Make a home battery trade the half-hourly power market, then sell the setup

How Agile worksOn Agile Octopus ↗ every half hour has its own price, set from the wholesale market. Tomorrow's 48 prices are published every afternoon (usually around 4pm). The shape repeats most days: cheap overnight, often cheap again in sunny or windy middays, sometimes negative (you're paid to use power), and expensive from 4pm to 7pm. Agile Outgoing does the same for power you export. A battery earns in two ways: it replaces expensive peak imports with cheap overnight energy, and it can sell energy back when export prices are high.

Get the plannerDownload agile_plan.py. It uses only Python's standard library. Octopus publishes tariff prices on a public API, so there's no key and no account. Run it with your region letter (it's at the end of your tariff code and on your bill; C is London):

python agile_plan.py --region C

python3 agile_plan.py --region C

It fetches Agile import (AGILE-24-10-01) and Agile Outgoing prices for every half hour available, then runs a small dynamic programme: for each half hour it tries charging, holding and discharging at every battery level, and keeps the cheapest path through the whole period. It respects the battery's size, its charge rate, round-trip losses and your house's typical use. Run the evening the next day's prices came out:

Agile AGILE-24-10-01 region C: 58 half hours, Sat 10 Oct 18:00 to Sun 23:00
  import  min  15.41p  avg  28.12p  max  51.46p per kWh (inc VAT)
  export  min  10.20p  avg  16.28p  max  27.56p per kWh
  cheapest half hours: 22:30 15.76p, 00:30 16.01p, 01:00 15.41p, 02:00 15.97p, 02:30 17.29p, 03:00 17.60p

Battery: 9.5 kWh usable, 3.6 kW, 90% round trip, house load 0.35 kWh per half hour, start 2.0 kWh, reserve 1.0 kWh

time         import p export p  action            battery kWh
Sat 22:30      15.76    10.37  charge 1.8 kWh            2.8
Sun 00:30      16.01    10.49  charge 1.8 kWh            4.3
Sun 01:00      15.41    10.20  charge 1.8 kWh            6.1
Sun 02:00      15.97    10.47  charge 1.8 kWh            7.9
Sun 02:30      17.29    11.10  charge 1.6 kWh            9.5
...
Sun 16:00      44.43    24.22  discharge 0.4 kWh         5.6
Sun 16:30      49.41    26.59  discharge 0.4 kWh         5.2
Sun 17:00      49.68    26.71  discharge 0.4 kWh         4.8
Sun 17:30      51.12    27.40  discharge 0.4 kWh         4.4
Sun 18:00      51.38    27.52  discharge 0.4 kWh         4.0
Sun 18:30      51.46    27.56  discharge 0.4 kWh         3.6
...
Value of the battery over these 29 hours: 186p (about 154p a day at these prices)

Read the plan: fill up overnight at 15 to 17p, then cover the house through the day and the 44 to 51p peak. Notice it never exports here: tonight's export peak (27.56p) is worth less than the import it replaces (51p), so using the energy yourself wins. On a day with very cheap or negative overnight prices, the plan changes on its own.

Make it yoursThe defaults describe a common setup (GivEnergy 9.5kWh battery, 3.6kW inverter). Set your own:

python agile_plan.py --region M --capacity 13.5 --rate 5 --load 0.4 --soc 3 --csv plan.csv

• --capacity usable kWh, --rate max charge/discharge kW, --efficiency round trip (default 0.90)
• --load your average use per half hour in kWh (annual kWh divided by 17,520 is a fair start)
• --soc how full it is now, --reserve the floor it never goes below
• --no-export if you aren't on an export tariff