IEEE Power & Energy Magazine - July/August 2021 - 23

was " hot unplugged, " i.e., when the vehicle was still charging
immediately before being unplugged. If demand management
caused an increase in the hot unplug rate, then this would indicate
that there was not sufficient flexibility available to accommodate
smart charging. Figure 7 compares the hot unplug rates
for CrowdCharge and GreenFlux in the different trial phases.
The hot unplug rate remained low in all of the trial phases.
In the GreenFlux group, there was only a small (and not statistically
significant) increase in the hot unplug rate between the
outside of management period and trial 1. In all of the other
trial phases, the rate declined despite smart charging reducing
the speed of charging (trial 2) or charging being delayed until
the nighttime period in response to the ToU incentive.
For CrowdCharge, the only increase (though relatively
small) in the hot unplug rate occurred between trials 2
and 3. This is likely due to the algorithm used and the low
level of interaction between the participants and the journeyplanning
app. The trial 3 CrowdCharge algorithm reduced
the charging speed by more than
what was necessary to satisfy the
network's capacity limits. Without
this additional limitation, it is
unlikely that the hot unplug rate
would have increased.
Figure 7 shows that despite
regular demand management and
the participants delaying charging
in response to a ToU incentive,
there was little impact on the
proportion of charge events that
were not completed before the
vehicle was unplugged. The flexibility
demonstrated in Figure 6
can be exploited without altering
the vehicle's state of charge when
the driver unplugs the vehicle.
Can Financial Rewards
Encourage Drivers to
Charge Flexibly?
In trial 3, the users were offered
a financial reward linked to the
time of day they charged via a
simulated ToU tariff. The app was
updated to support the users in
changing their behavior. Trial 3
demonst rated the extent to
which participants were willing
to change their charging behavior
(that is, being flexible) when they
were offered a financial incentive.
This section describes the trial 3
results for the GreenFlux cohort.
Figure 8 presents the load profile
from the GreenFlux trial 3
july/august 2021
1,000
900
800
700
600
500
400
300
200
100
Trial 3 (With ToU) (90th Percentile)
End of Trial 2 (No ToU) (90th Percentile)
Winter Trial 1 (No ToU) (90th Percentile)
100
80
60
40
20
group for mid-October to mid-December (blue) compared to
the two other periods:
✔ the month before trial 3 began (orange)
✔ the winter (January-early March) period during trial
1 (gray).
The difference in the profile between trial 3 and trials
1 and 2 is clear, particularly during the traditional evening
peak period. The evening peak demand in trials 1 and 2
was suppressed using demand management. The impact of
the ToU system is clear: There is a significant reduction in
demand during the evening peak period. This reduction in
evening demand means that demand management was not
required during trial 3 as there was always enough capacity
available. The demand that was shifted away from the
evening peak period all began at 10 p.m., causing a spike in
demand. In a business-as-usual application, the start of the
delayed charging events would be distributed evenly across
the start of the overnight period, so that this spike would
Hot Unplug (% of Charge Events)
Outside of Management Trial 1 Trial 2 Trial 3
17.6% 16.2%15.2%
17.4%
14.7% 15.3% 13.6% 11.8%
CrowdCharge
GreenFlux
figure 7. The hot unplug rates according to trial cohort and phase.
12 a.m. 3 a.m. 6 a.m. 9 a.m. 12 p.m.
Time of Day
3 p.m. 6 p.m. 9 p.m. 12 a.m.
figure 8. The weekday demand curve for the 90th percentile. Trial 3 (the ToU tariff)
is compared to the other trials.
ieee power & energy magazine
23
Power Drawn (W Per Charger in Group)
(%)
IEEE Power & Energy Magazine - July/August 2021

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - July/August 2021
