Use electricity price and CO2 in Node-RED without unnecessary clicking

Core logic explained early

The price logic is not “forecast only”. It deliberately combines real market data with forecast values so a flow does not react unnecessarily to pure estimates.

Recommended entry point: summary

Node-RED has its own summary path with current values, best windows and meta fields for limits and API key state. That is the most sensible starting point for first automations.

Alternative for your own price logic: node-red/prices

For more advanced flows there is also a second Node-RED path that returns only the price series. It is intended for your own window calculations, charts, thresholds, or combinations with PV, battery and house status.

What the API actually returns

The summary endpoint returns a compact automation view. That is useful for Node-RED because you can use it immediately for simple flows while still keeping enough structure for later extensions.

If you need a real price series for your own calculations instead of the compact summary, use /api/v1/node-red/prices. The summary intentionally stays compact. The price series is intended for advanced custom flow logic.

Price series for your own flows: node-red/prices

The extra price-series path intentionally returns prices only, without CO2 fields and without precomputed window logic. That is exactly the point: in Node-RED you can decide for yourself how to evaluate or combine the slots with other signals.

https://api.energypriceforecast.eu/api/v1/node-red/prices?country=de&hours=48&mode=mixed&resolution=15m

https://api.energypriceforecast.eu/api/v1/node-red/prices?country=de&hours=48&mode=mixed&resolution=15m&price_mode=retail&plz=10115

Important: in 15m mode, forecast values are not smoothed between hours. The API does not invent intermediate prices. Instead, it clearly marks that those quarter-hour slots come from one hourly forecast block.

Base price or total retail price?

The Node-RED summary path still deliberately returns base or market prices. The prices path also returns base prices by default, but can optionally output assumption-based retail totals.

Time horizon and resolution

The examples use hours=48 and, for the summary, window_hours=4. That is a sensible starting point, but not the full product limit.

source.day_ahead_entries and source.forecast_entries count slots, not hours. That is why values such as 97 day-ahead and 0 forecast can still be perfectly plausible.

Where to put the API key

With the native node, enter the optional API key once in the shared Energy Price Forecast connection. Node-RED stores it as a credential and does not include it when exporting a flow. Without a key, requests use the free horizon; an entitled key can unlock values such as hours=120.

const API_KEY = "";
const API_URL = "https://api.energypriceforecast.eu/api/v1/node-red/summary?country=de&hours=120&window_hours=4";

The code applies only to the alternative standard-node flow below. Do not share such an exported flow with a key filled in. The authoritative limit is not the requested URL value but meta.allowed_horizon_hours.

Quick start with the native package

1. Open Menu → Manage palette → Install in Node-RED.
2. Search for @backupbattery/node-red-energypriceforecast and install the package.
3. Add and connect Inject, EPF summary and Debug nodes.
4. Create an Energy Price Forecast connection in the summary node, select the market and horizon, then deploy.

Alternatively, run the command in the Node-RED user directory and restart Node-RED afterwards. The package requires Node.js 20 or newer.

Ready-made signals instead of Function code

Place EPF signal directly after the summary node. Output 1 means “condition matched”; output 2 means “condition did not match”. The original API response remains intact.

For start/stop events, the signal node can emit only on state changes. By default its first input establishes the initial state without treating a deploy as a new event.

Cheapest individual hours for heat pumps and EVs

EPF planner processes the output from EPF prices. It can select, for example, the 12 cheapest individual hours in every 24-hour block or four charging hours during the weekend.

Block boundaries automatically use the electricity market's local time zone and remain correct across daylight-saving changes. Enable a persistent Node-RED context store if plans must survive restarts.

Alternative without the package: import the summary flow

1. Click Import in the top right corner of Node-RED.
2. Paste the complete JSON block below.
3. Import it and then click Deploy.
4. Open the debug sidebar and trigger the Inject node once manually.

Import flow for the summary

Copy the full block and import it into Node-RED. In addition to debug data, the flow creates a safe Boolean signal under energypriceforecast/consumer_release for a wallbox, relay or further logic.

[
  {
    "id": "epf_summary_tab",
    "type": "tab",
    "label": "EPF Summary",
    "disabled": false,
    "info": ""
  },
  {
    "id": "epf_summary_inject",
    "type": "inject",
    "z": "epf_summary_tab",
    "name": "Load summary",
    "props": [],
    "repeat": "1800",
    "once": true,
    "onceDelay": "1",
    "topic": "",
    "x": 130,
    "y": 120,
    "wires": [["epf_summary_prepare"]]
  },
  {
    "id": "epf_summary_prepare",
    "type": "function",
    "z": "epf_summary_tab",
    "name": "Set API key and URL",
    "func": "const API_KEY = \"\";\nmsg.url = \"https://api.energypriceforecast.eu/api/v1/node-red/summary?country=de&hours=48&window_hours=4\";\nmsg.headers = {};\nif (API_KEY.trim()) {\n    msg.headers.Authorization = \"Bearer \" + API_KEY.trim();\n}\nreturn msg;",
    "outputs": 1,
    "noerr": 0,
    "initialize": "",
    "finalize": "",
    "libs": [],
    "x": 390,
    "y": 120,
    "wires": [["epf_summary_http"]]
  },
  {
    "id": "epf_summary_http",
    "type": "http request",
    "z": "epf_summary_tab",
    "name": "Fetch summary",
    "method": "GET",
    "ret": "obj",
    "paytoqs": "ignore",
    "url": "",
    "persist": false,
    "x": 630,
    "y": 120,
    "wires": [["epf_summary_extract"]]
  },
  {
    "id": "epf_summary_extract",
    "type": "function",
    "z": "epf_summary_tab",
    "name": "Expose key fields",
    "func": "const status = Number(msg.statusCode);\nif (!Number.isFinite(status) || status < 200 || status >= 300) {\n    msg.is_cheapest_window_now = false;\n    msg.automation_signal_fresh = false;\n    msg.error = 'HTTP ' + (Number.isFinite(status) ? status : 'unknown');\n    msg.payload = { error: msg.error };\n    return [null, msg];\n}\nconst data = msg.payload || {};\nconst flat = data.flat || {};\nconst meta = data.meta || {};\nconst price = data.price || {};\nconst co2 = data.co2 || {};\nmsg.current_price = flat.current_price;\nmsg.current_co2_g_kwh = flat.current_co2_g_kwh;\nmsg.cheapest_window_start = flat.cheapest_window_start || null;\nmsg.greenest_window_start = flat.greenest_window_start || null;\nmsg.is_cheapest_window_now = flat.is_cheapest_window_now === true;\nmsg.cheapest_window_remaining_minutes = flat.cheapest_window_remaining_minutes ?? null;\nmsg.next_full_cheapest_window_start = price.next_full_window ? price.next_full_window.start : null;\nmsg.next_full_greenest_window_start = co2.next_full_window ? co2.next_full_window.start : null;\nmsg.api_key_state = meta.api_key_state || 'missing';\nmsg.allowed_horizon_hours = meta.allowed_horizon_hours;\nmsg.used_horizon_hours = meta.used_horizon_hours;\nmsg.rate_limit_daily = meta.rate_limit_daily;\nmsg.used_calls_today = meta.used_calls_today;\nmsg.automation_signal_fresh = true;\nmsg.error = null;\nreturn [msg, null];",
    "outputs": 2,
    "noerr": 0,
    "initialize": "",
    "finalize": "",
    "libs": [],
    "x": 870,
    "y": 120,
    "wires": [["epf_summary_release", "epf_summary_debug"], ["epf_summary_release", "epf_summary_error_debug"]]
  },
  {
    "id": "epf_summary_debug",
    "type": "debug",
    "z": "epf_summary_tab",
    "name": "Show summary",
    "active": true,
    "tosidebar": true,
    "console": false,
    "tostatus": false,
    "complete": "true",
    "targetType": "full",
    "x": 1110,
    "y": 120,
    "wires": []
  },
  {
    "id": "epf_summary_release",
    "type": "function",
    "z": "epf_summary_tab",
    "name": "Safe consumer release",
    "func": "msg.topic = 'energypriceforecast/consumer_release';\nmsg.payload = msg.automation_signal_fresh === true && msg.is_cheapest_window_now === true;\nreturn msg;",
    "outputs": 1,
    "noerr": 0,
    "initialize": "",
    "finalize": "",
    "libs": [],
    "x": 1120,
    "y": 180,
    "wires": [["epf_summary_release_debug"]]
  },
  {
    "id": "epf_summary_release_debug",
    "type": "debug",
    "z": "epf_summary_tab",
    "name": "Release consumer (Boolean)",
    "active": true,
    "tosidebar": true,
    "console": false,
    "tostatus": true,
    "complete": "payload",
    "targetType": "msg",
    "x": 1400,
    "y": 180,
    "wires": []
  },
  {
    "id": "epf_summary_catch",
    "type": "catch",
    "z": "epf_summary_tab",
    "name": "Catch errors",
    "scope": null,
    "uncaught": false,
    "x": 140,
    "y": 220,
    "wires": [["epf_summary_error"]]
  },
  {
    "id": "epf_summary_error",
    "type": "function",
    "z": "epf_summary_tab",
    "name": "Make error readable",
    "func": "const message = msg.error && msg.error.message ? msg.error.message : 'Unknown error';\nmsg.is_cheapest_window_now = false;\nmsg.automation_signal_fresh = false;\nmsg.payload = { error: message, source: msg.error && msg.error.source ? msg.error.source : null };\nreturn msg;",
    "outputs": 1,
    "noerr": 0,
    "initialize": "",
    "finalize": "",
    "libs": [],
    "x": 400,
    "y": 220,
    "wires": [["epf_summary_release", "epf_summary_error_debug"]]
  },
  {
    "id": "epf_summary_error_debug",
    "type": "debug",
    "z": "epf_summary_tab",
    "name": "Summary error",
    "active": true,
    "tosidebar": true,
    "console": false,
    "tostatus": true,
    "complete": "payload",
    "targetType": "msg",
    "x": 660,
    "y": 220,
    "wires": []
  }
]

For testing, the release is visible in the Debug sidebar. Then connect the Boolean node to your own device or MQTT logic. Test both true and false before connecting a real load.

Advanced: import the price-series flow

If you prefer to calculate things yourself, the second flow is intended for the raw price series. It stores the full slot list in msg.price_entries.

[
  {
    "id": "epf_prices_tab",
    "type": "tab",
    "label": "EPF Prices",
    "disabled": false,
    "info": ""
  },
  {
    "id": "epf_prices_inject",
    "type": "inject",
    "z": "epf_prices_tab",
    "name": "Load price series",
    "props": [],
    "repeat": "1800",
    "once": false,
    "onceDelay": "1",
    "topic": "",
    "x": 130,
    "y": 120,
    "wires": [["epf_prices_prepare"]]
  },
  {
    "id": "epf_prices_prepare",
    "type": "function",
    "z": "epf_prices_tab",
    "name": "Set parameters",
    "func": "const API_KEY = \"\";\nconst BASE_URL = \"https://api.energypriceforecast.eu/api/v1/node-red/prices?country=de&hours=48&mode=mixed&resolution=15m\";\nconst USE_RETAIL = false;\nconst PLZ = \"10115\";\nmsg.url = USE_RETAIL ? BASE_URL + \"&price_mode=retail&plz=\" + encodeURIComponent(PLZ) : BASE_URL;\nmsg.headers = {};\nif (API_KEY.trim()) {\n    msg.headers.Authorization = \"Bearer \" + API_KEY.trim();\n}\nreturn msg;",
    "outputs": 1,
    "noerr": 0,
    "initialize": "",
    "finalize": "",
    "libs": [],
    "x": 360,
    "y": 120,
    "wires": [["epf_prices_http"]]
  },
  {
    "id": "epf_prices_http",
    "type": "http request",
    "z": "epf_prices_tab",
    "name": "Fetch price series",
    "method": "GET",
    "ret": "obj",
    "paytoqs": "ignore",
    "url": "",
    "persist": false,
    "x": 610,
    "y": 120,
    "wires": [["epf_prices_extract"]]
  },
  {
    "id": "epf_prices_extract",
    "type": "function",
    "z": "epf_prices_tab",
    "name": "Forward series",
    "func": "const data = msg.payload || {};\nmsg.price_entries = Array.isArray(data.entries) ? data.entries : [];\nmsg.first_price_slot = msg.price_entries.length ? msg.price_entries[0] : null;\nmsg.price_series_source = data.source || {};\nmsg.api_key_state = data.meta && data.meta.api_key_state ? data.meta.api_key_state : 'missing';\nmsg.allowed_horizon_hours = data.meta ? data.meta.allowed_horizon_hours : null;\nmsg.series_request = data.request || {};\nmsg.series_assumptions = data.assumptions || null;\nmsg.error = null;\nreturn msg;",
    "outputs": 1,
    "noerr": 0,
    "initialize": "",
    "finalize": "",
    "libs": [],
    "x": 860,
    "y": 120,
    "wires": [["epf_prices_debug"]]
  },
  {
    "id": "epf_prices_debug",
    "type": "debug",
    "z": "epf_prices_tab",
    "name": "Show price series",
    "active": true,
    "tosidebar": true,
    "console": false,
    "tostatus": false,
    "complete": "true",
    "targetType": "full",
    "x": 1110,
    "y": 120,
    "wires": []
  }
]

For retail mode in Germany, set USE_RETAIL = true and enter a suitable PLZ postcode in the Function node.

For automations: separate display and execution logic

For dashboards, msg.cheapest_window_start may already be enough. For real control decisions, however, it is often more important to know whether the best window is already active. That is why the summary flow also sets msg.is_cheapest_window_now and msg.cheapest_window_remaining_minutes.

if (msg.is_cheapest_window_now) {
  // enable the consumer now
} else if (msg.next_full_cheapest_window_start) {
  // use the next full window for planning
}

return msg;

Pragmatically, a boolean for “now” is usually more valuable for automations than only a timestamp in the future.

Which message fields you will have afterwards

Important notes