Monitor your power usage and costs with Prometheus and Grafana.
This prometheus exporter will connect to the Tibber API, subscribe to updates from Tibber Pulse or Watty and make the metrics available for Prometheus.
See the provided Grafana dashboard for examples on how they can be used.
Note that the consumption, cost and price heatmap panels requires the Hourly heatmap plugin.
If you don't have a device for live measurements, only the power price metrics will be available on that dashboard. It is possible to create a dashboard with historic consumption and cost using the ..._previous_day and ..._previous_hour metrics, but the availability of those metrics may vary between grid companies. See the dashboard without pulse for an example using the previous day metrics:
These metrics can also be used for a monthly overview dashboard:
The Tibber API provides power prices for the current and the next day, but storing future data can't be done easily in Prometheus. To be able to show future power prices, this exporter also has a JSON service which can be queried to get this data. In Grafana this can be used with the JSON API datasource plugin.
| Endpoint | Description |
|---|---|
/homes |
Available homes |
/homes/<home id>/prices |
Power prices for today and tomorrow (if available) |
/homes/<home id>/prices?period=now |
Power prices from now and through tomorrow (if available) |
/homes/<home id>/prices?period=today |
Power prices for today |
/homes/<home id>/prices?period=tomorrow |
Power prices for tomorrow (if available) |
Next day prices are usually available around 13:00 each day.
See the prices dashboard for some example panels which use this datasource.
Docker image is available on ghcr.io.
Prebuilt binaries for Linux, MacOS and Windows are available on the release page.
A TIBBER_TOKEN is required to use the Tibber API, go to developer.tibber.com to find yours.
Regardless of how you run this program, it is important to run it with an automatic restart mechanism. If the live feed from Tibber for some reason is interrupted or not available, the program will take a short pause and exit. The pause is to avoid a restart loop and trigger rate limiting in the Tibber API, and just exiting and restarting is better than trying to handle every possible error situation.
The automatic restart on errors in the live feed will make the exporter recover from most error situations. However, there is a bug in the Tibber service which sometimes fails to report that a home has the live feed feature. To handle this situation, the exporter has a command line option --live that can be used to force the exporter to always start the live feed. Therefore, if you have the live feature for your home, it's recommended to run the exporter with --live <home id> or set the environment variable TIBBER_LIVE_ENABLE_HOMES=<home id> to ensure that the live feed is always started.
If you don't already run Grafana and Prometheus, you can try a complete setup with docker compose.
cd docker-compose
TIBBER_TOKEN=... docker compose upThen go to http://localhost:3000/ and find the dashboards from the menu.
Install in your kubernetes cluster with Helm. First add the the helm repository:
helm repo add tibber-exporter https://terjesannum.github.io/tibber-exporter/
helm repo update
Then install the helm chart:
helm install tibber-exporter tibber-exporter/tibber-exporter --set tibberToken=...
This with install the exporter with the prometheus.io/scrape annotation set to true. If you run the Prometheus operator, install with serviceMonitor.enabled=true to create a ServiceMonitor instead:
helm install tibber-exporter tibber-exporter/tibber-exporter --set tibberToken=... --set serviceMonitor.enabled=true
docker run -d -p 8080:8080 -e TIBBER_TOKEN=... --restart always ghcr.io/terjesannum/tibber-exporter:latestRun with your Tibber token in the TIBBER_TOKEN environment variable, or specify it with the -token flag:
tibber-exporter -token ...Prometheus need to be configured to scrape the exporter, so add a scrape job to /etc/prometheus/prometheus.yml:
scrape_configs:
- job_name: "tibber-exporter"
scrape_interval: 5s
static_configs:
- targets: ["localhost:8080"]
How often the data is updated depends on your energy meter. Look at the logs from the exporter to see how often it receives updates, and adjust the scrape interval. Shorter scrape interval generates more data, so consider scraping less frequent if you only use a dashboard with a wide time range and don't need "live" updates.
Also remember that Prometheus is designed for monitoring and not precise calculation, so don't expect the result of the queries to excactly match your electricity bill.
Import the dashboard or use id 16804 and import from grafana.com. Then select the Prometheus datasource that scrapes the exporter.
| Name | Description | Type |
|---|---|---|
| tibber_home_info | Home info | Gauge |
| tibber_grid_info | Power grid info | Gauge |
| tibber_power_price | Current power price | Gauge |
| tibber_power_price_level | Current price level | Gauge |
| tibber_power_price_next_hour | Power price next hour | Gauge |
| tibber_power_price_level_next_hour | Price level next hour | Gauge |
| tibber_power_consumption | Current power consumption | Gauge |
| tibber_power_consumption_day_avg | Average power consumption today | Gauge |
| tibber_power_consumption_day_min | Minimum power consumption today | Gauge |
| tibber_power_consumption_day_max | Maximum power consumption today | Gauge |
| tibber_power_consumption_day_total | Total power consumption today | Counter |
| tibber_power_consumption_previous_hour | Total power consumption previous hour | Gauge |
| tibber_power_consumption_previous_day | Total power consumption yesterday | Gauge |
| tibber_power_consumption_reactive | Current reactive consumption | Gauge |
| tibber_power_cost_day | Total power cost today | Gauge |
| tibber_power_cost_day_total | Total power cost today | Counter |
| tibber_power_cost_previous_hour | Total power cost previous hour | Gauge |
| tibber_power_cost_previous_day | Total power cost yesterday | Gauge |
| tibber_power_production | Current power production | Gauge |
| tibber_power_production_day_min | Minimum power production today | Gauge |
| tibber_power_production_day_max | Maximum power production today | Gauge |
| tibber_power_production_day_total | Total power production today | Counter |
| tibber_power_production_previous_hour | Total power production previous hour | Gauge |
| tibber_power_production_previous_day | Total power production yesterday | Gauge |
| tibber_power_production_reward_day | Total power production reward today | Gauge |
| tibber_power_production_reward_day_total | Total power production reward today | Counter |
| tibber_power_production_reactive | Current net reactive production | Gauge |
| tibber_power_factor | Current power factor (active power / apparent power) | Gauge |
| tibber_power_profit_previous_hour | Total power profit previous hour | Gauge |
| tibber_power_profit_previous_day | Total power profit yesterday | Gauge |
| tibber_voltage | Phase voltage | Gauge |
| tibber_current | Line current | Gauge |
| tibber_signal_strength | Tibber Pulse signal strength | Gauge |
| tibber_exporter_build_info | Tibber exporter build and version info | Gauge |
More infomation can be found in the Tibber API documentation.
I would be happy if you use this referral code. That will give each of us a bonus to use on stuff like Tibber Pulse 😄
Or register the code qandobma directly in the app if you signed up without an invite.