro-transit supplying system a Smart Grid for a Smart Mobility. Referring to the city of Rome, whose metro-transit system has been analyzed, this proposal meets the market requirement. In fact, work is in progress for the implementation of the E-Mobility project, founded by two foreign automotive companies in collaboration with a big Italian electrical company. The project has the aim of the diffusion of city car EV, as a result of the realization of a capillary recharge network, whose main gures are summarized in Table 6: 3 large Italian cities are involved (Rome, Pisa and Milan) with their local DSO, 400 recharge points are scheduled, whose 100 private and 300 public, with the objective to 2020 to get a 34% of recharge points in public parking and 1 point of recharge for 8 EV. The project expects that these recharge points are supplied both by local DSO and by private External Service Providers. At the light of the proposal of the paper, it is clear that this project could represent a big change and challenge for the metro-transit city company. 7. Conclusions The recovering of trains braking energy in stationary storages nowadays would allow a global energy saving, with an additional improvement in case of optimization of the storage system design. In order to have a positive and real feedback on this statement, the authors performed an analysis for the assessment of the braking energy on a real metro-transit line, by home-made specialized software. The application results conrmed this thesis and pointed out a potential energy saving from 30% to 38% for the analyzed real metro-line. Considering the actual trend of diffusion of EV for the urban mobility, and the strengthened importance to implement IEMS for guaranteeing G2V and V2G functions, a new proposal for the design of a sustainable urban mobility system is formulated. The idea is focused on the integration of the two urban mobility systems: the metro transit system and EVs, through the construction of a Smart Grid for a Smart Transport. It is demonstrated that the proposed integration provides a resolution of the energy problems regarding the two systems and in general it can guarantee the minimization of environmental impact, as well as the optimization of the service as a result of the maximization of the transport bid in time. Moreover the solution presents the minimum economic impact, because of the optimization of the logistic and technological application of the existing electric power plant. References
[1] European Commission, European Energy and Transport Trends to 2030, DG TREN, 2008. [2] International Energy Agency, Energy Balance of OECD Countries, IEA, 2006. [3] European Commission, EU Energy and Transport in Figures, 2008. [4] L. Eudy, J. Zuboy, Overview of advanced technology transportation, 2004 update, Technical Report on Advanced Vehicle Testing Activity, Energy Efciency and Renewable Energy, U.S. Department of Energy, DOE/GO-1020041849, August 2004. [5] H.-J. Chuang, C.-S. Chen, C.-H. Lin, C.-H. Hsieh, C.-Y. Ho, Design of optimal coasting speed for saving social cost in Mass Rapid. Transit systems, in: Proceedings

2138

M.C. Falvo et al. / Electric Power Systems Research 81 (2011) 21272138 [20] A. Capasso, R. Lamedica, R. Giglioli, G. Zini, Electrical Sub-Stations with electrochemical storages for metro-transit systems, in: Ingegneria Ferroviaria n. 4, April, 1984. [21] A. Adinol, R. Lamedica, C. Modesto, A. Prudenzi, S. Vimercati, Experimental assessment of energy saving due to trains regenerative braking in an electried subway line, IEEE Transactions on Power Delivery 13 (October (4)) (1998) 15361542. [22] A. Capasso, R. Lamedica, G. Maranzano, R. Bartoni, M. Francisi, Deterministicprobabilistic simulation model for the calculation of the potential capacity of power systems for metro-transit transport, in: Proc. AEIT Annual Meeting, Capri, September, 2006. [23] A. Capasso, M.C. Falvo, U. Grasselli, R. Lamedica, R. Bartoni, M. Francisi, G. Maranzano, A planning study on power systems of metro-transit transportation system, in: Proceedings IEEE SPEEDAM 2008, Ischia, Italy, June 1113, 2008. [24] M.C. Falvo, R. Lamedica, R. Bartoni, G. Maranzano, Energy saving in metrotransit systems: impact of braking energy management, in: Proceedings SPEEDAM 2010 (IEEE), Pisa, Italy, June 1416, 2010. [25] M.C. Falvo, F. Foiadelli, Preliminary analysis for the design of an energy-efcient and environmental sustainable integrated mobility system, in: Proceedings IEEE PES General Meeting 2010, Minneapolis, USA, 2529 July, 2010. [26] A. Burke, Present Status and Marketing Prospects of the Emerging HybridElectric and Diesel Technologies to Reduce CO2 Emissions of New Light-Duty Vehicles in California, Institute of Transportation Studies, University of California, Davis, Online availability: http://repositories.cdlib.org/itsdavis/UCD-ITSRR-04-2, June 2004. [27] A.F. Burke, Batteries and ultracapacitors for electric, hybrid, and fuel cell vehicles, Proceedings of the IEEE 95 (April (4)) (2007) 806820. [28] J. Taylor, A. Maitra, M. Alexander, D. Brooks, M. Duvall, Evaluation of the impact of plug-in electric vehicle loading on distribution system operations, in: IEEE PES General Meeting, 2009. [29] L. Jinming, H. Peng, Modeling and control of a power-split hybrid vehicle, IEEE Transactions on Control Systems Technology 16 (November 6) (2008) 12421251. [30] M. Ceraolo, A. di Donato, G. Franceschi, A general approach to energy optimization of hybrid electric vehicles, IEEE Transactions on Vehicular Technology 57 (May (3)) (2008) 14331441. [31] P. Menga, M. Ceraolo, An evaluation of global environmental and energy value of vehicle technologies, in: 3rd European Ele-Drive Transportation Conference, Geneva Switzerland, March 1113, 2008. [32] M.D. Galus, G. Andersson, Demand management of grid connected plug-in hybrid electric vehicles (PHEV), IEEE Transactions on Vehicular Technology 58 (October (8)) (2009). [33] X. Li, L.A.C. Lopes, S.S. Williamson, On the suitability of plug-in hybrid electric vehicle (PHEV) charging infrastructures based on wind and solar energy, in: IEEE PES General Meeting, 2009, http://www.ev-charging-infrastructure.com.

[6]

[7] [8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

Third International Conference on Electric Utility Deregulation and Restructuring and Power Technologies, 69 April, 2008, pp. 28332839. Y.V. Bocharnikov, A.M. Tobias, C. Roberts, S. Hillmansen, C.J. Goodman, Optimal driving strategy for traction energy saving on DC suburban railways, Electric Power Applications, IET 1 (September (5)) (2007) 675682. F. Ruelland, K. Al-Haddad, Reducing Subways Energy, in: Proceedings Electrical Power Conference, IEEE Canada, 2526 October, 2007, pp. 261267. M. Steiner, M. Klohr, S. Pagiela, Energy storage system with ultracaps on board of railway vehicles, in: European Conference on Power Electronics and Applications, 25 September, 2007, pp. 110. M. Brenna, F. Foiadelli, E. Tironi, D. Zaninelli, Ultracapacitors application for energy saving in subway transportation systems, in: Proceedings International Conference on Clean Electrical Power, 2123 May, 2007, pp. 6973. F. Foiadelli, M.C. Roscia, D. Zaninelli, Optimization of storage devices for regenerative braking energy in subway systems, in: IEEE PES General Meeting, Montreal, Canada, 1822 June, 2006. H.-J. Chuang, C.-S. Chen, C.-H. Lin, C.-H. Hsieh, C.-Y. Ho, Design of optimal coasting speed for saving social cost in Mass Rapid. Transit systems, in: Proceedings 3rd International Conference on Electric Utility Deregulation and Restructuring and Power Technologies, 69 April, 2008, pp. 28332839. M. Brenna, F. Foiadelli, M.C. Roscia, D. Zaninelli, Prospective for energy saving by means of Ultracapacitors in electric systems for transportation, in: Fourth International Conference and Exhibition on Ecological Vehicles and Renewable Energies, Monaco, France, 2629 March, 2009. S. Samuel, L. Austin, D. Morrey, Automotive test drive cycles for emission measurement and real-world emission levels-a review, in: Proceedings Institution of Mechanical Engineers, Part D, Journal of Automobile Engineering, Professional Engineering Publishing 216 (7) (2002) 555564. M. Barcaglioni, F. Gaddi, R. Giglioli, R. Salutari, G.A. Saraceno, G. Zini, Battery storage plant to improve energy saving and security of subway electric supply, in: IEE 1995 Electric Railways in a United Europe Conference, 2730 March, 1995. B. Mellit, C.I. Goodman, R-I.M. Arthurton, Simulation for studying operational and power supply conditions in rapid transit railways, Proceedings IEE 125 (4) (1978) 298303. P. Minciardi, S. Savio, G. Sciutto, Models and tools for simulation and analysis of metro-rail transit systems, in: Proc. IV COMPRAIL, Madrid, 79 September, 1994. L. Aligata, et al., Automatic planning methodology of power systems for metrotransit transport. An application to Naples metro-transit system, in: Proc. AEI General Meeting, S. Margherita Ligure, 1979. R. Billinton, et al., Development of the ELDC and reliability evaluation of composite power system using Monte Carlo method, in: Proc. Power Engineering Society Summer Meeting, 1620 July, 2000. A. Capasso, R. Lamedica, E. Massara, E. Pettinelli, A simulation model for the electrical power consumptions calculation in a metro-transit system, in: Proceedings AEI General Meeting, Bologna, 1982.