Modelling electric vehicle mobility in energy service networks - Modelling Distributed Energy Resources in Energy Service Networks - page 159

7.4.2 Optimal power flow model results
The TCOPF program makes the most of the ABM forecast on PHEV load flexibil-
ity resulting, when compared to the results in Chapter 6, in a much more detailed
description of the energy flows required to charge the batteries. Table 7.4 compares
basic network metrics that compare how PHEV penetration impacts the local net-
work assessed. Scenario 1 with no PHEVs gives benchmark data on energy provided
and network losses. As expected, once PHEVs are included in the model, the energy
supplied by the slack bus and the network losses increases slightly due to the new
mobile load brought to the infrastructure. In scenarios 2 and 3 it is possible to cal-
culate the electricity cost and carbon content with which vehicles are charged. In
the second scenario charging costs are lower than in the third scenario because the
latter has higher spot prices and a carbon tax of £60/ t CO 2 for charging PHEVs with
fossil fuelled power. Nevertheless, the prowess of the TCOPF solver is seen when
scenario 3 charges the PHEV fleet with cleaner electricity when compared to sce-
nario 2. Based on this case study and for sake of discussion, as things stand today
PHEVs would charge with an average grid carbon content of 0.554 kgCO 2 /kWh. But
if low carbon generation increases and coal is displaced, charging emissions could be
reduced to 0.393 kgCO 2 /kWh (see Figure 7.4), thus having serious implications on
Well-to-wheel ( W2W ) efficiency studies [242].

Table 7.4 Techno-economic results of the urban energy system

Case Slack Losses Charging emissions Charging

scenario (MWh) (MWh) ( t CO 2 ) cost (£)

1 507.86 14.63 0.00 0.00

2 512.97 14.82 2.72 244.01

3 512.97 14.82 1.93 435.40

Figures 7.9 and 7.10 display the load variations seen from the grid supply
point once PHEVs are included. Since penetration is low and there is a correla-
tion between power demand and prices, it is expected that charging occurs mostly
during the valley of the static curves. There is an exception on Saturday when condi-
tions allow for some charging to be done during the afternoon hours. It is important
to clarify that although charging profiles vary for each scenario, the daily energy
requirements of vehicles are the same, this explains why the slack and energy losses
are the same for scenarios 2 and 3. However, as the figures show the impartial TCOPF
solver dispatches charging differently due to distinct economic signals - this sort of
exercise helps to clearly highlight PHEV load flexibility.
From these sorts of results energy stakeholders should be confident that with
robust communication links and advanced algorithms a smart-charging scheme for
PHEVs is possible.
Figures 7.11 and 7.12 depict where , when and how much power optimally charged
PHEVs would require to satisfy energy requirements for scenarios 2 and 3. As it can