A tool to analyze existing energy data to see how much money an energy storing system would have saved.
The software handles this:
- Storing excess solar power.
- Charge from grid.
- Use stored energy when most profitable.
Note that this is based on historical data, that you must have. And also of course historical prices, which you'll also have to provide. This doesn't tell the whole truth about the future. In fact, since this is based on real and historical data, where the prediction could be perfect, this should give a better result than in reality.
python3 -m analyzer.main sample-data/partial-year2.json.gz
Additional options:
usage: main.py [-h] [--window WINDOW] [--battery-capacity BATTERY_CAPACITY]
[--depth-of-discharge DEPTH_OF_DISCHARGE] [--charging-loss CHARGING_LOSS]
[--discharging-loss DISCHARGING_LOSS] [--max-grid-power MAX_GRID_POWER]
[--output-dir OUTPUT_DIR] [--no-grid-charge] [--start-time START_TIME]
[--end-time END_TIME]
input_file
Analyze potential battery savings from energy data
positional arguments:
input_file Input JSON file with energy data
options:
-h, --help show this help message and exit
--window WINDOW Optimization window size in minutes (default: 1440 = 24h)
--battery-capacity BATTERY_CAPACITY
Battery capacity in Wh (default: 24000 = 5 * 4800)
--depth-of-discharge DEPTH_OF_DISCHARGE
Minimum battery level as percentage (default: 5.0)
--charging-loss CHARGING_LOSS
Charging loss percentage (default: 7.5)
--discharging-loss DISCHARGING_LOSS
Discharging loss percentage (default: 7.5)
--max-grid-power MAX_GRID_POWER
Maximum grid charging power in watts (default: 17250 = 230V * 25A * 3
phases)
--output-dir OUTPUT_DIR
Output directory for visualization files
--no-grid-charge Disable grid charging
--start-time START_TIME
Start time in ISO format (e.g., 2024-05-01T00:00:00Z)
--end-time END_TIME End time in ISO format (e.g., 2024-05-02T00:00:00Z)
The software reads data in the following format.
{
"2024-05-01T00:01:00Z": {
"Wh": -92.49444444444445,
"importPrice": 1.4260000000000002,
"exportPrice": 0.5008
},
"2024-05-01T00:02:00Z": {
"Wh": -94.23333333333333,
"importPrice": 1.4260000000000002,
"exportPrice": 0.5008
},
...
}
The analytics will work best if you include all costs/benefits that scales with (k)Wh.
The code has been designed with minutely data in mind. It's untested how it behaves with lower and higher data resolutions. The higher the resolution, the higher the accurancy of the output though.
By using the --output-dir option the code generates a chart showing when different actions happens. With this you should be able to get a better understanding of the flow of energy and the battery level. This will be rendered as a single html file so that you easily can open it in your web browser.
python3 tests/test_simulation.py
Verifies that the output data is as expected with the given sample file.
python3 influx_fetcher.py \
--url "influx-url" \
--token "access-token" \
--org "your-org" \
--bucket "your-bucket" \
--start "2024-05-01" \
--end "2025-02-14T23:59:59Z" \
--output "sample-data/partial-year3.json" \
--tax-reduction 0 \
--network-benefits 0.08 \
--transfer-cost 0.8
Hint: Needs python3-influxdb-client on Debian. Or equivalents in other distributions.
This is output from the analyzer, based on real data.
$ time python3 -m analyzer.main sample-data/partial-year2.json.gz --output-dir ./out
Reading input file sample-data/partial-year2.json.gz...
Loaded 417600 data points
Grid charging parameters:
Combined losses: 14.4%
Loss multiplier: 1.2500
Required price ratio: 1.1805 (18.0%)
Processing data with 1440 minute windows...
Optimized Battery Simulation Summary
==================================================
Time period: 2024-05-01 00:00 to 2025-02-14 23:59
Days: 290 (0.79 years)
Battery Configuration:
Capacity: 24.0 kWh
Final level: 1.20 kWh
Energy Flows:
Excess energy stored: 3633.56 kWh
2024-05: 724.67 kWh
2024-06: 665.77 kWh
2024-07: 661.44 kWh
2024-08: 653.51 kWh
2024-09: 558.13 kWh
2024-10: 325.69 kWh
2024-11: 22.46 kWh
2024-12: 2.44 kWh
2025-01: 6.39 kWh
2025-02: 13.06 kWh
Grid energy charged: 3070.07 kWh
2024-05: 39.43 kWh
2024-06: 73.68 kWh
2024-07: 102.67 kWh
2024-08: 70.83 kWh
2024-09: 108.25 kWh
2024-10: 328.46 kWh
2024-11: 618.57 kWh
2024-12: 638.42 kWh
2025-01: 757.72 kWh
2025-02: 332.02 kWh
Battery energy used: 5735.79 kWh
2024-05: 653.79 kWh
2024-06: 632.70 kWh
2024-07: 653.79 kWh
2024-08: 619.77 kWh
2024-09: 570.17 kWh
2024-10: 559.71 kWh
2024-11: 548.48 kWh
2024-12: 548.34 kWh
2025-01: 653.79 kWh
2025-02: 295.26 kWh
Financial Summary:
Export value lost: 1014.14 SEK
2024-05: 170.46 SEK
2024-06: 209.76 SEK
2024-07: 177.04 SEK
2024-08: 128.21 SEK
2024-09: 157.04 SEK
2024-10: 117.35 SEK
2024-11: 17.92 SEK
2024-12: 3.63 SEK
2025-01: 13.81 SEK
2025-02: 18.93 SEK
Grid charging cost: 3903.10 SEK
2024-05: 35.87 SEK
2024-06: 71.05 SEK
2024-07: 122.69 SEK
2024-08: 76.24 SEK
2024-09: 104.61 SEK
2024-10: 358.60 SEK
2024-11: 856.96 SEK
2024-12: 774.62 SEK
2025-01: 944.06 SEK
2025-02: 558.39 SEK
Import cost saved: 13173.57 SEK
2024-05: 1400.99 SEK
2024-06: 1541.52 SEK
2024-07: 1362.50 SEK
2024-08: 1401.30 SEK
2024-09: 1032.44 SEK
2024-10: 998.24 SEK
2024-11: 1527.41 SEK
2024-12: 1384.75 SEK
2025-01: 1617.74 SEK
2025-02: 906.67 SEK
Negative Price Statistics:
WARNING: export_stored: 200.34 kWh during negative prices
Cost impact: -32.40 SEK
Net savings: 8256.33 SEK
Rough estimate: 10398.71 SEK/year (disclaimer: not a true value)
Battery State Statistics:
Number of times battery was full: 373
Number of times battery was empty: 329
Battery was full 18.8% of the time
Battery was empty 28.9% of the time
Battery was partially charged 52.4% of the time
Number of battery cycles: 239.0
Average cycles per day: 0.82
Visualization file has been created: ./out/battery_viewer.html
Open this file in a web browser to view the interactive visualization.