Charger¶

A Charger models a charging point: the plug that connects an ElectricVehicle to the grid. At every step, the optimizer decides which vehicle each charger serves, and the charger's rating caps how fast that vehicle charges or, with V2G, discharges.

Chargers and vehicles come together: a portfolio with electric vehicles needs at least one charger, and a portfolio with chargers needs at least one vehicle. Building the AssetPortfolio raises an OdysValidationError otherwise.

Basic usage¶

Let's give the depot of the ElectricVehicle page one 22 kW charger for its two vans:

from odys import Charger

charger = Charger(name="charger", max_power=0.022)  # 22 kW

The charger goes into the same portfolio as the vans:

from datetime import timedelta

from odys import AssetPortfolio, Battery, ElectricVehicle, Trip

van_1 = ElectricVehicle(
    name="van_1",
    battery=Battery(
        capacity=0.100,
        max_charge_power=0.022,
        max_discharge_power=0.011,
        efficiency_charging=0.95,
        efficiency_discharging=0.95,
    ),
    trips=(
        Trip(
            name="morning_delivery",
            start_time=timedelta(hours=2),
            end_time=timedelta(hours=4),
            energy_consumption=0.030,
            min_soc_at_departure=0.8,
        ),
    ),
)
van_2 = ElectricVehicle(
    name="van_2",
    battery=Battery(
        capacity=0.100,
        max_charge_power=0.022,
        max_discharge_power=0.011,
        efficiency_charging=0.95,
        efficiency_discharging=0.95,
    ),
    trips=(
        Trip(
            name="afternoon_route",
            start_time=timedelta(hours=6),
            end_time=timedelta(hours=9),
            energy_consumption=0.040,
            min_soc_at_departure=0.7,
        ),
    ),
)

portfolio = AssetPortfolio([van_1, van_2, charger])

Fields¶

efficiency has no effect today: charging through a charger loses nothing beyond the battery's own efficiencies. Put the charger's losses into the battery's efficiency_charging and efficiency_discharging instead. See CH-4.

Sharing chargers¶

With fewer chargers than vehicles, the chargers are the bottleneck, and the optimizer decides who charges when, from the trips and the prices. At every step:

• a charger serves at most one vehicle, and a vehicle uses at most one charger;
• a vehicle on a trip isn't connected to any charger;
• a vehicle without a charger can't charge or discharge;
• a connected vehicle charges or discharges at most at the charger's max_power, and at most at its own battery's ratings.

A vehicle can move to another charger from one step to the next, at no cost, and any charger lets a vehicle with V2G discharge. See CH-1 to CH-4.

Results¶

The depot is complete now. Let's connect it to the grid, with a price that changes through the shift (see EnergyMarket), and solve it. First, the vans' state of charge, through result.electric_vehicles:

from odys import (
    ElectricVehicleFinalState,
    ElectricVehicleInitialState,
    EnergyMarket,
    EnergySystem,
    PriceProfile,
    Scenario,
)

grid = EnergyMarket(name="grid", max_trading_volume_per_step=1.0)

energy_system = EnergySystem(
    portfolio=portfolio,
    markets=grid,
    scenarios=Scenario(
        profiles=(PriceProfile(market=grid, values=[60, 45, 70, 90, 80, 55, 50, 60, 85, 120, 110, 90]),),
    ),
    timestep=timedelta(hours=1),
    number_of_steps=12,
    initial_conditions=(
        ElectricVehicleInitialState(vehicle=van_1, soc=0.5),
        ElectricVehicleInitialState(vehicle=van_2, soc=0.5),
    ),
    final_conditions=(
        ElectricVehicleFinalState(vehicle=van_1, soc=0.5),
        ElectricVehicleFinalState(vehicle=van_2, soc=0.5),
    ),
)

result = energy_system.optimize()
print(result.electric_vehicles.soc.unstack().round(2))

ev    van_1  van_2
time
0      0.71   0.50
1      0.92   0.50
2      0.77   0.71
3      0.62   0.69
4      0.55   0.69
5      0.55   0.90
6      0.75   0.77
7      0.96   0.63
8      0.85   0.50
9      0.73   0.50
10     0.62   0.50
11     0.50   0.50

The trips shape the vans' day:

• van_1 charges in the first two hours and leaves at step 2 with 92 %, above the required 80 %, because it can sell the extra back at 80 $/MWh when it returns at step 4.
• Each trip lowers the state of charge evenly over its steps: 0.15 per step for van_1's 30 kWh over two steps.
• After charging again at steps 6 and 7, van_1 discharges at 11 kW (V2G) through the afternoon price peak, down to the required 50 %.

Now the charger, through result.chargers:

print(result.chargers.to_dataframe().round(3))

              assignment[van_1]  assignment[van_2]  power
time charger
0    charger                1.0                0.0  0.022
1    charger                1.0                0.0  0.022
2    charger                0.0                1.0  0.022
3    charger                0.0                1.0  0.000
4    charger                1.0                0.0  0.000
5    charger                0.0                1.0  0.022
6    charger                1.0                0.0  0.022
7    charger                1.0                0.0  0.022
8    charger                1.0                0.0  0.000
9    charger                1.0                0.0  0.000
10   charger                1.0                0.0  0.000
11   charger                1.0                0.0  0.000

Every row has a single 1: the charger serves one van at a time. van_2 gets it while van_1 is on its morning trip, at steps 2 and 3, and at step 5, just before its own route. The rest of the shift it goes to van_1.

result.chargers has two series, with a scenario level when there's more than one scenario:

• assignment: 1 when a vehicle is connected to a charger, indexed by time, charger and vehicle. The dataframe has one assignment[...] column per vehicle.
• power: the charging power drawn through each charger, in MW. It counts charging only: what a vehicle discharges through the charger (V2G) is in that vehicle's net_power, not here. That's why power is 0 at steps 3, 4 and 8 to 11: the connected van feeds power back then, as its falling state of charge shows.

The Depot Charging with V2G example runs three vans and two chargers over a full day.

How it's modelled¶

A charger adds one decision per vehicle and step: whether that vehicle is connected to it. CH-1 to CH-3 decide who may connect, and CH-4 caps the vehicle's power at the rating of its charger, or 0 without one. A charger has no term of its own in the power balance or the objective. The rules and their limitations are on the Electric vehicles and chargers formulation page.

Next steps¶

The depot traded with the grid through a market. Next, see how an EnergyMarket works.