techem-scraper

Purpose

Scrapes water/heating consumption data ("Verbrauchsinfo") from mieter.techem.de.

Background

Techem offers a portal to show the "Verbrauchsinfo" to tenants to inform them about their energy consumption. The "Verbrauchsinfo" is a monthly generated usage report for heating and cooling as well as cold and warm water.

If you live in a multi-tenant building, Techem might be the provider for water and heating meters.

Warning

Unfortunately Techem does not offer an API client, so this is an experimental way of fetching it programmatically. This project is not affiliated with Techem in any shape or form. The backend might change any time, so use at your own risk.

--help

usage: techem.py [-h] [-B] [-u USERNAME] [-p PASSWORD]

Scrapes consumption data (Verbrauchsinfo) from Techem website

options:
  -h, --help            show this help message and exit
  -B, --browser
  -u, --username USERNAME
  -p, --password PASSWORD

How to use

Have Python 3.x and the MSAL python library installed.

Either, you provide your login to the script, which will then use selenium and chrome headless to log You in:

./techem-eed-scraper.py -u email@example.com -p mysecurepassword

or You need to be logged out of the mieter.techem.de portal in your browser (no active session) and run

./techem.py -B

How it works

The portal mieter.techem.de manages user logins via Azure Active Directory B2C. It uses an OAuth 2.0 Auth Code Flow, for which the scraper will open a web browser for you to log in with your credentials.

After copying in the Redirect URL (starts with https://mieter.techem.de/auth?state), the script will automatically fetch your consumption data and create three types of files:

• period.json (containing the periods you have stored)
• YYYY-MM.json (containing the consumption data of the specific month)
• YYYY-MM-average.json (containing the consumption of other comparably-sized appartments

Sample data

The data provided by the API looks like this:

{
    "count": 4,
    "data": [
        {
            "period": "2023-01",
            "service": "HOT_WATER",
            "unitOfMeasure": "M3",
            "amount": 1.9,
            "quality": 1.0,
            "revision": 2,
            "status": "OK",
            "createdAt": "2025-04-27T17:13:07Z"
        },
        {
            "period": "2023-01",
            "service": "HOT_WATER",
            "unitOfMeasure": "KWH",
            "amount": 223.0,
            "quality": 1.0,
            "revision": 2,
            "status": "OK",
            "createdAt": "2025-04-27T17:13:07Z"
        },
        {
            "period": "2023-01",
            "service": "HEATING",
            "unitOfMeasure": "HCU",
            "amount": 2504.0,
            "quality": 1.0,
            "revision": 2,
            "status": "OK",
            "createdAt": "2025-04-27T17:13:07Z"
        },
        {
            "period": "2023-01",
            "service": "HEATING",
            "unitOfMeasure": "KWH",
            "amount": 1912.9,
            "quality": 1.0,
            "revision": 2,
            "status": "OK",
            "createdAt": "2025-04-27T17:13:07Z"
        }
    ]
}

The comparable data for average looks like:

{
    "count": 2,
    "revision": 2,
    "data": [
        {
            "service": "HEATING",
            "unitOfMeasure": "KWH",
            "amount": 2392.92,
            "createdAt": 1677926811000,
            "status": "OK"
        },
        {
            "service": "HOT_WATER",
            "unitOfMeasure": "KWH",
            "amount": 288.911507,
            "createdAt": 1677926811000,
            "status": "OK"
        }
    ]
}

Interesting findings about the backend

• Blocks user agents like python requests
• If the measurement for certain reasons is not plausible, the website won't show it. The API response shows the consumption with a status EED_NE_BLACKLIST_IMPLAUSIBLE.