IEEE Power & Energy Magazine - Grid Edge 2023 - 71

weekdays as well as the average
profile of 1,000 EVs. It highlights
different start-charging times,
charging durations (based on the
initial and final SOCs), and an EV
with two daily charges (EV Load 1).
The EV profiles are validated
comparing the average demand
of the created pool ( " Model "
in Figure 8) with the average
EV charging behavior monitored
during the MEA trial. Clearly,
the average charging behavior of
the created profiles matches that
of the monitored EVs. The average
peak demand using the profiles
(1.09 kW) occurred at nearly
the same time (between 8 and
9 p.m.) as the ave r age pe a k
demand monitored during the trial
(1.08 kW). Moreover, it was found
that the average energy consumption
using the created EV profiles
(12.33 kWh) differed by less
than 3% from the average energy
consumption shown by the EVs
monitored in the trial (12.63 kWh).
To highl ight the effects of
3.5
4
2.5
3
1.5
2
0.5
1
Monitored
12
(a.m.)
24 68 10 12
(p.m.)
24 68 10
Time of Day (1-min Resolution)
figure 8. EV profiles and average (1,000 EVs) demand (weekday).
1.2
1.4
1.6
1.8
2
0.2
0.4
0.6
0.8
1
ignoring the dependency of the
initial and final SOCs on the startcharging
time, a set of 1,000 EV
profiles was also created using the
whole-day normalized histogram
shown in Figures 5 and 6. The
average EV demand shown in Figure
8 demonstrates that, although
the evening peak is similar to the whole-day approach, the
morning peak is 0.5 kW as compared to 0.8 kW (EVs connected
in the morning are likely to charge for shorter periods
than those connected overnight). Finally, it was found
that because the dependency of the initial and final SOCs on
the start-charging time is being ignored, EV charging modeled
with this whole-day alternative can (mistakenly) show
approximately 13% more energy consumption (14.24 kWh)
compared to the monitored charging behavior.
12
(a.m.)
2468 10 12
(p.m.)
2468 10
Time of Day (1-min Resolution)
figure 9. The average demand of 1,000 households, EVs, and households plus EVs
(weekday).
EVs and the Grid
It is generally assumed that some people charge their EVs
at home shortly after returning from work. This means that
the corresponding EV demand could be coincident with that
from the use of multiple appliances, such as lights. Although
this will depend on weather and behavioral aspects (which
vary from one country to another), in the MEA project this
assumption holds true. As shown in Figure 9, the maximum
average charging demand from 1,000 EVs (using the prenovember/december
2018
vious model) coincided with that of nonelectrically heated
households. (The 1,000 households are assumed to have an
EV.) From the perspective of DNOs, this coincident demand
can translate to technical problems on their networks
because it can significantly exceed the values used for the
original distribution network design. For instance, if a residential
low-voltage network has been designed considering
an average peak demand (also known in the United Kingdom
as after-diversity maximum demand) of 1-1.5 kW per
house, then adding an average of 1 kW per house resulting
from the widespread adoption of EVs will likely lead to voltages
below the statutory limits and/or to asset thermal overloads.
It is, therefore, in the interest of DNOs to evaluate how
these networks will be affected by different numbers of EVs.
Impacts on Low-Voltage Networks
The MEA network studies first investigated the hosting
capacity of three-phase low-voltage networks [modeled in
OpenDSS (distribution system simulator)] considering EV
ieee power & energy magazine
71
EV Load 1
EV Load 2
EV Load 3
Whole Day
1.2
1.4
1.6
Model
0.2
0.4
0.6
0.8
1
Households
EVs
Households and EVs
Average Demand (kW)
Individual EV Demand (kW)
Average EV Demand (kW)
IEEE Power & Energy Magazine - Grid Edge 2023

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