CoolFace
Datasetpublic

PGIQ/pgiq-ratings

PowerGridIQ Rating dataset (headline) Free, redistributable headline ratings for 72 global power markets and sub-regions, for deciding where to site or when to schedule a large new electricity load (data centers, AI compute clusters, heavy industry). This is the open headline slice of the ratings behind PowerGridIQ. The rating is a proprietary, directional screen; evidence is drawn from public operator and regulator sources, and some inputs are modelled. What this… See the full description on the dataset page: https://huggingface.co/datasets/PGIQ/pgiq-ratings.

sourceHugging Facecc-by-4.0updated 3mo agoView on Hugging Face
0likes21downloads
Dataset Card

PowerGridIQ Rating dataset (headline)

Free, redistributable headline ratings for 72 global power markets and sub-regions, for deciding where to site or when to schedule a large new electricity load (data centers, AI compute clusters, heavy industry).

This is the open headline slice of the ratings behind PowerGridIQ. The rating is a proprietary, directional screen; evidence is drawn from public operator and regulator sources, and some inputs are modelled.

What this file contains

Per market: a tier, a reliability-adjusted pgiq_rating, a five-pillar fundamentals score, an outlook, a confidence level, and an overall reliability_level.

  • —fundamentals is the five-pillar score (Access, Availability, Cost, Momentum, Carbon).
  • —pgiq_rating is the reliability-adjusted rating: fundamentals minus a reliability dock.
  • —Tier 1 (Prime) is relative. Current Prime markets: Sweden North, United Arab Emirates, Norway.
  • —Sub-regions are rated where a country's internal divergence is decision-relevant (Germany, Japan, Sweden, Australia); country overviews are excluded.

The five individual pillar sub-scores, the detailed reliability reads, the point-in-time history and the interconnection-queue detail are not in this file. They are proprietary and available via the API for evaluation (not for redistribution) and under a commercial license for production use. Contact hello@powergridiq.com.

Columns

id, name, group, parent, tier, tier_name, pgiq_rating, fundamentals, reliability_level, reliability_dock, reliability_assessed, reliability_confidence, outlook, under_review, report_url, as_of.

Load it

python
from datasets import load_dataset
ds = load_dataset("PGIQ/pgiq-ratings")
df = ds["train"].to_pandas()
print(df[df.tier == 1][["name", "pgiq_rating", "fundamentals", "reliability_level"]])

License and attribution

The headline ratings in this file are licensed Creative Commons Attribution 4.0 (CC BY 4.0): free to use, share, adapt and redistribute, including commercially, with credit to PowerGridIQ Rating (https://powergridiq.com). The pillar sub-scores, detailed reliability reads, point-in-time history and queue detail are proprietary and are not covered by this license.

Source and updates

Maintained from powergridiq.com. Live JSON: https://powergridiq.com/data/pgiq-ratings.json. Methodology: powergridiq.com/methodology. The as_of field records the refresh date.