I’m not experienced (enough) with the coding, would be great to have an example of car charging based on remaining the current battery percentage.
Ie. if the car has 30% remaining, and default charging profile is set to max. 80%, would be nice to have a notification when the charge should start charging (which can be automated as well) for best price performace.
You would need the battery capacity (in my case 77kWh) and the default charging speed (11kW) in a helper I guess?
In most situation the time only is provided when the car is actually charging, which is not the case when you just plugin the chargeconnector, then it’s narmally ‘connected - waiting to charge’.
You can plug it in, let it charge for a few minutes until you know the remaining time, then use that to determine the best time to continue charging.
You will need to know what the remaining time is, as that is a factor to determine the best time.
Yes that’s true, but not very ‘hands-off automation’.
When you know the current percentage, full capacity of the battery (kWh), and charging speed (kW) you can make a estimation of the chargetime which can be used in the code. Right? Or am I missing an element?
Well, batteries usually charge slower the fuller they are.
I don’t have an electric vehicle myself, so I don’t have any practical experience with how to use the macro for it. The example is based on the information provided by another user of the macro.
There’s a lot wrong with that, but yes, you could use something like that.
I’m currently busy, but I’ll respond with a better version later today or tomorrow.
The template (binary) sensors you now have is the legacy template sensor format. I would advice to use the modern format.
There was also an issue with your end paramter. After midnight this would still add one day, and this could result in your car not being charged at 7:00 in the morning because the cheapest moment was eg at 15:00
It would look like this (note that template: is the integration, your current sensors are under sensor, the code below is how it could be placed directly in your configration.yaml):
template:
- sensor:
- unique_id: f48b4b86-a6c9-4791-93a3-daf9c3a89ff2
name: "EV estimated charge time"
unit_of_measurement: "minutes"
state: >
{% set battery_capacity = states('input_number.ev_battery_capacity') | float %}
{% set charge_speed = states('input_number.ev_charging_speed') | float %}
{% set current_charge_percentage = states('sensor.ev_state_of_charge') | float %}
{% set current_charge_kwh = (current_charge_percentage / 100) * battery_capacity %}
{% set remaining_kwh = battery_capacity - current_charge_kwh %}
{% set remaining_time_hours = remaining_kwh / charge_speed %}
{{ (remaining_time_hours * 60) | round(0) }}
availability: "{{ states('sensor.ev_state_of_charge') | is_number }}"
- binary_sensor:
- unique_id: 8ced58d7-8919-4a0f-8cd6-4b8b55b81470
name: "EV optimize charge"
state: >
{% set sensor = 'sensor.nordpool_kwh_nl_eur_3_10_0' %}
{% set hours = states('sensor.ev_estimated_charge_time') | float / 60 %}
{% set end = today_at('07:00') + timedelta(days=1 if now() > today_at('07:00') else 0) %}
{% from 'cheapest_energy_hours.jinja' import cheapest_energy_hours %}
{{ cheapest_energy_hours(sensor=sensor, hours=hours, start=now(), end=end, mode='is_now') }}
availability: >
{{
is_state('device_tracker.ev_tracker', 'home')
and is_state('binary_sensor.charger_connected', 'Locked, can connected')
and states('sensor.ev_estimated_charge_time') | is_number
}}
Possible adjustment / tweak to be made is that the code can be optimized to start at the beginning of the lowest price. Now it seems that it started charging at 02:40 when the first 40mins of the lowest price are already ‘gone’. The end of the charging session is slightly in more expensive tariffs.
Another aspect is that, when the dishwasher (example) kicks in, the power to the EV charger is limited (load balancing) this may cause a delay in charging the EV, and extending the period in more expensive tariffs at the end of the charging session.
Is that possible with a command, to start at the very beginning of the cheapest moment (02:00 this morning in the screenshot), instead of waiting?
Another option would be to manually add let’s say 15min to the estimated charge time, to incorporate any slight delays in charging.
Sorry for my late response.
I think you’re misreading the data.
The cheapest hour in your graph is from 4:00 to 5:00, so it starts at 2:40 to fully make use of that cheapest hour.
If it would have started 40 minutes earlier, you would have missed the 40 cheap minutes from 4:20 to 5:00
Apparently the charging did take like 1 or 2 minutes longer, so that’s why it charged a few minutes in the 5:00 to 6:00 time block. You could add maybe 5 minutes to the estimated charge time to account for that, that will make it start a bit earlier.
Add option to use 'all' as input for the hours parameter. When hours='all' is set, all hours between the start and end will be used. This can for example be practical when you want the average price of the whole day, and will avoid issues when hours=24 and there are only 23 hours in the day because of DST.
BUG FIXES
Fix a bug in the time_key finder when the data also includes yesterday’s prices.
I have made a few sensors such as expensive_hours_8_13 e.g. morning’s expensive hours and I want to avoid battery charge within these hours. Typicly in my area the electricity price is higher 07-12 than 12-17 so if I have inverted in self-use mode it will charge as soon as PV excess is present. I need to change inverter working mode to feed-in during expensive hours and get it back to self-use mode during cheap hours. I have tested automation based on these sensors but got stuck at trigger step. Shoud I use a state of an attribut in this case which one? There are only unavaillable and unknown. I have also cheched examples in the documentation but didn’t find them usefull.
IMPROVEMENTS
BUG FIXES