Electric Vehicles in District: Agent Based Modeling

Charging Electric Vehicles will have an impact in the electricity network infrastructure. What will be the influence at district level?

Charging Electric Vehicles will have an impact in the electricity network infrastructure. What will be the influence at district level?

Welcome to the Electric Vehicles in District: Agent Based Modeling simulation. The objective of this model is to estimate how Electric Vehicles (EVs) influence energy flows in a defined boundary (in this case, a district).
Unlike other simulations, this one uses agents (Electric Vehicles, in this case). Part of the model is dedicated to define the behavior of one of the EVs, that is, of one agent.

As it works with a variable "agent population", the model can change easily how many agents participate in the simulation, what is very useful to prototype different scenarios with a variable number of EVs.
The core of the model defines the behavior of the EV agent.  During the simulation, the behavior of each EV (40 as default) is modeled independently, including the status of its battery and whether the EV is charging or not.
The model displays the status of each one of the EVs:
  1. Charging: the EV is connected to a charging point, and withdrawing energy at a defined rate (Average power drawn).
  2. Full Battery: the EV battery is fully charged and it remains connected to the charging point.
  3. Absent: the EV is away from the charging point, and its battery is discharging.
The transitions from one state to the next one depend on the status of the battery and two additional factors:
  • How probable is for an EV to be conneced to a charging point right now?
  • If the EV is connected now, for how long will it remain connected?
As we are dealing with a large number of EVs, the characterization of these probabilites (mean and standard deviation for each hour of the day) can be extrapolated from preexisting data of charging points (This model does not use real data for the moment).
Although each EV is simulated separately, individual data can be extracted, too. As an example, this is the behavior of the EV number one.

Knowing how many EVs are charging at a given moment and their average power demand, it is possible to know the power required for EV charging in the district. What if there should be a cap to this power demand? The model allows to set a limit; then, if the numbers of EVs is too high, the maximum allowed power will be shared by each charging EV (and therefore their batteries will charge slowlier).
The simulation estimates the status of the EVs at each hour, in this case over ten days. The tab Total power demand indicates at each moment the load that the distribution network of the district will endure to charge the EVs.

This is the end of the explanation. Some ideas about how can I help you from here:
  • In your project, define custom agents and is features (for instance, define that half the Electric Vehicles shall have larger batteries).
  • Define a new category of agent for households, and study the potential of Vehicle to Grid (Showcase 2) at district level.
  • Scenarios: Study the impact of different policies and user behaviors.
  • Connect the simulation of several districts to model electricity flows in a whole city.

View the model in Insight Maker