IEEE Power & Energy Magazine - Grid Edge 2023 - 70

or more is higher than 70% (approximately 63% on weekend
mornings). On the other hand, second charging events,
which occurred mostly at night and represented fewer than
one-third of all events, are less likely to reach full charge
because using these EVs at night results in some requiring a
80
70
10
20
30
40
50
60
new charging event early in the morning. During weekdays,
the highest probability (approximately 75%) of reaching a full
charge occurred between 3 and 9 p.m. On weekends, this
happened mostly (approximately 80%) between 6 p.m. and
7 a.m. Overall, the probability of finishing an EV charging
event with eight units or fewer
is lower than 20%, suggesting
that users prefer to end charging
events with a high final SOC.
Number of Units (% of Battery SOC)
(a)
10
20
30
40
50
60
70
Number of Units (% of Battery SOC)
(b)
6-10 a.m.
9 p.m.-6 a.m., 10 a.m.-3 p.m.
3-9 p.m.
Whole Day
figure 6. Normalized histograms of the final SOC per charging event considering
time dependency (weekday). The (a) first and (b) second charging events.
40
35
30
25
20
15
10
5
Charging on the Same
Day as Your Neighbor?
EV users have different charging
needs; some are likely to charge
their EVs every day, but others
might not. This is shown in Figure 7,
which provides the probabilities
of charging on the same day for
different percentages of EVs on
weekdays and weekends. Regardless
of the type of day, the probability
of all EVs being charged
at least once on the same day is
approximately 7%, i.e., twice per
month. In addition, the probability
of the majority of EVs (half
or more) charging at least once
on the same weekday is more
than 75%, i.e., more than threequarters
of a month. Interestingly,
for approximately 6% of the days
(nearly twice per month), no EVs
were charged at all.
Weekday
Weekend
0102030405060708090 100
EV Penetration Level (%)
figure 7. The normalized histogram for different percentages of EVs (same-day
charging).
70
ieee power & energy magazine
Modeling EV
Charging Demand
The MEA project developed a
methodology that combines the
normalized histograms presented
here as well as the typical EV
demand (approximately 3.6 kW)
and its power factor (0.98 inductive)
to produce daily time-series
EV profiles. Following a random
selection approach, the methodology
defines the key parameters of
an EV profile: the number of connections
a day, the start-charging
time of each connection, and the
initial and final SOC of each connection
[see Quirós-Tortós et al.
(2018) in the " For Further Reading "
section].
Figure 8 shows examples of
three individual EV profiles for
november/december 2018
Probability (%)
Probability (%)
Probability (%)
0 (0)
1 (8.33)
2 (16.66)
3 (25)
4 (33.33)
5 (41.66)
6 (50)
7 (58.33)
0 (0)
1 (8.33)
2 (16.66)
3 (25)
4 (33.33)
5 (41.66)
6 (50)
7 (58.33)
8 (66.66)
9 (75)
10 (83.33)
11 (91.66)
12 (100)
8 (66.66)
9 (75)
10 (83.33)
11 (91.66)
12 (100)
IEEE Power & Energy Magazine - Grid Edge 2023

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - Grid Edge 2023
