C IRE D 20th International Conference on Electricity Distribution
Paper 0160
STOCHASTIC ANALYSIS OF THE IMPACT OF PLUG-IN HYBRID ELECTRIC
VEHICLES ON THE DISTRIBUTION GRID
Kristien CLEMENT Edwin HAESEN
KU Leuven - Belgium KU Leuven - Belgium
Kristien.Clement@esat.kuleuven.be Edwin.Haesen@esat.kuleuven.be
ABSTRACT
Alternative vehicles, such as plug-in hybrid electric vehicles
(PHEVs), become more popular. The batteries of these
PHEVs are designed to be charged at home, from a
standard outlet in the garage, or on a corporate car park.
These extra electrical loads have an impact on the
distribution grid The uncoordinatedpower consumption on
a local scale can lead to grid problems. Therefore
coordinated charging is proposed The exactforecasting of
household loads is not possible, so stochastic programming
is introduced The stochastic approach represents an error
in the forecasting of the daily load profiles. Two main
program techniques are analyzed: quadratic and dynamic
programming. The coordination ofthe PHEVs reduces the
power losses and improves the power quality. The
estimation of the costs of grid reinforcement must be
compared with the cost of the implementation of a smart
metering system for the coordination ofthe charging.
INTRODUCTION
Hybrid electric vehicles (HEVs) are commercialized and
plug-in hybrid electric vehicles (PHEVs) are becoming
more popular. PHEVs are charged by plugging into electric
outlets or on-board electricity generation. These vehicles
can drive at full power in electric-only mode over a limited
range. As such, PHEVs offer valuable fuel flexibility [1].
The maximum penetration degree is 30% by 2030 for
Belgium as predicted by Tremove [2].
There are two main places where the batteries of PHEVs
can be recharged: either on a car park, corporate or public,
or at home. The focus here lies on the latter. The electrical
consumption for charging PHEVs may take up to 5% ofthe
total electrical consumption taking in Belgium by 2030 [3].
For a PHEV with a range of60 miles (100 km), this amount
can increase up to 8%. For these computations, a utility
factor is taken into account [4].
The charging of PHEVs has an impact on the distribution
grid because these vehicles consume a large amount of
electrical energy and this demand of electrical power can
lead to extra large and undesirable peaks in the electrical
consumption.
The improvements in power quality that are possible by
using coordinated charging is emphasized in [5]. It also
indicates that not coordinating the charging of PHEVs
decreases the efficiency of the distribution grid operation.
Deterministic household profiles are used in [5], assuming
CIRED2009 Session 4 Paper No 0160
Prague, 8-11 June 2009
Johan DRIESEN
KU Leuven - Belgium
Johan.Driesen@esat.kuleuven.be
that there is a perfect knowledge of future data. For this
approach a sufficient number of measurement data must be
available. Most of the time, these measurements are not
adequate to do a perfect forecasting ofthe data. A stochastic
approach, in which an error in the forecasting of the daily
load profiles is considered, is therefore more realistic. The
same assumptions and modelling as in [5] are considered.
For the distribution grid, the IEEE 34 node test feeder [6] is
used.
QUADRATIC PROGRAMMING
The idea is to achieve optimal charging and grid utilization
to minimize power losses in a stochastic approach. This
optimization problem can be tackled by the quadratic
programming technique (QP) [7,8].
Optimization
The daily load profiles are the essential input parameters,
which are fixed for the deterministic approach. In the case
of a stochastic approach, the uncertainties of these
parameters can be described in terms ofprobability density
functions. In that way, the fixed input parameters are
converted into random input variables with assumed normal
distributions at each node. The standard deviation (J is
determined in such a way that for 99.7 % of the cases, the
value of the stochastic sample is within a band of5 or 25 %
of the actual value Il for each time step of the daily load
profiles.
Results
For 2000 independent samples of each daily load profile in
the grid, one optimal charging profile is calculated. This
optimal charging profile is used to determine the power
losses for the 2000 individual load profiles. This is the
stochastic optimum. For each of these 2000 load profiles,
the optimal charging profile and the corresponding power
losses are also computed, which is the deterministic
optimum. The power losses of the deterministic optimum
are subtracted ofthe power losses ofthe stochastic optimum
and divided by the deterministic optimum, defined as ~P
which is always positive. This is shown for a variation of
the household loads of 5 and 25% in respectively Figure 1
and Figure 2. The forecasting of the daily load profiles
introduces an efficiency loss because the charge profiles of
the PHEVs are not optimal for this specific daily load
profile.
 C IRE D 20 th International Conference on Electricity Distribution Prague, 8-11 June 2009

Paper 0160

If the standard deviation ofthe normal distribution and thus contrast in terms of power losses is not large between the
the variation of the household load are reduced, the 2000 deterministic and stochastic optimum.
charge profiles of the deterministic optimum will converge
to the optimal charge profile and the efficiency loss will - optimal charge profile
reduce indieating that the power losses of the differences - - - deterministic optimum
will go down as shown in Figure 1 compared to Figure 2. co
""'~ 1500
/
/,
\
0
250 ,--- ----,- - ---,- - ---.--- - -,--- - ,--- ----,- - , c
.E
~ \

200 ~ 1000
\
\ ",
ro
,
s:
o ,, ,
'"
= , ,,,
150
'0
:;;
,, ,
" ,, ,,,
500
0
a,
,
,, ," "
,,
100 , ,, ,
,
0
22h 23h OOh 01h 02h 03h 04h 05h 06h
Time [h]
50
Figure 3: The deterministic opt imum and opt imal charger
profil e for node 33.

00 0.002 0.004 0.006 0.008 0.01 0.012 0.014 The importance of the knowledge of the peak during the
li P [%] evening is demonstrated by Figure 4. For the simulation, the
Figure 1: Histogram of the difference between the stochastic peak is shifted 1.5 hours to a later time step which increases
and the deterministic optimum of an arbitrary day during the efficiency loss, because the vehicles are now charging
winter for a va r iati on of 5%. during the actual evening peak.
300 400 ,--- - ---.--- - ---,-- - - ,--- - --,--- - ---,- - ----,

350
250

300
200
250

150 200

150
100

100
50
50

0 0 '----
0 0.4
li P [%]
Figure 2: Histogram of the difference between the stochastic
and the deterministic optimum of an arbitrary day during
winter for a va riation of 25% .
In general, the difference between the power losses of the
stoehastic and the determini stic optimum is rather small. It
is clear that the error in forecasting does not have a large
impact on the power losses. The daily household load
profiles are having the same trend each day during winter
season resulting in an optimal charge profile which
resembles to a deterministic charge profile ofa specific day
as plotted in Figure 3. Both charge profiles are having the
same trend, which is a dip between 23h and OOh, i.e. the
peak for most residential loads in this case. Therefore the
-
0.5 1.5 2 2.5 3 3.5 4 4 .5
li P [%]
Figure 4: Histogram of the difference between the stochastic
and the determinist ic optimum for shifting peak load.
DYNAMIC PROGRAMMING
The optimal coordination of charging PI-IEVs can also be
tackled by the dynamic programming technique (OP). The
OP technique decomposes the original optimization
problem into a sequence of subproblems which are solved
backward over each stage . A classical implementation of
the OP technique is the shortest path problem [9]. For the
application of this paper, the model is represented as a
series of plug-in hybrid electric vehicles.

CIRED2009 Session 4 Paper No 0160

 C IRE D 20 th International Conference on Electricity Distribution Prague, 8-11 June 2009

Paper 0160

Optimization penetration degree of30% and the power losses are the ratio
The dynamic programming technique successive ofthe power losses to the total load. The coordination ofthe
approximation (OPSA) decomposes the multidimensional charging reduces the power losses and improves the power
problem in a sequence ofone-dimensional problems which quality considerably with respect to the uncoordinated
are much easier to handle [10]. The optimizations occur at charging to a level which is similar to the case without
one variable at a time meanwhile holding the other variables PHEVs.
at a constant value. All the variables are evaluated that way. Table 1: Power quality and losses for the IEEE grid.
Results Parameters Without Uncoordinated Coordinated
PHEVs charging charaine
In general, the difference between the results of both
Load IkVAI 40 56 40
techniques is negligible although the QP technique gives
Line
more accurate results because the values of the charge 182 266 183
current (AI
profile are continuous, in contrast to the OP technique Node
where there is a step size of400W introduced for the power 213 207 213
voltage IVI
of the charger, giving a discrete charge profile. Reducing Power
the step to an infinitesimal value for the OP technique 3.1 4.4 3.8
losses (%1
would give the same result as for the QP technique. This
step size is taken rather large to reduce the number oflevels The coordination of the charging can be done by a smart
and with that the computational time and storage metering system. The distribution grid must be enforced to
requirements. The storage requirements are heavier for the cope with the increased loads and voltage drops by charging
OP technique compared to the QP technique because every PHEVs if this coordination system is not applied. Both
possible path over each stage must be stored leading to very scenarios will introduce extra costs for the distribution
large matrices increasing also the computational time and system operators (OSO) and eventually for the customers.
making the QP technique faster. In Figure 5 are the charge
profiles for the QP and OP technique compared. The A global estimation is performed to indicate how many
profiles are following the same trend but are a bit different expenses are involved for Belgium for the upgrading of a
due to the discrete values of the DP charge profile . small distribution grid. For the argumentation , the IEEE 34
node test feeder is connected to each phase of a three phase
- charge profile OP
transformer of 160 kVA which is standard in Belgium,
- - - charge profile DP forming a global grid of 100 nodes. Under no PHEVs
conditions, the maximum load for the three grids together is
2000 I , 120 kVA. Considering no PHEVs in the future, the
,," '' ,,"" ,, transformer has enough reserve capacity for this global grid
:;; , ,
,, '' ,, ,, to meet additional peak load and the load growth for the
i? 1500
'"
,
, , ' ,
,," ,,
"
"5 ' ,, -'' next 10 years, which is predicted to be a few percent. A
'"
' ,, ,, ,, , 4x95 Aluminum underground conductor of 400 V is
~ 1000 "\I , ,
I
standard in Belgium. The maximum capacity of these
'~"
,

c,
V
,,,
conductors is 220 A [11]. For the case without PI-IEVs, the
,, standard underground conductor would be sufficient.
500
, , If PHEVs are introduced , the power for the global grid
, "
increases to 168 kVA, which is out of range for the 160
"
"
,,
" kVA transformer. This transformer must be replaced by a
22h 23h OOh 01h 02h
Time[h]
03h 04h 05h 06h standard transformer of 250 kVA to deal with the extra
PI-lEVs, the load growth and additional peak load. The price
Figure 5: The charge profile for node 1 for the QP and the of such a transformer is about 6000 including costs for
DP program technique. placement, transport, connection LV and HV,.... Due to the
PHEVs , the line currents increase to 266 A. The maximum
IMPACTS ON THE DISTRIBUTION GRID capacity ofthe current conductor is not enough and must be
replaced by a 4xl50 Al underground conductor with a
The uncoordinated charging of the batteries of plug-in capacity of 280 A. The cost price of these conductor is
hybrid electric vehicles has a not negligible impact on the about 60 /m including costs for placement, supervision,
performance of the distribution grid in terms of power trench,.... The total length of the grid is about 360 m. The
losses and power quality. The power quality and losses are total cost for the replacement is 21 600. The total cost
represented in Table 1 for three cases: without PHEVs, the price for the OSO is 27 600 due to the introduction of 30
uncoordinated and coordinated charging. The power quality PHEVs per 100 nodes. It must be emphasized that the cost
is given as the average of 1000 samples of peak load, prices are only an approximated value and gives only the
voltage drop and line current during winter season for a magnitude of the costs.

CIRED2009 Session 4 Paper No 0160

C IRE D 20 th International Conference on Electricity Distribution
Voltage deviations up to 10% in low voltage grids are
acceptable for 95% of the time according to the EN50I60
norm which is mandatory in Belgium [12]. In the case of
uncoordinated charging, this limit has been reached and
action must be taken to reduce the voltage drop. The
problem of the voltage drop can be tackled by placing a
capacitor bank or an on-load tap changing transformer
although the latter is not common at low voltages, but may
be necessary in the future.
There is also another cost involved: the power loss. These
losses increase reasonable in the case of uncoordinated
charging. The power losses and loads must also be
produced and transported over the transmission lines.
A smart metering system based on a multi-agent-system
(MAS) must be implemented to control the coordination
and communication between the PHEVs individually, the
distribution system operator and the transmission system
operator (TSO). The vehicles could also be grouped and
represented by a fleet manager to communicate with the
DSO and TSO. Smart metering will lead to opportunities to
make PHEVs a controllable load, to apply vehicle-to-grid
(V2G) and to combine PHEVs and renewable energy. This
technology is available for implementation, but capital
investments by the utilities are necessary [13]. The
realization of smart metering only for the coordination of
PHEVs would be unlikely but it would be rather an extra
incentive for the implementation of this technology.
Less grid enforcements could be necessary with the
coordination system. The maximum load is lower because
the vehicles are not charging if the household loads are
peaking. Therefore, the voltage drops, line currents and
power losses are considerably reduced. The cost price of
upgrading the grid must be compared with the cost price of
the execution of smart metering.
CONCLUSION
In general, coordinated charging of plug-in hybrid electric
vehicles can lower power losses and voltage deviations by
flattening out peak power, although in some cases, grid
enforcement will be necessary. The implementation of the
coordinated charging is also not without costs. It is difficult
to determine the economical optimum of both methods.
However the choice of charging periods is rather arbitrary,
the impact of the penetration level is large.
Stochastic programming is introduced to represent an error
in the forecasting which increases the power losses. This
efficiency loss is rather small if the trend of the household
load profiles is known, so charging during the peak load of
the evening can be avoided.
CIRED2009 Session 4 Paper No 0160
Prague, 8-11 June 2009
Paper 0160
These results are obtained by the quadratic programming
technique. The dynamic programming technique is also
implemented but does not improve the computational time
or the achieved accuracy. The applied techniques and
methods can be extended to other objective functions.
REFERENCES
[1] A. Raskin and S. Shah."The emerge ofhybrid electric
vehicles," Alliance Bernstein. (2006)
[2] S. Logghe, B. Van Herbruggen, and B. Van
Zeebroeck," Emissions of road traffic in belgium,
transport & mobility Leuven," Tremove, Jan. 2006.
[3] K. Clement, K. Van Reusel, and J. Driesen, "The
consumption of electrical energy of plug-in hybrid
electric vehicles in Belgium," EET-2007 2nd
European Ele-Drive Transportation Conference,
Brussels, Belgium, May 2007.
[4] M. Duvall and E. Knipping, "Environmental
assessment ofplug-in hybrid electric vehicles volume
1: Nationa wide greenhouse gas emissions," EPRI,
Tech. Rep., 2007.
[5] Clement K., Haesen E., Driesen J, "The impact of
uncontrolled and controlled charging ofplug-in hybrid
electric vehicles on the distribution grid," EET-2008
3rd European Ele-Drive Transportation Conference,
Geneva, Switzerland, March 11-13,2008.
[6] W. H. Kersting, "Radial distribution test feeders."
IEEE Power Engineering Society Winter Meeting,
January 28 - February 1, 2001.
[7] E. Haesen, J. Driesen, and R. Belmans, "Robust
planning methodology for integration of stochastic
generators in distribution grids," lET Journal on
Renewable Power Generation, vol.l, no.l , pp. 25-32,
March 2007.
[8] J. Linderoth, A. Shapiro, and S. Wright,"The
empirical behavior ofsampling methods for stochastic
programming," Annals of Operations Research, vol.
142, no. 1, February 2006.
[9] R. Bellman, Adaptive control processes: a guided
tour: Princeton University Press, 1962.
[10] 1. Yi, J. Labadie, and S. Stitt, "Dynamic optimal unit
commitment and loading in hydropower systems,"
Journal ofwater resource planning and management-
ASCE, vol. 129, no. 5, pp. 388-398, Sep-Oct. 2003.
[11] H. L. Willis, Power Distribution Planning Reference
Book. New York (N.Y.): Dekker, 1997.
[12] "En 50160, voltage characteristics of electricity
supplied by public distribution systems," 1999.
[13] U.S. Departement of Energy, "Summary report:
Discussion meeting on plug-in hybrid electric
vehicles," Tech. Rep., 2006.