Environmental Engineering Reference
In-Depth Information
7.1
Modelling PHEV mobility
7.1.1 Modelling methods
Optimal charging of EVs is an intriguing and relevant topic within the smart-grid
theme. The problem itself is susceptible to many factors that can influence results
to varying degrees. These factors include flexible electricity tariffs, carbon content
of the power used to run the vehicle, battery capacity, charging rate, vehicle travel
patterns and network features ( e.g. topologies) - just to mention a few. Therefore,
for power system engineers, optimal PHEV charging is a novel field of research with
unresolved issues from a technical-economical, market and policy perspective.
Summaries on measuring, managing, modelling and aggregating the impact of
PHEVs on the grid have been presented in References 225 and 226. Thus far an item
left unresolved when tackling this peculiar problem is the fact that researchers have
to make broad assumptions when addressing mobility issues in power flow calcu-
lations [227]. Vehicle usage and the resulting energy demands are generally based
on aggregation, for example composing census data and average trip distances [228]
or combining local travel surveys with comparable datasets from other countries
[229]. These are then translated into PHEV load curves that can be added to existing
residential and commercial loads. A similar approach was used in Chapter 6.
The research gap between modelling vehicle travel patterns and combining
them with power flow analysis lacks a granular depth because there is limited data
available on travel surveys for light duty vehicles. Furthermore, work based on survey
data does not detail how vehicles move through road networks, nor the power network
layout they interact with, and hence has been limited by assuming vehicles obey a
specific travel pattern. Power engineering papers developed thus far usually have the
goal of either improving grid performance [228] or reducing charging costs to PHEV
owners while also making use of low carbon electricity [229].
A recurrent topic in the literature is PHEV capability in demand response
strategies via third party coordination [230]. In these works limited spatial data is
sufficient to conduct a high-level analysis; however, greater granularity is needed for
a better depiction of PHEV flows in distribution networks. An alternative is using an
activity-based approach to trip generation which results in spatial-temporal informa-
tion on PHEV movements [231]. Such activities can then be translated into energy
requirements, with the advantage that spatial dimension is also known. In a similar
fashion, Reference 232 uses micro-simulation of traffic, including potential conges-
tion, to create vehicle profiles which are used to study the impact of PHEVs on the
power grid in the city of Zurich [233]. These approaches use the agent-based mod-
elling (ABM) paradigm, which is a bottom-up way to model actors in a system,
including their autonomous actions and interactions.
Given the socio-technical interactions of electricity systems, in particular when
including distributed energy resources (DERs), it is no surprise smart-grid studies
have been published in artificial intelligence literature since that is the place where
the agent-based paradigm originates from. For example the acceptance of PHEVs as
an alternative to cars with an internal combustion engine has been studied with an
 
Search WWH ::




Custom Search