Impact of Fleet Electriﬁcation and Charging Infrastructure on

Free-Floating Car Sharing in Milan

Soﬁa Borgosano

1 a

, Alessandro Nocera

, Michela Longo

1 b

and Wahiba Yaici

3 c

Department of Energy, Politecnico di Milano, Via Lambruschini 6, Milano, Italy

Politecnico di Milano, Via Lambruschini 6, Milano, Italy

CanmetENERGY Research Centre, Natural Resources Canada, Ottawa, Canada

ﬁ

Keywords:

Electric Vehicles, Electric Car Sharing, Public Transportation, Charging Methods.

Abstract:

The automotive industry’s transition toward sustainability has prioritized Electric Vehicles (EVs) due to their

potential to reduce pollution and improve energy efﬁciency. This evolution is particularly critical in urban

contexts such as Milan, where free-ﬂoating car sharing services present unique challenges and opportunities

for electriﬁcation. The integration of EVs into car sharing ﬂeets demands careful consideration of battery

autonomy, charging times, and the distribution of charging infrastructure to meet high vehicle utilization rates.

This study evaluates the feasibility of transitioning Milan’s internal combustion car-sharing ﬂeet to an electric

model, analyzing technical and operational challenges through a scenario-based simulation approach.

1 INTRODUCTION

Urbanization is rapidly reshaping global mobility. By

2050, an estimated 6.3 billion people will reside in

urban areas, posing signiﬁcant challenges for urban

mobility systems (Moss, 2012). Transportation net-

works, essential for the movement of people and

goods within and between cities, will need to evolve

to address growing demand, congestion, and environ-

mental concerns (Colombo et al., 2023). This con-

text underscores the necessity of rethinking mobil-

ity strategies to achieve more efﬁcient and sustain-

able urban transport systems. Traditional car sharing

has gained prominence as an effective solution to re-

duce private car ownership, optimize vehicle utiliza-

tion, and reduce the environmental impact of urban

transportation (Hensher, 2018) (Weibin et al., 2018).

Car sharing is experiencing rapid global growth, with

user bases and ﬂeets expanding signiﬁcantly. For ex-

ample, from 2022 to 2027, the number of car shar-

ing users is expected to increase at a Compound An-

nual Growth Rate (CAGR) of 16. 9% (Cederqvist,

2023). However, reliance on Internal Combustion En-

gine (ICE) vehicles in many traditional car sharing

services limits their overall environmental beneﬁts.

https://orcid.org/0009-0005-6334-4630

https://orcid.org/0000-0002-3780-4980

https://orcid.org/0000-0002-6142-9180

Although car sharing reduces the total number of ve-

hicles on the road, ICE-powered ﬂeets still contribute

to urban air pollution and greenhouse gas emissions.

To overcome these limitations, the industry is increas-

ingly turning to Electric Car Sharing (ECS), which

integrates the operational advantages of car sharing

with the environmental beneﬁts of battery electric ve-

hicles (BEVs) (Perboli et al., 2018). Electric car shar-

ing offers signiﬁcant potential to amplify the envi-

ronmental beneﬁts of shared mobility by incorporat-

ing zero-emission vehicles into ﬂeets. BEVs produce

no tailpipe emissions, reducing air pollution and con-

tributing to cleaner urban environments. In addition,

they support the transition to more sustainable energy

systems, particularly when charged using renewable

energy sources. As a result, ECS has become a grow-

ing focus in the shared mobility sector, with an in-

creasing number of operators adopting Electric Ve-

hicles (EVs). In 2019, 66% of the global car shar-

ing services included electric vehicles in their ﬂeets,

demonstrating the trend towards cleaner transporta-

tion solutions (Nicholas and Bernard, 2021). De-

spite its potential, the feasibility of ECS largely de-

pends on the availability, efﬁciency, and distribution

of charging infrastructure. Charging stations differ

in speed, with some enabling fast charging within

minutes, while others require several hours for a full

charge. This variability can be challenging for EV

Borgosano, S., Nocera, A., Longo, M. and Yaici, W.

Impact of Fleet Electriﬁcation and Charging Infrastructure on Free-Floating Car Sharing in Milan.

DOI: 10.5220/0013360300003941

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 11th International Conference on Vehicle Technology and Intelligent Transport Systems (VEHITS 2025), pages 537-542

ISBN: 978-989-758-745-0; ISSN: 2184-495X

537

users, who must plan charging around station avail-

ability while accounting for potential queuing times

(Rauf et al., 2023).

The effectiveness of electric car-sharing depends

on overcoming critical challenges such as charging

point accessibility, system interoperability, and com-

patibility with vehicle usage patterns (Liao and Cor-

reia, 2021). Various studies have proposed models

to simulate vehicle charging and assess its impact on

the power grid (Hammerschmitt et al., 2024), as well

as simulate energy consumption (Genikomsakis and

Mitrentsis, 2017; Gerossier et al., 2019). However,

for vehicles used in car-sharing, predicting routes and

driving styles is difﬁcult due to the fact that they

can be rented by different individuals for a variety

of purposes. This variability signiﬁcantly impacts the

charging needs.

This study aims to tackle a critical issue for the

city of Milan, focusing on facilitating the widespread

adoption and effectiveness of an all-electric car-

sharing solution. It achieves this by analyzing real-

world demand and evaluating various charging tech-

nologies, such as conventional charging stations,

wireless charging systems, and battery swapping so-

lutions. The study assesses whether transitioning the

ﬂeet to EVs can be sustainable and evaluates which

charging methods are best suited to meet the system’s

needs.

2 METHODOLOGIES

This section outlines the methodologies adopted to

simulate the feasibility of transitioning a car shar-

ing service from an ICE ﬂeet to a BEV ﬂeet, taking

into account the compatibility with the autonomy of

the BEV and the distribution and performance of the

charging infrastructure.

2.1 Scope and Simulation Algorithm

The simulation aims to verify whether the car shar-

ing demand observed with ICE vehicles can be effec-

tively met using BEVs. The system must ensure that

each BEV chosen by a user has sufﬁcient energy to

complete the rental. Modeling car sharing demand is

particularly complex due to the interdependence be-

tween vehicle availability and the number of trips, es-

pecially in free-ﬂoating services where vehicles can

be left in any parking space within the service area.

This creates uncertainty regarding vehicle availabil-

ity for subsequent users. To address this, the simu-

lation focuses on representing vehicle availability at

a local level and tracking individual trips with high

spatial and temporal resolution. In addition to ana-

lyzing origin-destination ﬂows, it also estimates en-

ergy consumption and kilometers traveled per rental.

Since two rentals with the same start and end points

may involve different routes, statistical models of mo-

bility demand would be imprecise for calculating en-

ergy consumption. Therefore, a data-driven approach

is employed, relying on existing usage data. Despite

its limitations, this approach aligns with the goal of

comparing BEV and charging infrastructure scenar-

ios, rather than planning operational details of the

car sharing service. Key simplifying assumptions in-

clude:

• Users must recharge BEVs only when the State

of Charge (SOC) falls below a scenario-deﬁned

threshold, potentially requiring them to detour to

the nearest charging station.

• No staff assistance for vehicle recharging is con-

sidered.

• The existing charging infrastructure in Milan is

assumed to have uniform characteristics and com-

patibility with all BEVs in the scenarios.

• BEV range, charging times, and management are

not assumed to impact the demand trends ob-

served with ICE vehicles.

The simulation comprises the following steps:

1. Energy Consumption Calculation (E): The energy

consumption for each trip was calculated based

on the real consumption of the diesel car on that

speciﬁc route, using Equation 1. In this equation,

G represents the liters of gasoline used per trip,

gas

denotes the fuel consumption rate of an inter-

nal combustion engine (ICE) vehicle, expressed in

liters per kilometer (L/km) and c

represents the

speciﬁc energy consumption of the selected EV.

E =

gas

· c

[kW h]

(1)

2. SOC Calculation: The initial and ﬁnal SOC for

each trip are computed. If the SOC after a trip

falls below the threshold, recharging is simulated.

The SOC for the next rental depends on the energy

recharged between rentals.

3. Charging Management: Charging is simulated to

stop either when the SOC reaches 100% or a

minimum of 20%. Charging during a trip as-

sumes sufﬁcient time to reach a charging station

and continue the journey. The percentage of bat-

tery recharge (SOC

rec

) over a time interval (∆t)

depends on the power provided by the charging

method (P) and the battery capacity of the se-

lected vehicle (C

batt

) as reported in Eq. 2.

VEHITS 2025 - 11th International Conference on Vehicle Technology and Intelligent Transport Systems

538

SOC

rec

∆t · P

batt

· 100

(2)

Figure 1 visually represents the algorithmic ﬂow

for managing the SOC in an electric car sharing sce-

nario.

Figure 1: Recharging algorithm.

It outlines how vehicle energy levels are calcu-

lated and adjusted throughout a rental process, as well

as during transitions between rentals. The starting

point of the algorithm calculates the SOC at the begin-

ning of a rental, then based on the kilometers driven

and vehicle consumption rate, the energy used during

the rental is determined. The ﬁnal SOC is updated

by subtracting the consumed energy from the initial

SOC. If the SOC falls below a speciﬁed threshold,

the algorithm incorporates a recharging event, either

mid-rental or at the end of the trip. When a recharge

occurs, the SOC is adjusted to reﬂect the additional

energy gained, constrained by battery capacity. If the

ﬁnal SOC of one rental is sufﬁcient for the next user,

the car is made available; otherwise, it is assumed to

undergo recharging.

2.2 Performance Metrics

The algorithm is applied to various scenarios, each us-

ing a single type of BEV and charging infrastructure.

The performance of each scenario is assessed using

several key metrics that provide insight into the oper-

ational feasibility and efﬁciency of transitioning to a

BEV-based car sharing system.

One important metric is the number of unfeasible

rentals. These represent instances where a rental can-

not be completed because the vehicle’s battery charge,

or SOC, is insufﬁcient to meet the energy demands

of the trip. Such cases highlight the limitations of

BEV autonomy under speciﬁc conditions or charg-

ing infrastructure availability. Another critical mea-

sure is charging during rentals. This metric reﬂects

how often users would need to interrupt their trips to

recharge the vehicle mid-journey. It provides an indi-

cation of the practicality of the BEV ﬂeet, especially

in scenarios with longer rental distances or sparse

charging infrastructure. Similarly, charging at rental

end is evaluated, referring to the number of rentals

that require a recharge at the conclusion of the trip

due to the SOC falling below the minimum threshold.

This metric captures the impact of low battery levels

on subsequent vehicle availability for the next user.

The analysis also considers the average distance to

charging points, which represents the typical detour a

user would need to make to reach the nearest charging

station when recharging is necessary. A longer detour

can increase inconvenience for users and potentially

deter them from adopting the service. Finally, the

number of feasible rentals is examined. This metric

indicates how many rentals can be completed without

the need for mid-trip recharging or disruptions due

to SOC limitations. A higher proportion of feasible

rentals suggests better alignment between BEV capa-

bilities, user demand, and the charging infrastructure.

These metrics collectively provide a comprehen-

sive evaluation of the system’s performance, helping

to identify strengths, weaknesses, and areas for im-

provement in different scenarios.

3 CASE STUDY

The dataset was compiled from car-sharing records,

focusing on ICE vehicles. Vehicle availability was

tracked at regular intervals, capturing essential details

such as location, timestamp, and fuel level. Although

the data does not allow for full route reconstruction, it

provides valuable insights into mobility patterns by

identifying trip start and end points, duration, and

fuel consumption. This information serves as a basis

for analyzing vehicle utilization and operational efﬁ-

ciency. Trip duration was inferred by examining the

time elapsed between consecutive stops for each vehi-

cle. However, only key trip attributes—such as depar-

ture and arrival locations, travel time, and fuel level

variations—are available, while the exact routes taken

remain unknown. The subsequent analysis focuses on

Impact of Fleet Electriﬁcation and Charging Infrastructure on Free-Floating Car Sharing in Milan

539

the spatial distribution of stops, vehicle availability

trends, trip and stop durations, refueling patterns, and

fuel consumption.

Table 1 outlines the vehicle categories considered

in this study, organized by battery capacity and en-

ergy consumption. It includes several types of small

vehicles, emphasizing their versatility and appropri-

ateness for urban settings.

Table 1: Vehicles characteristics.

Category C

batt

cons

[kWh] [kWh/km]

Small A 17.6 0.175

Small B 42 0.146

Small C 10.3 0.075

Medium 57.5 0.143

Large 90 0.208

The analysis considers three main charging meth-

ods, with varying power capacities, commonly used

in Milan for both quick and fast charging. The lo-

cation of all the charging infrastructure has been as-

sessed based on the existing positions of public charg-

ers, as shown in Figure 2 (Electromaps, 2024).

Figure 2: Actual distribution of charging station in Milan.

A total of 143 charging stations have been added

in the last six months, and 304 in the last 12 months.

This data provides valuable insight into the rapid pace

at which charging stations are being installed across

the city. There are currently 2,643 connectors, with

the majority being of type 2. The following charging

infrastructures has been analyzed:

• Conductive Charging: This method involves ﬁxed

charging points at speciﬁc locations throughout

all scenarios. The power capacities for these

points are 7 kW, 22 kW, and 110 kW, catering to

different charging speed needs.

• Wireless Charging: This technology allows for a

more convenient charging process, as it doesn’t

require physical connection to the vehicle. It is

assumed that the wireless system will be available

for vehicles with a SOC below 40%. The power

levels considered for wireless charging are 3 kW,

7 kW, and 11 kW.

• Battery Swap: This method resembles tradi-

tional refueling, where the vehicle’s battery is ex-

changed for a fully charged one in just a few min-

utes. For vehicles in category Big, it is assumed

that the infrastructure consists of automated Bat-

tery Swapping Stations, enabling a quick and

seamless transition. In contrast, for vehicles in

category Small C, the process requires operator

intervention to perform the battery swap.

3.1 Analysis of the Results

Figure 3 provides a comprehensive summary of the

simulation results across all scenarios, with data aver-

aged on a daily basis. Each row corresponds to a sce-

nario derived from a combination of a speciﬁc BEV

model and charging infrastructure, while the columns

contain key performance indicators used to evaluate

the scenario.

The study revealed signiﬁcant variability in the

feasibility and efﬁciency of electric carsharing trips

based on the combination of charging methods and

BEV models. Across all scenarios, the percent-

age of feasible trips—those completed without re-

quiring user behavior changes—ranged from 55% to

92%. Scenarios utilizing Battery Swap and high-

power conductive charging (110kW) performed the

best, with feasible trip percentages exceeding 90% in

most cases. Conversely, Inductive charging at 3kW

showed the poorest performance, with feasible trips

dropping to as low as 55% for smaller battery vehi-

cles, like the ForTwo. A key priority for improving

carsharing systems is the reduction of unfeasible trips,

as shown in Figure 4, which ranks the 32 scenarios

based on the Unfeasible Rentals metric. In all the

cases the percentage reminas below the 16% across

all scenarios with values dropping to less than 1% in

favorable scenarios (e.g., Battery Swap and 110kW

conductive charging). Battery Swap and Conductive

110kW consistently rank as the top-performing meth-

ods, recording unfeasible trip rates of approximately

1%, regardless of the vehicle type. However, in sub-

sequent scenarios, vehicles like the XEV YOYO and

the Tesla Model 3 emerge as strong performers. When

paired with Conductive 22kW, Inductive 11kW, or In-

ductive 7kW charging, these models maintain unfea-

sible trip rates of around 3%, showcasing their adapt-

ability to mid-tier charging solutions.

VEHITS 2025 - 11th International Conference on Vehicle Technology and Intelligent Transport Systems

540

Figure 3: KPI analysis for each vehicle.

Figure 4: Unfeasible trips scenario comparison.

While BSS and Conductive 110kW charging are

clearly superior in terms of performance, their imple-

mentation poses signiﬁcant challenges. These include

high costs, extensive infrastructure requirements, and

compatibility issues, especially with existing electric

vehicle models and networks. To maximize their po-

tential, careful planning, targeted investments, and a

phased deployment strategy will be critical. Mid-tier

solutions like Conductive 22kW and Inductive 11kW

provide a compelling compromise, especially for ve-

hicles with moderate to large batteries such as the

Tesla Model 3. These options strike a balance be-

tween performance and practicality, delivering rela-

tively low unfeasible trip rates while being more af-

fordable and easier to deploy at scale. This rein-

forces the need for a tiered approach to charging in-

frastructure, where high-power solutions are reserved

for high-demand locations, and mid-power solutions

are implemented more broadly to ensure accessibil-

ity and efﬁciency. Lastly, it is important to highlight

that the average distance between end-of-rental points

requiring charging and the nearest charging station re-

mains relatively stable across both the days analyzed

and the simulated charging scenarios. These distances

are well within the 500-meter threshold that users are

typically willing to walk to access a vehicle (Her-

rmann et al., 2014).

Impact of Fleet Electriﬁcation and Charging Infrastructure on Free-Floating Car Sharing in Milan

541

4 CONCLUSIONS

This study analyzed the feasibility of transitioning

Milan’s free-ﬂoating car-sharing ﬂeet from ICEVs to

BEVs. Through a scenario-based simulation, it con-

sidered the interplay between vehicle energy auton-

omy, charging infrastructure distribution, and user de-

mand.

The results demonstrate that the integration of

BEVs into car-sharing services is achievable, pro-

vided that charging infrastructure is strategically

planned. Battery swapping and high-power conduc-

tive charging (110 kW) emerged as the most effec-

tive solutions for ensuring high operational perfor-

mance, with feasible rental percentages exceeding

90%. However, these solutions also face signiﬁcant

implementation challenges, including infrastructure

costs and compatibility issues.

Mid-tier options, such as 22 kW conductive and

11 kW inductive charging, present a balanced com-

promise, offering acceptable performance with lower

costs and scalability. Moreover, the study under-

scores the importance of maintaining short distances

between rental endpoints requiring charging and the

nearest charging stations, ensuring alignment with

user convenience thresholds.

Ultimately, the ﬁndings highlight the need for a

tiered and phased deployment strategy for charging

infrastructure to support the successful electriﬁcation

of car-sharing ﬂeets. This approach can maximize en-

vironmental beneﬁts while ensuring operational feasi-

bility, positioning Milan as a model city for sustain-

able urban mobility.

ACKNOWLEDGEMENTS

This study was carried out within the MOST –

Sustainable Mobility Center and received funding

from the European Union Next-GenerationEU (PI-

ANO NAZIONALE DI RIPRESA E RESILIENZA

(PNRR) – MISSIONE 4 COMPONENTE 2, IN-

VESTIMENTO 1.4 – D.D. 1033 17/06/2022,

CN00000023). This manuscript reﬂects only the

authors’ views and opinions, neither the European

Union nor the European Commission can be consid-

ered responsible for them.

REFERENCES

Cederqvist, M. (2023). The carsharing telematics market.

Technical report, Berg Insight AB.

Colombo, C. G., Borghetti, F., Longo, M., and Foiadelli,

F. (2023). Electriﬁcation of motorway network: A

methodological approach to deﬁne location of charg-

ing infrastructure for ev. Sustainability, 15(23).

Electromaps (2024). https://www.electromaps.com/it.

Genikomsakis, K. N. and Mitrentsis, G. (2017). A computa-

tionally efﬁcient simulation model for estimating en-

ergy consumption of electric vehicles in the context of

route planning applications. Transportation Research

Part D: Transport and Environment, 50:98–118.

Gerossier, A., Girard, R., and Kariniotakis, G. (2019).

Modeling and forecasting electric vehicle consump-

tion proﬁles. Energies, 12(7).

Hammerschmitt, B. K., Unsihuay-Vila, C., Sausen, J. P.,

Capeletti, M. B., Aoki, A. R., Teixeira, M. D., Bar-

riquello, C. H., and Abaide, A. d. R. (2024). Adaptive

charging simulation model for different electric vehi-

cles and mobility patterns. Energies, 17(16).

Hensher, D. A. (2018). Tackling road congestion – what

might it look like in the future under a collaborative

and connected mobility model? Transport Policy,

66:A1–A8.

Herrmann, S., Schulte, F., and Voß, S. (2014). Increas-

ing acceptance of free-ﬂoating car sharing systems us-

ing smart relocation strategies: A survey based study

of car2go hamburg. In Gonz

alez-Ram

ırez, R. G.,

Schulte, F., Voß, S., and Ceroni D

ıaz, J. A., edi-

tors, Computational Logistics, pages 151–162, Cham.

Springer International Publishing.

Liao, F. and Correia, G. (2021). Electric carsharing and

micromobility: A literature review on their usage pat-

tern, demand, and potential impacts. International

Journal of Sustainable Transportation, 16(3):269–

286.

Moss, M. L. (2012). Urban mobility in the 21st century.

Technical report, NYU Rudin Center for Transporta-

tion Policy.

Nicholas, M. and Bernard, M. R. (2021). Success factors

for electric carsharing. Technical report, INTERNA-

TIONAL COUNCIL ON CLEAN TRANSPORTA-

TION.

Perboli, G., Ferrero, F., Musso, S., and Vesco, A. (2018).

Business models and tariff simulation in car-sharing

services. Transportation Research Part A: Policy and

Practice, 115:32–48.

Rauf, H., Zehra Zaidi, S. S., Naveed, H., Mehmood, D.,

Jabeen, F., and Malik, A. W. (2023). Urbanevsim:

Open-source electric vehicle mobility and charging

simulation platform. In 2023 International Confer-

ence on Frontiers of Information Technology (FIT),

pages 79–84.

Weibin, Z., Yuhang, S., Yong, Q., Qianmu, L., and Minglei,

S. (2018). Traveler behavior analysis based on car2go

sharing operation data. In nternational Conference

on Transportation and Development 2018: Planning,

Sustainability, and Infrastructure Systems.

VEHITS 2025 - 11th International Conference on Vehicle Technology and Intelligent Transport Systems

542