1234 IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 5, NO.

4, OCTOBER 2014

Urban Scale Photovoltaic Charging Stations

for Electric Vehicles
Morris Brenna, Member, IEEE, Alberto Dolara, Student Member, IEEE, Federica Foiadelli, Member, IEEE,
Sonia Leva, Senior Member, IEEE, and Michela Longo

Abstract—An energy integration between photovoltaic (PV) Through an intelligent integrating of vehicles, electricity,
systems and electric vehicles (EVs) can help to overcome prob- free renewable fuel, and market opportunities, it is possible to
lems related to the feasibility of a more sustainable mobility, most make a zero-carbon future. Renewable energy such as photo-
of all in urban context. This paper aims to examine the potential
and the technical benefits of using PV systems as energy supplier voltaic (PV) integrated in a vehicle-to-grid (V2G) concept
for charging EVs. For this purpose, an urban scale integrated could be a solution to alleviate carbon footprint. An effective
system is presented through a mathematical model that considers way to maximize the utilization of space and capacity of the
the power flows related to the PV generator, the charging station, V2G parking lot (VPL) is the integration of PV rooftops as a
and electric grid. The most significant result is the evaluation of micro-resource in the VPL facility [4]. A plug-in hybrid EV
the self-consumption in order to optimize the interaction between
the PV system and charging station for EVs. This analysis has (PHEV) is essentially a hybrid vehicle with a larger battery
been conducted for different vehicles typologies and different PV pack [5], [6]. Therefore, it runs on electricity when its battery
installations, giving rise to more than 9000 different cases and state-of-charge (SOC) is high. Otherwise, the ICE takes over
allowing to consider the wideness of the self-consumption range and the vehicle consumes gasoline similar to a hybrid vehicle.
for different months. The battery pack can be recharged through a charging station
Index Terms—Charging infrastructure, electric vehicles (EVs), connected to an electric power grid. PHEVs are characterized
photovoltaic (PV) energy. by their all-electric range (AER). The quantity of electricity
demanded by PHEVs raises concerns about their potential
negative impacts on the grid. Therefore, it is important to
I. I NTRODUCTION analyze possible alternative solutions to supply the EVs.
Considering the wide diffusion of renewable sources, it comes
I NFLUENCED by recent developments in the oil market
and greater sensitivity in terms of pollutants, the last
decade has seen a growing interest in electric mobility. Despite
to mind the idea to exploit these energies to help the EVs
charging [7], [8]. Other sources can also be considered for
major technological developments in various areas of research, charging EVs. For example, in the metropolitan areas, braking
there are still many issues to be addressed. Among these, the energy coming from the urban transportation systems can be
need for a reliable and diverse charging infrastructure which exploited for the EVs charge [9].
meets different user needs is placed at the forefront. In its These studies often refer to scenarios in which renewable
directive on introduction of renewable energy, the European solutions benefit of national economical incentives. Recently,
Parliament states that 20% of the energy use within the union due to the economical crisis, these funds have been
should be covered by renewable energy sources by 2020 [1], dramatically reduced, especially in Europe, and therefore,
[2]. Today, there are several suggested alternatives to fossil it is important to reach an economical sustainability of the
fuels that are being well investigated in order to replace fossil so-called green solutions. A further step is therefore thinking
fuels in many aspects of human life and industries. Several about applications to increase the self-consumption in order
economical, ecological, and political ambitions lie behind this to improve the profitability of renewable energy power plants
replacement, one of which is greenhouse gas pollution leading also in the absence of economical incentives. Considering the
to global warming. A sector mostly involved from this issue is scientific background, this idea is already under study and
the transportation sector that sees the replacing of a number of development in building applications where the PV production
conventional internal combustion engine (ICE) vehicles with can directly supply building devices as heat pumps or can be
electric vehicles (EVs) [3]. stored in dedicated batteries to reach a self-sustainability of
the system [10], [11].
Manuscript received February 27, 2014; revised May 15, 2014; accepted This paper addresses to this concern and it analyzes the
July 19, 2014. Date of publication August 14, 2014; date of current version benefits of integrating a PV carport in the charging station
September 16, 2014. and how to better exploit the available energy. Since most
The authors are with the Department of Energy, Politecnico di Milano,
Milano 20133, Italy (e-mail: morris.brenna@polimi.it; alberto.dolara@ cars used for work purposes remain parked during the day,
polimi.it; federica.foiadelli@polimi.it; sonia.leva@polimi.it; michela.longo@ a common scenario could be to charge PEVs during work
polimi.it). hours. The paper focuses on those charging infrastructures
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org. dedicated to commuters with access to reserved parking areas
Digital Object Identifier 10.1109/TSTE.2014.2341954 at their workplace. Considering an installation having a good

1949-3029 c 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications standards/publications/rights/index.html for more information.

Authorized licensed use limited to: UNIVERSITY TENAGA NASIONAL. Downloaded on July 15,2020 at 04:04:52 UTC from IEEE Xplore. Restrictions apply.
BRENNA et al.: URBAN SCALE PV CHARGING STATIONS FOR EVs 1235

TABLE I
PARAMETERS U SED FOR A NALYSIS

availability of space, a good exposure to solar radiation,

and a reduced presence of shading [12], an outdoor parking
justifies the development of a PV carport in replacement
of the traditional shelter. However, the overall efficiency of
renewable energy systems requires a high level of optimization
of their components [13]. The efficient exploitation of nonpro-
grammable renewable sources is also related to the accuracy
in the forecasting of their production [14]. Finally, the load
profile depends on several variables that can be assumed to be
stochastically distributed, including actual SOC of the battery,
parking duration, parking type, and vehicle powertrain [15]. A
statistical approach based on measurements campaign is useful
to improve the accuracy in load profile predictions [16].
In this study, a PV charging system has been analyzed. In
Section II, a mathematical law that estimates the production
from PV shelter as a function of different parameters based
on a statistical approach is defined. Section III presents the
mathematical model of the system that considers the power
flows related to the PV generator, the charging station, and Fig. 1. Daily energy production in different regions of Italy. The horizontal
electric grid. Section IV contains details of the proposed charg- lines represent the average values.
ing system, whereas Section V applies the previous analysis.
The inputs required by the program are the following.
1) The geographical coordinates of the site.
II. E NERGY P RODUCTION F ROM PV S YSTEM 2) Peak power of PV plant. In this work, a base power of
This section presents an energy production estimation of 1 kWp is considered to easily extend the results to PV
PV systems, considering different factors such as localization, systems of any peak power.
technology, orientation, and period. In particular, three geo- 3) Installation: free-standing.
graphical locations in Italy representing different solar radiation 4) Estimated system losses: 14%.
levels (north Italy—low solar radiation, center Italy—medium 5) PV technology, azimuth, and tilt angles as reported in
solar radiation, and south Italy—high solar radiation), different Table I.
orientations of the PV generator (in terms of tilt and azimuth The outputs of the program are the following.
angles), and various PV technologies have been considered. 1) Ed : Average daily electricity production from the given
Table I summarizes the values used in this analysis. system (kWh).
The goal of this analysis is to define a mathematical law that 2) Em : Average monthly electricity production from the
allows assessing the production of a PV system as a function given system (kWh).
of the month and the orientation of the PV generator. In this 3) Hd : Average daily sum of global irradiation per square
paper, the regression approach is applied. meter received by the modules of the given system
2
The data used to assess the production of the PV sys- (kWh/m ).
tems have been extracted from the database of the software 4) Hm : Average sum of global irradiation per square meter
2
Photovoltaic Geographical Information System (PVGIS) [17], received by the modules of the given system (kWh/m ).
developed by the Joint Research Center, a Directorate General In particular, in this work only Ed has been considered.
of the European Commission. The outputs provided by the Fig. 1 shows the variation of the daily electricity production
program are based on a climatic database that considers Ed as a function of azimuth and tilt angles and installation
measured data from 2000 to 2009. region.

Authorized licensed use limited to: UNIVERSITY TENAGA NASIONAL. Downloaded on July 15,2020 at 04:04:52 UTC from IEEE Xplore. Restrictions apply.
1236 IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 5, NO. 4, OCTOBER 2014

TABLE II
C OEFFICIENTS FOR (1) FOR D IFFERENT PV T ECHNOLOGY AND G EOGRAPHICAL A REAS

It is possible to observe that the pairs of tilt and azimuth

angles that maximize the energy production are around 30◦
and 0◦ , respectively, for all regions. In terms of average value
over a year, north of Italy is characterized by the lower value,
2.94 kWh/(kWp·day), whereas south of Italy is characterized
by the highest value, 3.45 kWh/(kWp·day), highlighting that
the energy production in Italy increases from the North to the
South. The PV technologies do not considerably influence the
production and the applied method. Therefore, the choice of
the crystalline silicon PV technology is based on its diffusion
in the installed PV plants.
A general mathematical law to estimate the daily production Fig. 2. Example of the noncoordinability of the production curve and load
can be obtained from the analysis of the curves represented in curve.
Fig. 1
Ed (m, T, A) = k1 · T 2 + k2 · A2 + k3 · m2 the average energy consumption of an electric vehicle is about
+ k4 · T + k5 · m + k6 · T · m + k7 (1) 0.2 kWh/km, the annual energy production estimated starting
from (1) is comparable to the energy required by EVs. In
where m is month, T is the tilt in degrees, A is the azimuth
the worst condition (north of Italy, winter months and both
in degrees, and k1 , k2 , k3 , k4 , k5 , k6 , and k7 are weight
vehicles in charge), it is possible to estimate that 30% of
coefficients. In Table II, the values of the coefficients for
the recharge energy have to be absorbed from the grid. For
the different regions using Minitab software are reported. It
these reasons, the solar energy can be suitably exploited for
is possible to observe that the values of the coefficient of
charging EVs.
determination (R2 ), that indicates how the mathematical model
However, there is a limit in the annual analysis, because
fits the data provided by PVGIS, are quite high.
the PV production curve (strongly related to the weather
conditions) and the load curve (that depends by EV charging
III. M ATHEMATICAL M ODEL OF THE P OWER F LOWS
and working conditions) do not match.
The assessment of the energy flows within the charging In this work, the production curve of a PV carport sized for
system requires the knowledge of the daily production curves two EVs has been considered. As per load curves, combina-
of the PV system and the demand curves of the EVs. tions of two different EV typologies are considered: a car (C)
A first analysis could be based on the annual energy balance. with battery capacity of 24 kWh and a quadricycle (Q) with
In this case, the average daily energy produced by the PV battery capacity of 9 kWh. Charging power has been assumed
system could be evaluated by applying the method presented constant throughout the charging period.
in Section II to PV systems dedicated to the EVs charging To assess in detail the potential of PV charging stations, it
systems installed on carports. Usually, these parking areas is necessary to carry out the comparison between these two
have a footprint between 10 and 15 m2 available for each power curves. Fig. 2 shows an example of the comparison
vehicle. Therefore, it is possible to install on their roofs between the charging power curves and the PV production
a PV generator that has a peak power between 1.5 and curve in the case of two cars charged sequentially—the first
2.5 kWp/car. one in the morning and the second one in the afternoon—with
Considering the typical roof pitches of the carports and con- the same amount of energy but with different charging powers.
sidering that the average distance travelled annually in urban The matching between the two curves cannot be generalized
and suburban areas by the commuters is about 10 000 km and by an equation, but it has to be done analyzing case by case.

Authorized licensed use limited to: UNIVERSITY TENAGA NASIONAL. Downloaded on July 15,2020 at 04:04:52 UTC from IEEE Xplore. Restrictions apply.
BRENNA et al.: URBAN SCALE PV CHARGING STATIONS FOR EVs 1237

For this reason, the following two main scenarios are The energy produced by the PV plants during the charging
analyzed as follows. process, called “self-consumption” Eself -cons , has been esti-
1) Contemporary charging of two vehicles: Charging be- mated for each month considering the daily average production
gins on arrival at work, at 9:00 A . M. curve  tf
2) Sequential charging of vehicles: The company organizes Eself -cons = PPV (t)dt (5)
shifts for charging EVs, one that starts at 9:00 A . M. and ti
one at 1:30 P. M. where PPV (t) is the power produced by the PV system. It can
Each scenario considers: be evaluated by using the monthly average curve to determine
1) Three combinations of EVs: two cars (C1 and C2), a car the sun–rise and the sun–set and the average daily electricity
(C) and a quadricycle (Q), and two quadricycles (Q1 and production from (1).
Q2). Equation (5) is applied in all scenarios because the PV
2) Different values of charging power: 3.7, 7, 11, and generator peak power is always lower than the minimum power
22 kW. However, current market available quadricycles required for the recharge. The condition where the energy
cannot be charged with power exceeding 3.7 kW; there- produced by the PV systems is at the same time supplied to
fore, the number of scenarios that contain one or two vehicles and to the electric grid never occurs.
quads is reduced. Consequently, for every month and every case, the energy to
3) The energy that must be provided to the vehicles will withdraw from the grid Egrid to complete the recharge process
depend on the difference between the initial and final can be estimated as follows:
states of charge of the battery (ΔSOC). Three values of
Erech
ΔSOC (0.6, 0.4, and 0.2) will be considered in C1–C2 Egrid = − Eself -cons . (6)
ηrech
and C–Q combinations. Instead, considering that there
is only one available charging power for the quadri- Finally, the amount of energy produced by the PV system
cycles, five values of ΔSOC (0.6, 0.5, 0.4, 0.3, and and injected into the grid can be calculated as the difference
0.2) will be taken in account for the two quadricycles between the estimated energy produced by the PV system
combination in order to increase the number of the given in (1) and the estimated self-consumption for EVs
cases. recharging
The different combinations give rise to more than 9000 EPV−>grid = Ed − Eself -cons . (7)
different cases here analyzed. The values so obtained are calculated for each day. The
The evaluation of the energy flows shall be according to the following equations are used for the monthly value:
following scheme.
Starting from the combination of vehicles and the difference Erech,m = F · nday,month · Erech
between the initial and final state of charge ΔSOC and their Eself -cons,m = F · nday,month · Eself -cons
battery capacity Cbatt , the energy Erech required in a day to (8)
Egrid,m = (Erech,m − Eself -cons,m ) · F · nday,month
recharge both vehicles v1 and v2 is obtained
EPV−>grid,m = (Ed − Eself -cons ) · F · nday,month .

2
where nday, month is the number of days in the considered
Erech = Erech,v1 + Erech,v2 = Cbatt,v · ΔSOCv . (2)
month.
v=1

Charging time is calculated considering the charging power IV. D ETAILS OF PV C HARGING S YSTEMS
Prech , the energy required Erech , and the efficiency of the
charging process ηrech . In case of contemporary charge, charg- In this work, the charging system analyzed consists of a
ing time Δt and consequently the initial (ti ) and final (tf ) charge point for two PEVs that can be supplied from the
charging instants are grid and/or from the PV generator installed on the roof of
the carport. In particular, it has been considered as a PV
Erech shelter installed in a company car park area where the EVs
Δt = ; tf = ti + Δt. (3)
Prech · ηrech are available for recharge from 9:00 A . M. to 5:00 P. M.
The PV shelter is designed for two EVs and takes up an
In case of sequential charge, charging time for each
area of about 25 m2 . The PV generator is made up by 15
vehicle is
monocrystalline PV modules, for a total power of 3675 Wp.
Erech,v1 It is assumed that the tilt angle is equal to 20◦ and azimuth is
Δt1 = ; tf,1 = ti,1 + Δt1
Prech,v1 · ηrech equal to 0◦ .
(4) Fig. 3 shows the PV system, the elements of the charging
Erech,v2
Δt2 = ; tf,2 = ti,2 + Δt2 . station, and the energy flows. The red arrows represent the
Prech,v2 · ηrech
flows of energy produced by the PV system, whereas the blue
It is important to note that the charging time of each arrows represent the flows of energy drawn from the electric
vehicle has to be always lower than the time slot assigned grid. When the energy demand from the user is nil, the energy
for its recharge. As a consequence, the vehicle that starts its produced from the PV system is injected into the grid. When
recharge at 9:00 A . M. reaches the state of full charge before the one or two EVs are connected to the recharge station, the
beginning of the recharge of the second vehicle at 1:30 P. M. energy produced by the PV system is completely supplied to

Authorized licensed use limited to: UNIVERSITY TENAGA NASIONAL. Downloaded on July 15,2020 at 04:04:52 UTC from IEEE Xplore. Restrictions apply.
1238 IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 5, NO. 4, OCTOBER 2014

Fig. 3. Power flows of the proposed charging system.

Fig. 5. (a) ΔSOC = 0.6. (b) ΔSOC = 0.4. (c) ΔSOC = 0.2. Contempo-
rary charging of a car (C) and a quadricycle (Q) as a function of ΔSOC.
The recharge power of the car (C) and quadricycle (Q) are reported on the
category axis, starting from the top.

Fig. 4. (a) ΔSOC = 0.6. (b) ΔSOC = 0.4. (c) ΔSOC = 0.2. Contempo-
rary charging of two cars (C1 and C2). The recharge power of car 1 and car Fig. 6. Contemporary charging of two quadricycles (Q1 and Q2). The ΔSOC
2 are reported on the category axis, starting from the top. of each quadricycle are reported on the category axis.

the onboard batteries. This operating condition is related to the The evaluation of self-consumption has been made consid-
size of PV generator. In fact, its peak power is lower than the ering several values of ΔSOC, recharge power, combinations
minimum power required for the recharge; therefore, it is not of cars (C) and quadricycles (Q), and several sites. Charging
possible that the PV system supplies both vehicles and grid. powers are chosen as a function of ΔSOC. Taking into account
that high charging powers reduces the batteries life they have
to be avoided if there is not the need to provide significant
V. R ESULTS AND D ISCUSSION amounts of energy. Therefore, the maximum charging power
This section presents the main results of the analysis in term (22 kW) is considered only when the energy that has to
of the quantification of energy flows into the PV charging be supplied to the batteries is maximum (ΔSOC = 0.6).
stations. Among of the energy flows, the self-consumption Charging power of 11 kW is excluded when the energy that has
Eself -cons is the most important flow; it represents the amount to be supplied to the batteries is the minimum (ΔSOC = 0.2).
of energy that is produced by the PV charging station that is The ainnual value of Eself -cons is calculated starting from the
directly stored in the PEVs batteries. monthly Eself -cons .

Authorized licensed use limited to: UNIVERSITY TENAGA NASIONAL. Downloaded on July 15,2020 at 04:04:52 UTC from IEEE Xplore. Restrictions apply.
BRENNA et al.: URBAN SCALE PV CHARGING STATIONS FOR EVs 1239

Fig. 7. (a) ΔSOC = 0.6. (b) ΔSOC = 0.4. (c) ΔSOC = 0.2. Sequential charging of two cars. The recharge power of car 1 (C1) and car 2 (C2) are
reported on the category axis, starting from the top.

In the following, the annual average, minimum, and maxi-

mum value of self-consumed energy are reported. These values
are expressed as a percentage of the recharge energy Erech and
they are grouped in histograms where the bars indicate the
annual value, whereas the error bars indicate the minimum
and maximum values achieved in the months of December
and June, respectively.

A. Scenario 1: Contemporary Charging

The first set of graphs (Figs. 4–6) shows the results obtained
for the contemporary charging scenario. Fig. 4 shows the self-
consumption in case of contemporary charge of two cars,
considering the values of ΔSOC of 0.6 (case a), 0.4 (case b),
and 0.2 (case c) for each car; charge powers taken into account
for each car are shown on the category axis.
Fig. 5 shows the self-consumption in case of contemporary
charge of a car and a quadricycle, considering the values of
ΔSOC of 0.6 (case a), 0.4 (case b), and 0.2 (case c); car and
quadricycle charge powers taken into account are shown in the
first and in the second row on the category axis, respectively.
Fig. 6 shows the self-consumption in case of contemporary
charge of two quadricycles, considering the values of ΔSOC
form 0.2 to 0.6 as reported on the category axis.

B. Scenario 2: Sequential Charging

The second set of graphs (Figs. 7–9) shows the results
obtained for the sequential charging scenario. The same com-
binations of vehicles, charging power, and ΔSOC considered
in the scenario 1 are taken into account. In the case of sequen- Fig. 8. (a) ΔSOC = 0.6. (b) ΔSOC = 0.4. (c) ΔSOC = 0.2. Sequential
charging of a cars and a quadricycle. The recharge power of the car (C) and
tial charging with different recharging power or with different quadricycle (Q) are reported on the category axis.
energy demand, each case has to be considered twice because
the order of recharge has to be taken into account. Fig. 7
shows the self-consumption in case of sequential charge of in case of contemporary charge of a car and a quadricycle,
two cars, considering the values of ΔSOC of 0.6 (case a), 0.4 considering the values of ΔSOC of 0.6 (case a), 0.4 (case b),
(case b), and 0.2 (case c). Fig. 8 shows the self-consumption and 0.2 (case c). Fig. 9 shows the self-consumption in case

Authorized licensed use limited to: UNIVERSITY TENAGA NASIONAL. Downloaded on July 15,2020 at 04:04:52 UTC from IEEE Xplore. Restrictions apply.
1240 IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 5, NO. 4, OCTOBER 2014

VI. C ONCLUSION
This paper aims to examine the potential and the technical
benefits of using PV systems as energy supplier for charging
PHEV. For this purpose, first, suitable mathematical functions
have been developed to estimate the energy production from
PV systems in different geographical locations in Italy as a
function of tilt, azimuth, and month.
Successively, to determine the strategy that maximizes the
self-consumption recharging with the lowest power, a PV
Fig. 9. Sequential charging of two quadricycles. The ΔSOC of each quadri- carport combined with a charging system for two electric cars
cylce are reported on the category axis.
has been analyzed. The aspects linked to the energy flows,
considering the production of PV system, the absorption by
of contemporary charge of two quadricycles, considering the charging EVs, and the absorbed and injected energy to the
values of ΔSOC form 0.2 to 0.6. utility grid are evaluated, highlighting the technical sustain-
ability of the project. The comparability between the values
of the energy produced by PV and the one demanded by the
PHEVs is difficult to obtain due to the noncoordination of the
C. Comments on the Results solar source and the load.
A first analysis of these values shows that only a small Different scenarios and more than 9000 cases are analyzed
proportion of energy produced from PV can be supplied in terms of absorption profiles and energy. The most significant
directly into the electric vehicle batteries. The sequential results of this work is the percentage of energy coming from
charging allows a self-consumption higher than the simulta- the PV system to the EVs with respect to the energy required
neous charging. This increase steps around 10%–15% points by the charging point, that ranges from 1%–3% to 56%–72%.
and is far from taking a double absorption value (except Moreover, the energy flows strongly depend on month. The
in the case of the two quadricycle). This effect penalizes, maximization of the energy flow from PV system to electric
especially in the winter months. For both types of recharging, vehicle requires quite long and low power charges that allow
the combination of car and quadricycle, with maximum ΔSOC exploiting the hours when production of the PV shelter is high.
and minimum charging power presents the most favorable However, an energy storage system is necessary.
case. The combination of two cars with a maximum charging
power of 22 kW is the most unfavorable case regarding.
R EFERENCES
For both types of recharging, the combination of car and
quadricycle with maximum ΔSOC and minimum charging [1] J. Widén, “Correlations between large-scale solar and wind power in a
future scenario for Sweden,” IEEE Trans. Sustain. Energy, vol. 2, no. 2,
power presents the most favorable case. The combination of pp. 177–184, Apr. 2011.
two cars, with a maximum charging power of 22 kW is the [2] S. Rezaee, E. Fariah, and B. Khorramdel, “Probabilistic analysis of
most unfavorable case regarding. In fact, because the peak plug-in electric vehicles impact on electrical grid through homes and
parking lots,” IEEE Trans. Sustain. Energy, vol. 4, no. 4, pp. 1024–1033,
power of the PV shelter is equal to 3.7 kW and the maximum Oct. 2013.
average power is slightly higher than 2.6 kW, it is evident [3] U. C. Chukwu and S. M. Mahajan, “V2G parking lot with PV rooftop
that, by requiring more charging power, most of the energy is for capacity enhancement of a distribution system,” IEEE Trans. Sustain.
Energy, vol. 5, no. 1, pp. 119–127, Jan. 2014.
drawn from the grid. [4] Z. Darabi and M. Ferdowsi, “Aggregated impact of plug-in hybrid
The combination of two quadricycles is the case that takes electric vehicles on electricity demand profile,” IEEE Trans. Sustain.
greater advantage in moving from simultaneous to sequential Energy, vol. 2, no. 4, pp. 501–508, Oct. 2011.
[5] M. Falahi, H.-M Chou, M. Ehsani, L. Xie, and K. L. Butler-Purry,
charging. In fact, the sequential recharge shifts the second “Potential power quality benefits of electric vehicles,” IEEE Trans.
vehicle in the time slot that starts at 13.30, when the PV Sustain. Energy, vol. 4, no. 4, pp. 1016–1023, Oct. 2013.
production is higher; therefore, it is possible to have a better [6] L. Cheng, Y. Chang, J. Lin, and C. Singh, “Power system reliability
assessment with electric vehicle integration using battery exchange
exploitation of the renewable source. In the case of simulta- mode,” IEEE Trans. Sustain. Energy, vol. 4, no. 4, pp. 1034–1042,
neous charging, there are groups of different scenarios that Oct. 2013.
lead to the same results. In particular, the self-consumption [7] X. Li, L. A. C. Lopes, and S. S. Williamson, “On the suitability of
plug-in hybrid electric vehicle (PHEV) charging infrastructures based
(at equal ΔSOC) depends only on the lower charging power on wind and solar energy,” in Proc. Power Energy Soc. Gen. Meeting,
considered for the particular scenario. Since the power from 2009, pp. 1–8.
PV generator is always lower than the minimum power re- [8] M. Brenna, F. Foiadelli, M. Roscia, and D. Zaninelli, “Synergy between
renewable sources and electric vehicles for energy integration in distri-
quired for charging one EV, the charging time depends only bution systems,” in Proc. 15th IEEE Int. Conf. Harmonics Qual. Power
on the lower power considered for charging one EV. (ICHQP), 2012, pp. 865–869.
It can be concluded that the most important aspect to max- [9] Y. Zong et al., “Model predictive controller for active demand side
management with PV self-consumption in an intelligent building,” in
imize the self-consumption, useful to maximize the payback Proc. 3rd IEEE PES Int. Conf. Exhib. Innovative Smart Grid Technol.
time of the PV shelter, is to spread the EVs recharge during the (ISGT Eur.), 2012, pp. 1–8.
maximum production periods of PV system. This is possible [10] D. Carli, M. Ruggeri, M. Bottarelli, and M. Mazzer, “Grid-assisted
photovoltaic power supply to improve self-sustainability of ground-
only if there are no time constraints, such as in a situation of source heat pump systems,” in Proc. IEEE Int. Conf. Ind. Technol.
working periods. (ICIT), 2013, pp. 1579–1584.

Authorized licensed use limited to: UNIVERSITY TENAGA NASIONAL. Downloaded on July 15,2020 at 04:04:52 UTC from IEEE Xplore. Restrictions apply.
BRENNA et al.: URBAN SCALE PV CHARGING STATIONS FOR EVs 1241

[11] M. Brenna, F. Foiadelli, and D. Zaninelli, “Integration of recharging Federica Foiadelli (M’06) received the M.Sc. and
infrastructures for electric vehicles in urban transportation system,” Ph.D. degrees in electrical engineering from the
in Proc. 2nd IEEE Int. Energy Conf. Exhib. (ENERGYCON), 2012, Politecnico di Milano, Milano, Italy, in 2003 and
pp. 1060–1064. 2008, respectively.
[12] A. Dolara, G. C. Lazaroiu, S. Leva, and G. Manzolini, “Experimental Currently, she is an Assistant Professor with the
investigation of partial shading scenarios on PV (photovoltaic) modules,” Department of Energy, Politecnico di Milano. Her
Energy, vol. 55, no. 15, pp. 466–475, 2013. research interests include electric power systems and
[13] M. Brenna, A. Dolara, F. Foiadelli, G. C. Lazaroiu, and S. Leva, electric traction.
“Transient analysis of large scale PV systems with floating dc section,” Dr. Foiadelli is a member of CIFI (Italian Group
Energies, vol. 5, no. 10, pp. 3736–3752, 2012. of Engineering about Railways) and AEIT (Italian
[14] E. Ogliari, F. Grimaccia, S. Leva, and M. Mussetta, “Hybrid predictive Electric Association).
models for accurate forecasting in PV systems,” Energies, vol. 6, no. 4,
pp. 1918–1929, 2013.
[15] A. Ashtari, E. Bibeau, S. Shahidinejad, and T. Molinski, “PEV charging
profile prediction and analysis based on vehicle usage data,” IEEE Trans.
Smart Grid, vol. 3, no. 1, pp. 341–350, Mar. 2012.
[16] S. Shahidinejad, E. Bibeau, and S. Filizadeh, “Statistical development
of a duty cycle for plug-in vehicles in a North American urban setting
using fleet information,” IEEE Trans. Veh. Technol., vol. 59, no. 8,
pp. 3710–3719, Oct. 2010. Sonia Leva (M’01–SM’13) received the M.S. and
[17] (2013). Software Photovoltaic Geographical Information System Ph.D. degrees in electrical engineering from the
[Online]. Available: http://re.jrc.ec.europa.eu/pvgis Politecnico di Milano, Milano, Italy, in 1997 and
2001, respectively.
She is currently an Associate Professor of Elec-
trical Engineering with the Department of Energy,
Politecnico di Milano. Her research interests include
Morris Brenna (M’07) received the M.S. degree in electromagnetic compatibility, power quality, and re-
electrical engineering and the Ph.D. degree from the newable energy analysis and modeling. She is mem-
Politecnico di Milano, Milano, Italy, in 1999 and ber of the Italian Standard Authority (CEI/CT82)
2003, respectively. and of the IEEE Working Group “Distributed Re-
Currently, he is an Associate Professor with the sources: Modeling & Analysis”, as well as the Task Force on “Modeling and
Department of Energy, Politecnico di Milano. His Analysis of Electronically Coupled Distributed Resources”.
research interests include power electronics, dis-
tributed generation, electromagnetic compatibility,
and electric traction system.
Dr. Brenna is a member of Italian Electrical
Association (AEIT) and CIFI (Italian Group of
Engineering about Railways).
Michela Longo received the M.Sc. degree in in-
formation engineering and the Ph.D. degree in
mechatronics, information, innovative technologies,
and mathematical methods from the Department of
Alberto Dolara (S’09) received the M.S. and Ph.D. Engineering, University of Bergamo, Bergamo, Italy,
degrees in electrical engineering from the Politec- in 2009 and 2013, respectively.
nico di Milano, Milano, Italy, in 2005 and 2010, Currently, she is a Fellow Researcher with
respectively. the Department of Energy, Politecnico di Milano,
Currently, he is a Temporary Researcher with the Milano, Italy. Her research interests include electric
Department of Energy, Politecnico di Milano. His power systems, electric traction, and mechatronics.
research interests include traction systems, power Dr. Longo is a member of AEIT (Italian
quality, electromagnetic compatibility, and renew- Electric Association).
able sources.

Authorized licensed use limited to: UNIVERSITY TENAGA NASIONAL. Downloaded on July 15,2020 at 04:04:52 UTC from IEEE Xplore. Restrictions apply.