GivEnergy Smart Home Display

Introduction

This is a web-based application for showing a live summary of energy data from a GivEnergy inverter. It is designed to run on a Raspberry Pi (in Chromium) and sized for rendering on a Raspberry Pi 7" touchscreen display when loaded in Chromium in fullscreen mode.

Web Application

The user interface of the web app is rendered completely using SVG. The only external dependencies for the user interface are a little use of jQuery. The rest of the app consists of pure JavaScript and CSS.

Data is obtained from the inverter by using GivTCP on the local network and this is captured using a local Home Assistant server. The user interface of the smart home display web app grabs the data from this Home Assistant server (using a WebSocket connection).

The reason Home Assistant is used is it offers various energy-monitoring capabilities and automatic tariff switching, and can help store energy usage data throughout the day and calculate the usage costs (taking into account the usage during different parts of the day - i.e. peak vs. off-peak).

Prerequisites

Hardware prerequisites:

• GivEnergy inverter - you obviously need a GivEnergy inverter connected to your local home network :)
• Raspberry Pi 3/4 - for loading this GivEnergy smart home display web app. You may also choose to install the software dependencies (below) on the Pi
• Raspberry Pi 7" Touchscreen Display - for attaching to the Raspberry Pi board
• Pibow Touchscreen Frame - if you want to mount the Raspberry Pi display in a frame/stand. This also allows you to attach the Raspberry Pi to the rear of the screen
• Raspberry Pi 12.5W Power Supply - you will need a decent power supply to power both the Pi and 7" screen

Software prerequisites:

• Home Assistant - running on the same local network (e.g. on the Raspberry Pi or a separate home server)
• GivTCP - this connects directly to your GivEnergy inverter and provides realtime data via an MQTT broker that Home Assistant is able to read. This needs to be running on the same local network as your inverter (e.g. on the Raspberry Pi or a separate home server)

Hardware Setup

The Raspberry Pi 7" touchscreen display attaches directly to the Raspberry Pi board using the included ribbon cable. The Pi can be mounted to the rear of the touchscreen using the included standoffs. The picture below shows the screen mounted within the Pibow Touchscreen Frame (detailed above):

Here is another close-up view:

Software Setup

1. Follow the installation instructions for GivTCP and set this up somewhere on your local network so that it's running 24/7. This runs inside a Docker container.
2. Once GivTCP is setup, you will need to define a few settings inside the docker-compose.yml:
   1. INVERTOR_IP - set the IP address of your GivEnergy inverter (assign a static IP to the inverter via your router if possible)
   2. NUM_BATTERIES - set the number of batteries connected to your GivEnergy inverter
   3. MQTT_OUTPUT - set this to True so Home Assistant is able to retrieve inverter data at regular intervals
3. If you don't already have Home Assistant, you will need to install this on a device on your local network, such as a Raspberry Pi, home server, or NAS device. This can also run inside of Docker.
4. In Home Assistant, go to Configuration >> Devices & Services >> Integrations. Click on Add Integration, and search for MQTT. In the settings for MQTT you will need to define the Broker options - this will be the IP address of the device running GivTCP. If this is on the same device as Home Assistant you can use 127.0.0.1. Also set the port number, e.g. 1883. This should detect all the different power and energy metrics provided by the GivEnergy inverter.
5. In the configuration.yaml for Home Assistant, you will need to enable the Utility Meter (for defining your tariffs), and create some sensors that the web app requires. Here are the full contents of my configuration.yaml as an example. You will also need to change the peak and off-peak tariff rates in the code below.

default_config:

tts:
  - platform: google_translate

utility_meter:
  daily_energy:
    source: sensor.override_import_energy_today_kwh
    name: Daily Energy
    cycle: daily
    tariffs:
      - peak
      - offpeak

script: !include scripts.yaml
scene: !include scenes.yaml
automation: !include automations.yaml

sensor:
  - platform: template
    sensors:
      override_import_energy_today_kwh:  # Sometimes GivTCP returns a 0 kWh import energy reading, causing Home Assistant to double up the numbers when it returns to normal
        unit_of_measurement: kWh
        friendly_name: Import Energy Today kWh
        value_template: >
          {% set value = states('sensor.givtcp_import_energy_today_kwh') | float %}
          {% if value == 0 %}
            {{ states('sensor.override_import_energy_today_kwh') }}
          {% else %}
            {{ value }}
          {% endif %}

  - platform: template
    sensors:
      daily_energy_cost_peak:
        unit_of_measurement: '£'
        friendly_name: Today's Peak Tariff Cost
        value_template: "{{ (states('sensor.daily_energy_peak') | float * 0.138 ) | round(2) }}"  # Define peak tariff rate here

  - platform: template
    sensors:
      daily_solar_generated_income:
        unit_of_measurement: '£'
        friendly_name: Today's Generated Solar Income
        value_template: "{{ (states('sensor.givtcp_pv_energy_today_kwh') | float * 0.138 ) | round(2) }}"  # Define peak tariff rate here

  - platform: template
    sensors:
      daily_energy_cost_offpeak:
        unit_of_measurement: '£'
        value_template: "{{ (states('sensor.daily_energy_offpeak') | float * 0.05 ) | round(2) }}"  # Define off-peak tariff rate
        friendly_name: Today's Off-Peak Tariff Cost

  - platform: template
    sensors:
      daily_energy_cost_all:
        unit_of_measurement: '£'
        value_template: "{{ (states('sensor.daily_energy_cost_peak') | float) + (states('sensor.daily_energy_cost_offpeak') | float) }}"
        friendly_name: Today's Total Cost

  - platform: template
    sensors:
      daily_energy_export_income:
        unit_of_measurement: '£'
        value_template: "{{ (states('sensor.givtcp_export_energy_today_kwh') | float * 0.041 ) | round(2) }}"  # Define export tariff rate
        friendly_name: Today's Export Income

  - platform: template
    sensors:
      battery_state:
        value_template: >-
          {% if states('sensor.givtcp_discharge_power') | float(0) > 0 %} Discharging
          {% elif states('sensor.givtcp_charge_power') | float(0) > 0 %} Charging
          {% else %} Idle
          {% endif %}
        friendly_name: Battery State

6. The configuration.yaml also references automations.yaml - create this file if it doesn't exist, and define the automations to ensure Home Assistant automatically switches tariffs at the appropriate times. For example I am on the Octopus Go tariff from Octopus Energy, which has an off-peak period of 12:30am to 4:30am:

- alias: Switch to off-peak tariff
  trigger:
  - platform: time
    at: "00:30:00"
  condition: []
  action:
  - service: utility_meter.select_tariff
    data:
      tariff: offpeak
    target:
      entity_id: utility_meter.daily_energy
  mode: single
  id: offpeak1

- alias: Switch to peak tariff
  trigger:
  - platform: time
    at: "04:30:00"
  condition: []
  action:
  - service: utility_meter.select_tariff
    data:
      tariff: peak
    target:
      entity_id: utility_meter.daily_energy
  mode: single
  id: peak1

7. The automations above will create triggers to automatically switch the tariff at the appropriate times, ensuring Home Assistant correctly calculates your energy costs throughout the day. Important: Check to make sure Home Assistant is set to the right timezone (e.g. London). If it's set to UTC for example, the tariff will switch at the wrong time. You can check the timezone setting under Configuration >> Settings >> General.

8. In Configuration >> Dashboards >> Energy, configure the sensors for Electricity grid, Solar panels, and Home battery storage. I used these settings:

   1. Grid consumption - GivTCP Import Energy Today kWh
   2. Return to grid - GivTCP Export Energy Today kWh
   3. Solar production - GivTCP PV Energy Today kWh
   4. Battery systems:
      1. Energy going in to the battery - GivTCP Battery Charge Energy Today kWh
      2. Energy coming out of the battery - GivTCP Battery Discharge Energy Today kWh

9. Restart Home Assistant and you should soon see data under the Overview and Energy sections.

10. Generate a Long-Lived Access Token within Home Assistant, via your User Profile screen >> Long-Lived Access Tokens. This is required by the GivEnergy smart home display web app.

11. On the Raspberry Pi, run this command in Terminal to clone the repository:

    git clone https://github.com/DanielGallo/GivEnergy-Smart-Home-Display.git

12. "cd" into the directory:

    cd GivEnergy-Smart-Home-Display

13. Run the install script:

    sudo ./install.sh

14. The install script will download and install dependencies, and configure the web server to launch on startup.

15. Copy the Home Assistant access token created in step 10 above and add it in app.json within the app folder.

16. Set the domain (and port number) of your Home Assistant server in app.json.

17. You can then launch Chromium and try running the app by visiting http://localhost:3000

18. If you want Chromium to launch the app automatically when the Pi boots up, run this command in Terminal:

    sudo nano /etc/xdg/lxsession/LXDE-pi/autostart

19. Add these lines to the above file - this will force Chromium to launch in kiosk mode at startup and load the smart home display app. The mouse cursor is also configured to hide after 5 seconds of inactivity, so you just see the web app.

    @xset s off
    @xset -dpms
    @xset s noblank
    @chromium-browser --kiosk http://localhost:3000
    @unclutter -idle 5 -root
    

20. Type Ctrl + X to exit, type y to confirm changes, then press Enter to save.

21. Reboot the Raspberry Pi by running this command in Terminal:

    reboot

22. On startup, Chromium should launch and the web app should be loaded.

23. If you have any feedback or spot an issue with the instructions above, please share your comments by creating a new issue here.

Current Issues and Limitations

• There is normally a ~15% disparity between the reported battery discharge rate (givtcp_discharge_power) and the load rate (givtcp_load_power), for example the battery may show a discharge rate of 1.6 kW, but the house is only using 1.4 kW. Perhaps the GivEnergy team know why this might be happening - is it simply the loss of converting DC to AC, or an issue with GivTCP?
• Sometimes the arrows show power flow, when the sources (e.g. grid) are 0.00 kW - need to zero-out these values

Upcoming Plans

Here are the upcoming plans for this smart home display app:

• Investigate alternative to Home Assistant to ease setup - maybe a custom local process that logs stats and calculates daily costs based on the time of use (peak vs. off-peak)
• Have all the dependencies container-based, making it easier and quicker to set everything up on the Raspberry Pi, without needing to setup external dependencies
• Preconfigure as many dependencies as possible, to minimise the number of configuration steps
• Add error handling (show friendly errors) when things go wrong, e.g. wrong URL, no access token, etc.
• Implement a power saving mode to allow the user to define the hours when the screen is powered off (see below)

How to

Launch app in full screen mode on startup

See https://raspberrypi.stackexchange.com/a/76336.

Turn the screen off

See https://community.home-assistant.io/t/how-to-turn-off-raspberry-pi-screen/126060/5.

This could be automated via a Cron task (e.g. turn the display off at 6pm every day). Tapping the Raspberry Pi 7" touchscreen turns it back on.

Other Notes

• SVG icons were designed and generated using https://yqnn.github.io/svg-path-editor/