TenneT AFRR Bid Ladder Archive

A Python application for downloading and archiving TenneT's AFRR (Automatic Frequency Restoration Reserve) bidladders at regular intervals.

Overview

This project automatically downloads and archives bidladder snapshots from TenneT's API, which only provides the latest data. The archive enables historical analysis of ladder changes over time.

Sampling

To keep the size of the archive manageable the sampling frequency varies by hours-to-delivery (HTD):

• 12h-24h: Every 2 hours (HTD = 14, 16, ..., 22)
• 3h-12h: Every hour (HTD = 3, 4,, ..., 12)
• <3h: Every 15 minutes (HTD = 0, 0.25, 0.5, ..)

This results in uniform hours-to-delivery across ISPs.

Note: Runs on VPS using cron-scheduled bash script for reliable execution timing.

Data Structure

data/
└── YYYY-MM-DD/
    └── HHMM_DST/
        └── snapshot_HHMM_htd_X.Xh.parquet

The data is stored based on the date and time of the ISP. Each snapshot is saved in a Parquet file with a filename that includes the time of the snapshot and the hours-to-delivery (htd).

• YYYY-MM-DD: ISP date
• HHMM_DST: ISP time (24h format) with DST indicator (CET or CEST)
• snapshot_HHMM_htd_X.XX.parquet: Snapshot data, where:
  • HHMM: Time of the snapshot
  • X.XX: Hours-to-delivery

Parquet file contents:

• index: ISP start timestamp
• capacity_threshold
• price_down
• price_up
• snapshot_timestamp: Exact snapshot timestamp
• minutes_to_delivery: Minutes until ISP start (rounded to nearest quarter hour within 5-minute tolerance)

Acknowledgements

Thanks to fboerman for tenneteu-py!