A Python module to charge/discharge Tesla Powerwall based on solar forecast and peak/off peak tariffs.
Project description
PwForecast
A Python module to charge/discharge Tesla Powerwall based on solar forecast and peak/off-peak tariffs.
Utilises the excellent TeslaPy by Tim Dorssers.
PwForecast is primarily designed for those with dual rate tariffs. If there is interest in expanding this to provide more flexible control for those with more complex tariffs, please let me know.
Getting Started
PwForecast requires a Solcast API key and Site ID. You can sign up for a free Hobbyist account which allows up to two sites and 50 API calls per day.
TeslaPy has great documentation on getting started. It even works with two-factor authentication enabled.
Rather than inherit, PwForecast requires an instance of a TeslaPy Tesla class passed to it. This allows you to configure the object based on your credentials before passing to PwForecast.
When using PwForecast, Self Powered mode is recommended. Time Based Control could be used if you need to charge faster than 1.7kW per Powerwall, but results may be unpredictable.
Overview
PwForecast has two main methods, set_peak_mode
and set_off_peak_mode
. Both can be configured using a PwForecast
instance.
Here is an example showing how to configure teslapy and PwForecast:
import teslapy
from pwforecast import PwForecast
tesla_email = 'name@example.com'
solcast_api_key = '1234-1234'
solcast_site_ids = {'Site_A': '1234-1234',
'Site_B': '1234-1234'}
with teslapy.Tesla(email=tesla_email) as tesla:
pw_forecast = PwForecast(teslapy_session=tesla,
solcast_api_key=solcast_api_key,
solcast_site_ids=solcast_site_ids)
Setting Peak Mode
Use this method when your peak rate starts. You can configure what backup reserve the Powerwall can discharge down
to by setting the min_reserve_peak_rate
value.
pw_forecast.min_reserve_peak_rate = 10 # Default 20
pw_forecast.set_peak_mode()
PwForecast will try to ensure power flow behaves as expected. Please see Retry Logic for more info.
A summary report will then be printed:
-----------------------------------
Powerwall state of charge: 19.7%
Powerwall backup reserve: 100%
Powerwall capacity: 26.59kWh
Powerwall state of health: 95.0%
-----------------------------------
Setting Off-Peak Mode
Use this method when your off-peak rate starts. This requires a little more setup.
- Configure what backup reserve the Powerwall can discharge down to by setting the
min_reserve_off_peak_rate
attribute. - Configure the maximum backup reserve allowed by setting the
max_reserve
attribute. - Configure the amount of energy you require during the peak-rate by setting the
required_energy_peak_rate
attribute.
When setting peak mode, PwForecast will determine how much solar will be generated tomorrow. It will then calculate how
much to fill the batteries based on the remaining energy requirement that solar will not satisfy. If solar generation
completely satisfies required_energy_peak_rate
, the backup reserve will be set to min_reserve_off_peak_rate
.
get_solar_forecast_tomorrow
is called internally as part of set_off_peak_mode
. This will consume a Solcast API call
per site ID you have provided.
pw_forecast.min_reserve_off_peak_rate = 25 # Default 30
pw_forecast.max_reserve = 95 # Default 100
pw_forecast.required_energy_peak_rate = 20000 # Default 30000
pw_forecast.set_off_peak_mode()
PwForecast will try to ensure power flow behaves as expected. Please see Retry Logic for more info.
A summary report will then be printed:
-----------------------------------
Solar forecast tomorrow: 10.2kWh
Powerwall state of charge: 19.7%
Powerwall backup reserve: 100%
Powerwall capacity: 26.59kWh
Powerwall state of health: 95.0%
-----------------------------------
Getting Solar Forecast
If you're only interested in getting the solar forecast, you can call get_solar_forecast_tomorrow
. This will
consume a Solcast API call per site ID you have provided. The value returned will be an int of the estimated
solar forecast tomorrow in Wh.
pw_forecast.get_solar_forecast_tomorrow()
Retry Logic
When setting backup reserve, it is common to take two or three attempts before power flow changes to the expected figures. As the Tesla API is unofficial, it is difficult to determine why. Waiting longer does not seem to fix this issue, although waiting upwards of 45 minutes does sometimes result in power flow eventually changing correctly. Re-applying the setting does appear to fix this issue. Perhaps the unofficial Tesla API is missing a commit command. Please do let me know if you have any ideas.
Calling set_peak_mode
and set_off_peak_mode
will set the backup reserve, wait 20 seconds, and then check site
power flow to confirm the setting has been applied. If an incorrect power flow is detected, the method will retry
up to the set_backup_reserve_retry_limit
limit. If the set_backup_reserve_retry_limit
is reached, an exception
will be raised and caught by the global retry logic. PwForecast will then attempt to re-apply the setting up to the
global_retry_limit
limit, eventually raising an exception.
Both set_backup_reserve_retry_limit
and global_retry_limit
can be configured on the PwForecast instance. The
set_backup_reserve_retry_limit
has been split from the global_retry_limit
to try and avoid exhausting Solcast API calls.
The amount of time between each global_retry
can be configured via global_retry_sleep
. The amount of time
between each set_backup_reserve_retry_limit
can be configured via set_backup_reserve_response_sleep
.
Advanced Configuration
PwForecast has a few advanced configuration options. Please check the class docstring for more info.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file pwforecast-1.3.1.tar.gz
.
File metadata
- Download URL: pwforecast-1.3.1.tar.gz
- Upload date:
- Size: 8.6 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/5.0.0 CPython/3.9.19
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 33d38c9ab23e2d8bb98583b18c79d95e88d619efd35e96f0272ee8f1274c9fd3 |
|
MD5 | 1e801ee82cc24132a2de96d1b96cd325 |
|
BLAKE2b-256 | a82dde794704c88aa2b3afddde3859aacb7f1c5d282f9759353e9a880cdeb5ed |
File details
Details for the file pwforecast-1.3.1-py3-none-any.whl
.
File metadata
- Download URL: pwforecast-1.3.1-py3-none-any.whl
- Upload date:
- Size: 9.1 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/5.0.0 CPython/3.9.19
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 4f3d47f87e087c3672d663525f700cda48204aa878e022bc33c2bac961364896 |
|
MD5 | 38b0717c943292f571a723e026f9b2cd |
|
BLAKE2b-256 | 8d3b252e24ee69c44ec905dd006db72bff8294098b81ed1770d96d182cad76b9 |