THE MERGE PROJECT

Impacts of Electric Vehicles on the Distribution System Steady-state Operation

F. J. Soares, P. N. Pereira Barbeiro, C. Gouveia, P. M. Rocha Almeida, C. Moreira

and J. A. Peças Lopes

INESC TEC – INESC Technology and Science (formerly INESC Porto) and FEUP – Faculty of Engineering,

University of Porto, Campus da FEUP, Rua Dr. Roberto Frias, 378, 4200 – 465, Porto, Portugal.

Keywords: Charging Strategies, Critical Mass, Distribution Grid, Electric Vehicle, Steady-state Operation.

Abstract: This paper describes the main results of the MERGE project relative to Electric Vehicle (EV) charging

strategies and the impacts of EV integration on the steady-state grid operation. MERGE is a €4.5m,

collaborative research project supported by the European Commission’s Seventh Framework Programme

(FP7). The consortium includes utilities, regulators, commercial organisations and universities with interests

in the power generation, automotive, electronic commerce and hybrid and electric vehicle sectors across the

entire European Union (EU). One of the MERGE project missions is to evaluate the impacts that EV will

have on EU electric power systems, exploring EV and SmartGrid/MicroGrid simultaneous deployment,

together with renewable energy increase, to achieve CO

emission reduction through the identification of

enabling technologies and advanced control approaches. The work presented proposes three charging

strategies, dumb charging, multiple price tariffs and smart charging, and uses EV integration scenarios of

adherence to these charging schemes. The resulting scenarios are tested using an algorithm coded with

Python and using PSS/E, created within the MERGE framework to study EU grids steady-state behaviour.

Additionally, the critical mass of EV adherence to smart charging schemes that brings positive impacts to

the distribution grids operation was also evaluated.

1 INTRODUCTION

The changes that the actual electric power systems

are facing, namely in what regards renewables and

Electric Vehicles (EV) integration, will require

system operators to develop new network planning

and operation strategies in order to deal with the new

challenges arising from their large scale deployment.

In what regards networks planning and operation,

especially at the distribution level, the integration of

EV should be seen not only as a challenge, but also

as big opportunity to exploit the power systems’

infrastructures in a more effective manner and to

safely integrate larger quantities of renewables in the

systems.

The large scale integration of EV is very likely to

provoke several impacts in the power systems,

namely at the distribution level, like changes in the

branches loading, voltage profiles and load

diagrams.

Since EV are expected to be plugged-in in the

distribution systems, namely in Low Voltage (LV)

and Medium Voltage (MV) grids, these type of

networks are the ones where the EV charging

impacts will be strongly noticed. Congestion

problems are expected in already heavily loaded

grids, while in radial networks voltage limits

violations are likely to appear more frequently. The

changes in the energy losses is also a matter of great

concern, since the increase in the energy demand

owed to EV charging will probably make their value

rise considerably. The negative impacts referred are

more prone to appear if uncontrolled EV charging

strategies are used. In addition, the non-

controllability of the EV charging will also impact

negatively the profit that the EV

Supplier/Aggregators (EVSA) might achieve from

the markets negotiations, as they will not have

flexibility to shift the EV load towards the lower

demand periods, being thus incapable of profiting

from lower energy prices (Lopes et al., 2011).

On the other hand, the possibility of controlling

the EV charging will be of great benefit for both

EVSA and Distribution System Operator (DSO).

Soares F., Barbeiro P., Gouveia C., Almeida P., Moreira C. and Lopes J..

THE MERGE PROJECT - Impacts of Electric Vehicles on the Distribution System Steady-state Operation.

DOI: 10.5220/0003953600910100

In Proceedings of the 1st International Conference on Smart Grids and Green IT Systems (SMARTGREENS-2012), pages 91-100

ISBN: 978-989-8565-09-9

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

The EVSA will have the possibility of exploit the

EV flexibility for charging, namely the EV that are

parked during large periods of time overnight, thus

profiting from lower energy prices. Under these

circumstances, the EV charging management

performed by the EVSA will naturally shift a

significant amount of the EV load from the peak

hours towards lower demand periods, contributing to

improve the network operating conditions, to reduce

the energy losses and to diminish the DSO need to

invest in network reinforcements (Lopes et al.,

2011).

Nevertheless, in order to develop adequate

strategies to control the EV charging, it is first

required to evaluate the impacts that this new

elements will provoke in the power systems’

operation.

The project MERGE (Mobile Energy Resources

in Grids of Electricity) was created in order to seek

for solutions for the aforementioned problems. Its

mission is to evaluate the impacts that electric

vehicles will have on the European Union (EU)

electric power systems, in what regards planning,

operation and market functioning. The focus is

placed on EV and SmartGrid/MicroGrid

simultaneous deployment, together with renewable

energy increase, leading to CO

emission reduction

through the identification of enabling technologies

and advanced control approaches (Bower et al.,

2011).

One of the tasks of the MERGE project

involved the development of an evaluation suite

composed of several simulations tools capable of

identifying the impacts that the EV integration will

provoke in the electric power systems. The

assessment has been performed by analysing several

EU real networks using future EV integration

scenarios for 2020 and 2030.

This paper presents the results obtained with one

of the tools of the referred evaluation suite, which

was developed to perform steady-state simulations

in distribution networks (Rosa et al., 2011). The tool

incorporates several EV models, allowing estimating

the EV charging impacts in a given network, during

a week period, when different charging strategies are

adopted.

Five real MV networks were used as case

studies. For each network it was calculated the

maximum number of EV that can be safely

integrated, the changes provoked by EV in the

voltage profiles, lines loading, energy losses and in

the load diagrams. A comparison between the results

obtained for the maximum number of EV that can be

safely integrated in the studied MV networks and the

foreseen EV integration scenarios for 2020 and 2030

was also performed. Additionally, the critical mass

(percentage) of EV owners that need to adhere to

controlled charging schemes in order to enable the

safe operation of the networks was also analysed.

A brief description of the methodology followed

in the simulations, as well as the MV networks used

as case studies, is presented in the next section.

2 METHODOLOGY AND CASE

STUDIES

The methodology followed during the steady-state

simulations can be divided in two parts: one to

quantify the maximum EV number that can be safety

integrated in a given network (section 2.1) and

another to analyse the critical mass (section 2.3).

2.1 Methodology to Quantify the

Maximum Number of EV that can

be Safely Integrated

The quantification of the maximum number of EV

that can be safety integrated in the distribution

networks analysed was performed for each network

considering three charging strategies: dumb

charging, multiple tariff and smart charging.

In the dumb charging approach it is assumed that

EV owners are completely free to connect and

charge their vehicles whenever they want. The

charging starts automatically in the moment when

EV plug-in and only stops when the battery is fully

charged or when the EV is disconnected from the

grid by its owner. This approach should be described

as a no control strategy but it is particularly

important as it provides a measure for the

assessment of the efficacy of the other management

procedures.

The dual tariff intends to simulate a situation

where electricity is cheaper during some specific

hours of the day. For the simulations performed

within the MERGE project, the cheaper period was

assumed to be enclosed between 1h and 7h.

The smart charging strategy envisions an active

management system, where the EV aggregating

entities are capable of managing the EV charging

according to the market negotiations, always taking

into account the EV owners’ requests. In addition, it

is assumed that the DSO is capable of monitoring all

the elements connected to the grid and its state,

having also the capability of interfering in the EV

charging schedules in order to solve eventual

SMARTGREENS2012-1stInternationalConferenceonSmartGridsandGreenITSystems

technical problems that might appear in the grid.

This type of management provides the most efficient

usage of the resources available at each moment,

enabling congestion prevention and voltage control.

During the simulations performed, the following

EV charging scenarios were considered:

 All EV in dumb charging mode;

 All EV in multiple tariff mode;

 All EV in smart charging mode.

The implementation of the different EV charging

strategies in the simulation platform is thoroughly

described in (Rosa et al., 2011).

The simulations for each network and for each

charging strategy starts assuming an EV integration

that corresponds to a replacement of 1% of the

conventional vehicles fleet by EV (see Table 1). The

impacts of the referred EV integration level are then

evaluated. If no violations of the network

components’ technical limits were detected, the EV

integration percentage is increased by 1% and a new

evaluation of the network operating conditions is

performed in order to detect if any violation

occurred. This process is repeated until one of the

following conditions is verified: violation of the

voltage limits specified and/or branch overloading.

The maximum percentages of EV that can be

safely integrated in the MV networks analysed, for

each charging strategy, is recorded in the end of the

simulations.

In all the case studies, it was assumed the

existence of one fast charging station per network.

As it will be further demonstrated through the results

presented in section 3, fast charging stations have a

considerable impact in branches’ congestion levels

and in the voltage profiles. For this reason, the

network bus to which the fast charging station was

assumed to be connected, in each case study, was

selected among the network buses with the highest

voltage values. The EV resort to the fast charging

stations when, during a journey, their battery SOC is

not enough to complete the journey.

For the EV charging scenario that considers the

multiple tariff, it was assumed that the period of

lower energy prices is between 1h and 7h, every day

of the week.

2.2 MV Networks used as Case Studies

As referred previously, a set of five real MV

networks were used as case studies. These networks

were carefully chosen in order to evaluate systems

with different characteristics, like their topology

(rural or urban) and their type of consumers

(industrial, commercial or residential). In Table 1 it

is presented the most relevant characteristics of the

tested networks. A detailed description of these

networks can be found in (Sánchez et al., 2010).

Table 1: Networks’ characteristics.

Besides the variables presented in Table 1, load

diagrams may also have a significant influence in

the results obtained. The load profiles of each

network, during a typical week, are presented in

Figure 1. As it can be seen, the load diagrams of the

networks tested vary significantly. This variation

may be explained by the different climate, social-

cultural and economic conditions of each area.

Despite de differences, well defined daily patterns

are easily identified for all the networks except the

rural, where the load consumption along the week is

more irregular.

The identification of the daily load patterns is

very important for the implementation of the dual

tariff and the smart charging. For the former, the

daily load patterns can be used to define the period

during which the electricity price is lower, seeking

to incentivize the EV owners to shift their EV

existence charging to the lower demand periods. For

the latter, the knowledge of daily load patterns will

considerably ease the EVSA tasks in what regards

the prediction of the EV load, since it will allow

them to reduce forecasting errors and schedule the

EV charging with higher accuracy.

100

125

150

175

200

225

250

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

Power (MW)

MV Network 1 MV Network 2 MV Network 3 MV Network 4 MV Network 5

Figure 1: MV networks load profiles for a typical week.

THEMERGEPROJECT-ImpactsofElectricVehiclesontheDistributionSystemSteady-stateOperation

2.3 Methodology and Case Study used

in the Critical Mass Simulation

For the critical mass study, the first step of the

procedure followed consisted on the consideration of

a fixed EV integration percentage, of which one half

of the EV were assumed to be dumb charging

adherents and the other half multiple tariff

adherents. Then, if problems were not detected, the

EV integration percentage was increased by 10%,

assuming the same proportion of dumb charging and

multiple tariff adherents (50% of each). This

procedure was repeated until a problem in the

network was detected (either a voltage lower limit

violation or a branch overloading).

After detecting a technical problem, the second

step of the procedure consisted on iteratively

increasing the percentage of smart charging

adherents, in steps of 5%, while the dumb charging

and multiple tariff adherents percentage was

decreased accordingly, as explained in Figure 2.

First Iteration

• % of dumb charging adherents Æ 50%

• % of multiple tariff adherents Æ 50%

• % of smart charging adherents Æ 0%

Second Iteration

• % of dumb charging adherents Æ 47,5%

• % of multiple tariff adherents Æ 47,5%

• % of smart charging adherents Æ 5%

Third Iteration

• % of dumb charging adherents Æ 45%

• % of multiple tariff adherents Æ 45%

• % of smart charging adherents Æ 10%

Figure 2: Flowchart of the steps followed for critical mass

estimation

The second step of the procedure was repeated

until the technical problems previously identified

were solved. In the end of the procedure, the

percentage of smart charging adherents that allowed

solving the problems detected (the critical mass of

smart charging adherents) was recorded.

The network used as case study for the critical

mass simulation was the MV network 1.

3 RESULTS AND ANALYSIS

3.1 Changes in Load Diagrams

The results presented in this subsection are referred

to the maximum percentages of EV that can be

safely integrated in the MV networks analysed, for

each charging strategy, as well as to the changes in

the weekly load diagrams verified. Due to space

restrictions, only charts for one of the networks

analysed will be presented (MV Network 1).

However, the results presented in this section can be

generalized for the remaining networks.

The maximum allowable EV integration

percentages in the MV Network 1 are depicted in

Figure 4. The percentages are relative to the total

number of conventional vehicles enclosed in the

geographical area covered by this network, which was,

in this case, 21135 vehicles. For the dumb charging,

multiple tariff and smart charging, the number of EV

that can be safely integrated in this network is,

respectively, 5072, 7186 and 11836, which correspond

to the percentages indicated in Figure 3.

0% 10% 20% 30% 40% 50% 60%

Dumb Charging

Multiple Tariff

Smart Charging

24%

34%

56%

Maximum EV Percentage

Figure 3: Maximum EV integration percentage in the MV

Network 1.

It is important to note that even when the smart

charging is considered, the continuous load growth

due to the increase of the EV integration provokes,

at a certain moment, at least one technical constraint

violation. In the case of the MV Network 1, the first

violation to occur was a branch overloading. The

technical violations detected in the all the networks

are presented in Table 4 and Table 5.

The EV power demand profile for the entire

week, in MW Network 1, for the three charging

strategies, is shown in Figure 4. When considering

the dumb charging strategy, the EV tend to charge

essentially at the end of the day, which is the time

period when people arrive home from work. In the

multiple tariff strategy, the EV owners tend to

charge their vehicles between 1h and 7h, which is

the period of time when the energy prices are

assumed to be lower. With the smart charging, the

EV are charged mostly during the night, as this is the

period when the EV availability is higher and the

demand is lower. These two facts combined, make it

possible to integrate a large number of EV in this

grid without causing any technical constraints

violations.

SMARTGREENS2012-1stInternationalConferenceonSmartGridsandGreenITSystems

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

Power (MW)

Dumb Charging (25% EV) Smart Charging (57% EV) Multiple Tariff (35% EV)

Figure 4: EV load demand profiles in the MV Network 1.

Adding the EV load depicted in Figure 4 to the

conventional load of this network, makes it possible

to compute the total load diagrams for the three

charging strategies addressed, as presented in Figure

5. The load diagram for the scenario without EV

reveals a relatively constant pattern during the week

and the weekend days. A significantly large valley

period is notorious during the nights, while during

the days two small peaks are easily identifiable, one

occurring during lunch time and the other during the

evening.

100

120

140

160

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

Power (MW)

Dumb Charging (25% EV) Without EV

100

120

140

160

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

Power (MW)

Multiple Tariff (35% EV) Without EV

100

120

140

160

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

Power (MW)

Smart Charging (57% EV) Without EV

Figure 5: Load profiles without and with EV (MV

Network 1).

In the scenario without EV, this network has a

peak load of 128.5 MW, which is incremented to

135.6 MW using the dumb charging, to 133.9 MW

using the multiple tariff and to 132.1 MW using the

smart charging. The latter can be considered an

outstanding achievement, since the peak load only

increased 3.6 MW with an EV integration of 57%,

representing ca. 12047 EV.

It is interesting to notice that the EV charging,

for the dumb charging and the multiple tariff,

provokes changes in the hour at which the networks’

peak load occurs. In the particular case of this

network, the peak load occurrence changes from 14h

to 19h of Thursday. For the smart charging, the hour

at which the peak load occurs remains unchanged.

In Table 2 is presented an overview of the

maximum EV integration percentage and the

correspondent absolute value of EV allowed in each

of the MV networks studied.

Table 2: Maximum EV allowed integration.

Dumb

Charging

Multiple

Tariff

Smart

Charging

MV Network 1

24%

(5072 EV)

34%

(7186 EV)

56%

(11836 EV)

MV Network 2

40%

(2081 EV)

57%

(2965 EV)

74%

(3850 EV)

MV Network 3

(2193 EV)

(4386 EV)

(8771 EV)

MV Network 4

28%

(6090 EV)

24%

(5220 EV)

42%

(9135 EV)

MV Network 5

10%

(3416 EV)

(1708 EV)

24%

(8197 EV)

From the results obtained, it can be observed that

the analysed systems can handle, up to a certain

level, the penetration of EV without concerns to the

networks’ infrastructures. However, it was verified

that the maximum number of EV that can be safely

integrated in the networks depends on the charging

schemes adopted by the EV owners. From the three

strategies analysed, smart charging yielded better

results in all the case studies addressed, as with it

was possible to reach higher EV integration levels

without violating the networks´ technical

restrictions, meaning that higher investments

deferral can be obtained. The dual tariff can be

classified as the second best strategy, as in three of

the five networks it attained better results than the

dumb charging.

The fact of the dumb charging yielding better

results than the multiple tariff in some of the

networks can be explained by the instantaneous

increase of the EV load verified around 1h when the

multiple tariff implemented. This occurs due to a

large number of multiple tariff adherents starting

their charging almost simultaneously. This load

increase might occur in specific locations of the grid,

where some grid components are already operating

very near their limits, provoking the occurrence of

technical violations.

THEMERGEPROJECT-ImpactsofElectricVehiclesontheDistributionSystemSteady-stateOperation

3.2 Feasibility of the Foreseen EV

Integration Scenario for 2020 and

2030

The purpose of this section is to evaluate the

feasibility of the foreseen EV integration scenarios

for 2020 and 2030, which were defined by the

MERGE partners in (Hasset et al., 2011).

In the referred deliverable of the project, three

possible EV penetration scenarios for the period

between 2010 and 2030 were defined. From the

three scenarios defined, the one recommended to be

used as reference by the MERGE partners was

scenario 2, as it is in-between a rather pessimistic

scenario and the most optimistic one. Following this

recommendation, and taking into account that the

networks analysed are Spanish, the values obtained

in scenario 2 for Spain were selected to be used as

comparison basis for the results presented in this

section.

The feasibility of the EV integration levels

defined for the years 2020 and 2030 was evaluated

as follows:

1) It was calculated the total number of EV that

are expected to be integrated in each of the

MV networks analysed, for 2020 and 2030.

These values were computed assuming that

the number of EV present in each network is

proportional to the number of conventional

vehicles enclosed in the networks’

geographical area.

2) Then, it was calculated the maximum number

of EV that can be safely integrated in the

analysed networks, with the different charging

strategies (as presented and discussed in

section 3.1).

3) Finally, it was performed a comparison

between the values obtained in steps 1) and 2).

The EV integration level forecasted is

assumed to be unfeasible if the values

obtained in step 1) are higher than those

obtained in step 2) and feasible otherwise.

All the results obtained are presented in Table 3.

The feasibility of the integration levels forecasted by

the MERGE partners, for each charging strategy, is

indicated by a cross.

As it can be seen, the only networks that are not

capable of coping with the EV integration levels

forecasted are the MV Network 3 and the MV

Network 5, for the year 2030, both due to branches’

overloading problems. While the MV Network 5

only presents problems when the dumb charging is

implemented, the MV Network 3 has some branches

overloaded with both dumb charging and multiple

tariff. The smart charging yields the best results,

since with it all the EV integration levels forecasted

are feasible.

Table 3: Feasibility of the foreseen EV integration

scenario for 2020 and 2030.

3.3 Impacts in the Voltage Profiles

Table 4 depicts, the voltage values obtained in the

worst bus of the networks analysed, when the

maximum allowable EV integration is reached. The

values presented are referred to the hour at which the

worst voltage conditions in the networks are

verified, which can be different from the hour of the

peak load.

Table 4: Voltage in the worst bus (p.u.).

Without

Dumb

Charging

Multiple

Tariff

Smart

Charging

Network 1

1.0238

(0% EV)

1.0235

(25% EV)

1.0228

(35% EV)

1.0234

(57% EV)

Network 2

0.9460

(0% EV)

0.9295

(41% EV)

0.9306

(58% EV)

0.9310

(75% EV)

Network 3

0.9721

(0% EV)

0.9715

(3% EV)

0.9721

(5% EV)

0.9704

(9% EV)

Network 4

0.9866

(0% EV)

0.9853

(28% EV)

0.9866

(24% EV)

0.9848

(43% EV)

Network 5

0.9722

(0% EV)

0.9705

(10% EV)

0.9722

(5% EV)

0.9712

(24% EV)

As it can be observed, with the exception of the

MV Network 2, the EV extra demand provokes

almost insignificant voltage drops when comparing

with the initial scenario (with no EV present in the

grids). It is important to recall that in MV networks

the R/X ratio is low, contrarily to LV networks, what

makes the impacts of the active power consumed by

EV less relevant regarding voltage drops. In

addition, as the majority of the MV networks studied

are from urban areas, they are more prone to

congestion problems than to undervoltage issues.

The voltage values attained for the MV Network

1, MV Network 3, MV Network 4 and MV Network

5 are within acceptable values, while for the MV

Network 2 they reach values near or even below the

minimum limit allowed (defined for these networks

as 0.93 p.u.).

From these results, it is possible to conclude that

the voltage lower limit is very likely the technical

SMARTGREENS2012-1stInternationalConferenceonSmartGridsandGreenITSystems

constraint that impedes a higher EV integration level

in the MV Network 2.

Although the voltage values regarding the use of

different charging strategies are presented in the

same table, for all the networks, it should be stressed

that they are referred to different scenarios of EV

integration. Thus, the only possible fact that can be

concluded from the values presented is that the

smart charging provides better results, as it is the

charging strategy that allows safely integrating a

larger number of EV in all the case studies

evaluated.

3.4 Impacts in the Branches

Congestion Levels

Differently to what was verified for the voltage

profiles, branches’ congestion levels were the most

critical aspect in the generality of studied networks,

with especially emphasis in the networks with urban

characteristics. Looking at Table 5, where the rating

percentage of the most congested branch of each

network is presented, it is possible to observe the

effects of the EV charging when the three different

charging methods are applied. The maximum rating

limit allowed was assumed to be 100%.

The results obtained show, in all the networks,

that the branches’ load levels considerably worsen

with the growth of the number of EV present in the

grids. In fact, branches overloading is the factor that

limits a further EV integration in the MV Network 1,

MV Network 3, MV Network 4 and MV Network 5.

The MV Network 2, besides having low voltage

problems, also presents branches’ overloading

issues.

Likewise to the voltage profiles, the rating

values presented in Table 5 for the different

networks are referred to different scenarios of EV

integration. Thus, the only possible fact that can be

concluded is that the smart charging provides better

results, as it is the charging strategy that allows

safely integrating a larger number of EV in all the

case studies evaluated. If it was considered a fixed

number of EV in the grids, the worst rating

percentage obtained with the smart charging would

be significantly lower than the value obtained with

the dumb charging and the multiple tariff.

The dumb charging strategy is the charging

scheme that accounts for the worst results in the MV

Network 1, MV Network 2 and MV Network 3,

while multiple tariff strategy accounts for the worst

results in the MV Network 4 and MV Network 5.

As referred previously, the worst results of the

multiple tariff obtained in the MV Network 4 and

MV Network 5, in comparison with the dumb

charging approach, might be explained by the

instantaneous increase of the EV load verified when

a large number of multiple tariff adherents start their

charging, almost simultaneously, in the beginning of

the lower electricity price period.

The location of the fast charging stations is also

a very important variable in what regards branches’

overloading, as the large amount of power absorbed

by these facilities might overload the branches

upstream. This problem has in fact occurred in the

MV Network 1 and MV Network 5, where the

branches overloading registered was due to EV

charging in the fast charging stations. For this

reason, it is advisable that the installation of a fast

charging station is always preceded by a detailed

impact study.

Table 5: Rating in the worst branch.

Without

Dumb

Charging

Multiple

Tariff

Smart

Charging

Network 1

96.5%

(0% EV)

100.0%

(25% EV)

101.6%

(35% EV)

100.4%

(57%

EV)

Network 2

84.8%

(0% EV)

100.6%

(41% EV)

101.1%

(58% EV)

100.4%

(75%

EV)

Network 3

97.9%

(0% EV)

101.7%

(3% EV)

101.0%

(5% EV)

101.3%

(9% EV)

Network 4

79.1%

(0% EV)

100.1%

(28% EV)

102.4%

(24% EV)

100.5%

(43%

EV)

Network 5

97.7%

(0% EV)

100.5%

(10% EV)

105.1%

(5% EV)

100.5%

(24%

EV)

3.5 Energy Losses

The weekly energy losses in the networks analysed,

for all the scenarios studied, are presented in Table

6. The first value presented in each cell is referred to

the absolute value of the losses, while the second is

relative to the ratio between the losses and the

overall energy consumption in the networks.

Table 6: Weekly energy losses (MWh) and Losses/Total

Energy (%).

Without

Dumb

Charging

Multiple

Tariff

Smart

Charging

Network 1

50.0

MWh

0.34%

(

0% EV

)

54.0 MWh

0.34%

(

25% EV

)

53.1

MWh

0.34%

(

35% EV

)

54.7

MWh

0.33%

(

57%

Network 2

45.5

MWh

2.17%

(

0% EV

)

58.0 MWh

2.39%

(

41% EV

)

63.1

MWh

2.49%

(

58% EV

)

66.7

MWh

2.54%

(

75%

Network 3

82.5

MWh

0.68%

(

0% EV

)

87.5 MWh

0.69%

(

3% EV

)

87.9

MWh

0.69 %

(

5% EV

)

92.4

MWh

0.68%

(

9% EV

)

Network 4

50.6

MWh

0.40%

(

0% EV

)

54.3 MWh

0.39%

(

28% EV

)

53.0

MWh

0.39%

(

24% EV

)

54.6

MWh

0.39%

(

43%

Network 5

465.9

MWh

1.72%

(

0% EV

)

483.1 MWh

1.75%

(

10% EV

)

471.9

MWh

1.73%

(

5% EV

)

497.3

MWh

1.77%

THEMERGEPROJECT-ImpactsofElectricVehiclesontheDistributionSystemSteady-stateOperation

A significant increase in the absolute value of

the weekly losses is easily identifiable when

comparing the scenarios with and without EV. As

the energy losses are directly proportional to the

square of the current, when the demand increases,

due to the EV charging, the current flowing along

the grids’ branches raises as well, provoking an

increase in the losses.

Although the absolute value of the energy losses

increases with the smart charging (due to a larger

EV integration), its relative value reveals that this

charging strategy is the one that yields best results in

the majority of the cases studied.

The adoption of the multiple tariff strategy could

also lead to some positive results. As it can be

observed, when comparing this strategy with the

dumb charging, it is possible to decrease losses

relative value in four of the analysed networks (MV

Network 1, MV Network 3, MV Network 4 and MV

Network 5), mainly due to the load valleys in the

load diagrams that occur between 1h and 7h. The

exception is the MV Network 2, where the valley

hours occur in the late afternoon, not coinciding with

the period when the majority of the multiple tariff

adherents charge their EV: between 1h and 7h.

Generally, the charging method that yields worst

results is the dumb charging, since it leads to the

occurrence of the highest peak loads, which,

expectably, lead to the higher increases in the energy

losses.

3.6 Critical Mass Analysis

The main goal of this study is to identify the

percentage of EV owners that need to adhere to the

smart charging in order to safely integrate a given

number of EV.

The first step of the methodology implemented,

as referred in section 2.3, consisted on the

consideration of an initial EV integration percentage,

of which one half of the EV were assumed to be

dumb charging adherents and the other half multiple

tariff adherents. Then, if problems were not detected

in the network, the EV integration percentage was

increased by 10 until a problem in the network was

detected. For the MV Network 1, used as test case,

the initial EV integration percentage assumed was of

10% and the first technical violation was detected

with a 30% EV integration.

The second step of the methodology consisted on

iteratively increasing the percentage of smart

charging adherents, in steps of 5%, while decreasing

the dumb charging and multiple tariff adherents

accordingly. This procedure was repeated until the

technical problems identified were solved. For the

case study under analysis, the percentage of smart

charging adherents that allowed solving the

problems detected – the critical mass – was of 45%.

The differences between both scenarios referred

have a direct influence on the EV load profiles, as

presented in Figure 6. In the first scenario (in blue),

the EV power consumption has two daily peaks: one

in the late afternoon (due to dumb charging

adherents) and other during the first hours of the

night (due to multiple tariff adherents). When the

value of the smart charging adherents is incremented

to 45%, a decrease in EV power during the late

afternoon peak can be noticed.

Monda

Tuesda

Wednesda

Thursda

Frida

Saturda

Sunda

Power (MW)

50% Dumb + 50% Multiple Tariff 27.5% Dumb + 27.5% Multiple Tariff + 45% Smart Charging

Figure 6: EV load demand profiles in the MV Network 1

(30% EV).

In Figure 7 are depicted the load diagrams for

both cases studied. The peak load in the scenario

with 45% of smart charging adherents slightly

decreases, in comparison with the scenario with 50%

dumb charging and 50% multiple tariff.

100

120

140

160

Monda

Tuesda

Wednesda

Thursda

Frida

Saturda

Sunda

Power (MW)

Without EV 50% Dumb + 50% Multiple Tariff 27.5% Dumb + 27.5% Multiple Tariff + 45% Smart Charging

Figure 7: Load profiles without and with EV (MV

Network 1, 30% EV).

Figure 8, Figure 9 and Figure 10 show,

respectively, the voltages in the worst bus, the rating

in the worst branch and the weekly energy losses for

both scenarios simulated. As it can be noticed, the

increase in the number of smart charging adherents

yields benefits in all the indexes analysed.

As it can be seen in Figure 9, when considering

30% of EV integration, with 50% dumb charging

and 50% multiple tariff, there are some branches

SMARTGREENS2012-1stInternationalConferenceonSmartGridsandGreenITSystems

1.0238

1.0237

1.0238

1.0236

1.0237

1.0238

1.0239

Without EV 50% Dumb + 50% Multiple

Tariff

27.5% Dumb + 27.5% Multiple

Tariff + 45% Smart Charging

Voltage (p.u.)

Figure 8: Voltage in the worst bus (30% EV).

96.5

101.6

98.4

93.0

94.0

95.0

96.0

97.0

98.0

99.0

100.0

101.0

102.0

Without EV 50% Dumb + 50% Multiple

Tariff

27.5% Dumb + 27.5% Multiple

Tariff + 45% Smart Charging

Rating (%)

Figure 9: Rating in the worst branch (30% EV).

already overloaded. The worst branch is 1.6% above

its maximum rated capacity. By incrementing the

share of smart charging adherents to 45% (critical

mass value), while decreasing both dumb and

multiple tariff adherents to 27.5 %, the worst branch

rating decreases to 98.4%, value within the allowed

limits.

50.0

53.5

52.9

0.34

0.28

0.29

0.3

0.31

0.32

0.33

0.34

0.35

0.36

Without EV 50% Dumb + 50% Multiple

Tariff

27.5% Dumb + 27.5% Multiple

Tariff + 45% Smart Charging

Losses / Total Energy (%)

Losses (MWh)

Figure 10: Weekly losses (30% EV).

A rather obvious assumption about the critical

mass is that its value is expected to increase as the

number of EV connected to the grid raises. In order

to prove it, a second scenario with a higher EV

integration (40%) was analysed. This EV integration

level leads to a considerable aggravation of the

branches congestion levels. The worst branch is ca.

10% above its maximum rated capacity, against the

1.6% verified in the previous case (with 30% of EV

integration). Under these conditions, the worst

branch rating can only be decreased to acceptable

values if the smart charging adherents’ percentage

reaches 60% (critical mass value). As expected, this

result proves that the critical mass increases as the

EV integration level rises.

4 CONCLUSIONS

By analysing the results obtained from the steady-

state simulations performed in the MERGE project,

it was possible to verify that the magnitude of the

EV impacts are influenced by several factors, like

the EV integration level, the EV owners’ behaviour,

mobility patterns, the networks’ load profiles and

technical characteristics, the number and location of

fast charging stations in the grid and the EV

charging modes, among others. These factors have

been carefully analysed, being possible to reach

some important conclusions.

The analysed systems can handle, up to a certain

level, the penetration of EV without concerns to the

networks’ infrastructures. However, it was verified

that the maximum number of EV that can be safely

integrated in the networks depends on the charging

schemes adopted by the EV owners. From the three

strategies analysed (dumb charging, dual tariff

charging and smart charging), smart charging

yielded better results in all the case studies

addressed, since it was possible to reach higher EV

integration levels without violating the networks´

technical restrictions.

In what regards the EV impacts in networks with

different topologies, some important conclusions

were also attained. Concerning urban networks, as

they are usually composed by short lines and are

subjected to high power demand levels, they are

very likely to face branch/transformer overloading

problems faster than voltage drop issues. The results

presented in this report prove this fact, as overload

problems were identified in all networks studied

with urban topologies. Differently from urban

networks, rural networks have usually long radial

lines, which provoke considerable voltage drops.

Thus, low voltage problems are expected in these

grids, namely in the buses farthest from the feeding

points. The results obtained prove this fact, as low

voltage problems were only detected in the rural

network analysed.

The extra power demanded by EV also provokes

several changes in the networks’ load diagrams,

which are more pronounced as the EV integration

level rises. Nevertheless, the analysis performed

THEMERGEPROJECT-ImpactsofElectricVehiclesontheDistributionSystemSteady-stateOperation

allows concluding that it is impossible to generalise

results in a rigorous manner, as the changes induced

in the load diagrams depend of a large number of

factors that are different from network to network.

The location of the fast charging stations should

be carefully analysed, as they might provoke severe

voltage violations or branches overloading, due to

the large amount of power that they may consume

when in full operation. In fact, the studies performed

have demonstrated that the overload problems

identified in two of the studied networks were likely

provoked by the power consumed in fast charging

stations.

As it happened with the load diagrams, the

simulations performed for the critical mass allow

concluding that it is impossible to generalise results

in a rigorous manner. From the analysis of the

results obtained, it is only possible to conclude that

the critical mass, besides being dependent of the

network considered, increases with the EV

integration level.

In what regards the feasibility of the forecasted

EV integration scenario for 2020 and 2030, it was

possible to conclude that, independently of the

charging strategy adopted, no relevant problems in

the MV networks are expected to occur until 2020.

Conversely, in 2030, several problems are expected

to arise, namely if dumb charging or dual tariff

approaches are adopted. However, as results

presented in section 3.2 show, the forecasted

problems may be entirely solved if the smart

charging is implemented on a large scale.

From the results obtained with the steady-state

analysis performed within the MERGE project, it is

clear that the path to safely integrate large quantities

of EV in distribution networks, without making

large investments in grid reinforcements, is to

implement mechanisms that allow managing the EV

charging not only taking into account their owners’

requests, but also the networks’ technical

restrictions. Nevertheless, it should be remarked that

the adherence to these controlled charging schemes

will ultimately be always a decision of the EV

owners. Thus, it is of utmost importance to timely

define and implement adequate incentives’ policies,

attractive enough to make EV owners willing to

participate in such controlled charging schemes.

ACKNOWLEDGEMENTS

This work was supported in part by Fundação para a

Ciência e Tecnologia under Grants

SFRH/BD/48491/2008 and SFRH/BD/47973/2008

and by the European Union within the framework of

the European Project MERGE – Mobile Energy

Resources in Grids of Electricity, Contract 241399

(7th Framework Programme).

REFERENCES

Bower, E. T., Lopes, J. A. P., Soares, F. J., Rua, D.,

Hatziargyriou, n., Strunz, K. & Ferdowsi, m. 2011.

Initial findings of ‘merge’ (mobile energy resources in

grids of electricity). JSAE EVTeC’11. Japan.

Hasset, B., Bower, E. & Alexander, M. 2011. Evaluation

of the impact that a progressive deployment of EV will

provoke on electricity demand, steady state operation,

market issues, generation schedules and on the volume

of carbon emissions Deliverable D3.2 of the European

Project MERGE.

Lopes, J. A. P., Soares, F. J. & Almeida, P. M. R. 2011.

Integration of Electric Vehicles in the Electric Power

System. Proceedings of the IEEE, 99, 168-183.

Rosa, M., Issicaba, D., Gil, N., Soares, F. J., Almeida, P.

M. R., Moreira, C., Ribeiro, P., Heleno, M., Ferreira,

R., Sumaili, J., Meirinhos, J., Seca, L., Lopes, J. A. P.,

Matos, M., Soultanis, N., Anestis, A., Karfopoulos, E.,

Mu, Y., Wu, J., Ekanayake, J. & Narayana, M. 2011.

Functional Specification for Tools to Assess Steady

State and Dynamic Behaviour Impacts, Impact on

Electricity Markets and Impact of High Penetration of

EV on the Reserve Levels. Deliverable D2.2 of the

European Project MERGE, .

Sánchez, C., Gonzalez, A., Rosa, M., Ferreira, R., Cabral,

P., Batista, F., Issicaba, D., Gil, N., Moreira, C.,

Ribeiro, P., Diaz-Guerra, B., Papadopoulos, P., Grau,

I., Voumvoulakis, E., Zountouridou, E., Karfopoulos,

E., Bourithi, Ferdowsi, M. & Abbasi, E. 2010.

Scenarios for the evolution of generation system and

transmission, distribution grid evolution requirements

for different scenarios of EV penetration in different

countries. Deliverable D3.1 of the European Project

MERGE.

SMARTGREENS2012-1stInternationalConferenceonSmartGridsandGreenITSystems

100