How to know it's healthy
UK Grid is a small, scheduled system. There's no fancy status page because there's nothing complicated to surface. Here's where to look and what to do if something seems off.
Live freshness signal
Open the dashboard and look at the top-right chip in the header:
- green dot · updated HH:MM — last ingest was within the last 10 minutes
- amber dot · stale · Xm old — the 5-minute ingest hasn't run in the last 10 minutes
- amber dot · syncing… — first load, snapshot not yet retrieved
What runs, and when
| Function | Schedule | What it does |
|---|---|---|
ingestLive |
Every 5 min | Pulls the last 90 min from Elexon + Carbon Intensity + NESO, writes the live snapshot, regional snapshot, and the last 3 settlement periods to Firestore. |
gapCheck |
Daily at 03:30 UTC | Scans the last 7 days for any settlement period missing fields and backfills from the authoritative APIs. |
ingestHistorical |
Manual | HTTPS endpoint for bulk backfill across arbitrary date ranges. Token-gated, not publicly callable. |
If the dashboard looks stuck
- Check the header chip. If it's amber, the next ingest tick (within ~5 min) usually clears it.
- Check whether the Carbon Intensity API, Elexon BMRS, and NESO portal are themselves responding. If an upstream is down, we will be too.
- If the chip stays amber for >15 min and the upstream APIs are up, something's wrong on our side. Email hello@andelle.co or open an issue.
Why some fields lag the live time
A settlement period's record fills in over ~60 minutes as upstream feeds publish. The earliest field to land is generation (~1–3 min after the start of a 5-min sub-tick). The wholesale price lands 10–25 min into the half-hour. The settled carbon actual lands 30–60 min after the period ends. NESO's authoritative demand can be hours behind. The dashboard never blanks a card — it shows the latest available value and flags staleness when the next print is overdue.
See the Data Sources & Methodology page for the per-source publish cadence in detail.