Attribution & licensing

Where the data comes from, what you may redistribute, and what we deliberately do not publish.

Nothing on this page is legal advice. It is a plain account of where each series comes from and what we believe you may do with it. You can make your own call, and a procurement reviewer can follow the chain without emailing us.

The chain

Every public series traces back to the ENTSO-E Transparency Platform. Two are series ENTSO-E publishes, and the third is our sum of one of them. The Energy Data page lists which series are served in which zones.

• imbalance-prices: ENTSO-E’s imbalance settlement prices. As Article 17 balancing data under the Transparency Regulation, they sit on ENTSO-E’s free re-use list under CC-BY 4.0, and it is the same item in every zone we serve.
• balancing-state: ENTSO-E’s Current Balancing State, one value a minute per control area. ENTSO-E publishes it under Article 12.3(a) of the Electricity Balancing Guideline. That item is on the same free re-use list under CC-BY 4.0, and it is the same item in every zone we serve.
• imbalance-volumes: a series we derive. Each value is the sum of balancing-state’s minutes over one 15-minute imbalance settlement period in the Netherlands, divided by 60, in MWh. The sums are ours, not an ENTSO-E figure. The series carries CC-BY 4.0 from its input.

The free re-use list makes both upstream series freely redistributable with attribution. We fetch them, store the raw payload immutably, and derive everything we serve from that stored copy. We never discard or rewrite a stored payload, so you can trace any value we publish back to the exact upstream document it came from. For imbalance volumes, those are the documents behind the minutes it sums.

What you may redistribute

Every series we hold carries a licence tag. We record the tag at ingest, from the terms we obtained the series under. The public API serves only series tagged as redistributable. That filter is fail-closed: the API refuses to serve a series with an unrecognised or missing tag, rather than serving it by default. Whatever you can fetch from api.tolven.eu, we believe you may use and pass on under the upstream terms, with attribution. A key changes how much you can fetch, not those terms.

Credit each series with the attribution string that every series response carries. For imbalance-prices and balancing-state it reads “ENTSO-E Transparency Platform, CC-BY 4.0”. It credits the source and names the licence, as CC-BY 4.0 requires. For imbalance-volumes it reads “Derived from ENTSO-E Transparency Platform Current Balancing State, CC-BY 4.0; summed per 15-minute ISP by Tolven”. That string names us because we did the summing, so nobody takes the sums for an ENTSO-E figure.

We ask for no credit beyond the string the API returns. We claim no rights over ENTSO-E’s data. Whatever rights we hold in the sums in imbalance-volumes, we license under CC-BY 4.0, the terms of its input, so its credit string covers both. If you want to credit the cleaning and revision history as well, “via Tolven Energy Data” is welcome but never required.

What we deliberately do not publish

NL day-ahead prices. We do not serve them, and the reason is licensing rather than effort.

Dutch day-ahead prices are set on EPEX SPOT through Single Day-Ahead Coupling (SDAC). ENTSO-E does carry them, but that document is not on its CC-BY free re-use list: ENTSO-E’s terms make it free to access and not free to redistribute. We therefore cannot build anything public on it. The public API serves no day-ahead series, and the fail-closed licence filter above is what keeps it that way.

SMARD publishes day-ahead under CC-BY 4.0, but only for the German market areas: DE-LU, AT and LU. Its day-ahead API does not serve NL or the other neighbouring zones. The “neighbouring countries” panel on the SMARD website does display NL, but that is a front-end view rather than a downloadable CC-BY series.

That leaves the downstream republishers that appear to offer NL day-ahead under CC-BY. They assert that licence themselves, and we could not verify the chain behind it. One stamps its NL series as coming from SMARD while its own chart page credits ENTSO-E as the source. None, as far as we can verify, holds a right to sublicense the underlying EPEX/ENTSO-E data. The rights holders may tolerate that for public research; we could not rely on it commercially, so we do not build on it.

Publishing NL day-ahead properly means a distribution licence from the exchange, negotiated case by case rather than bought off a shelf. The line there is redistribution, not price. Handing the data to a third party is redistribution whether we charge for it or not, so no arrangement sidesteps it by giving the data away.

If you need NL day-ahead today, license it directly. If a German benchmark will do, SMARD publishes DE-LU day-ahead openly and you do not need us in the middle.

Until such a licence is in place, we declare the absence rather than fill it — the same principle we apply to gaps in the data. We would rather serve nothing than data we cannot defend.

Gaps and revisions are part of the record

We report gaps, never interpolate them. Where the upstream record has an interior gap, /v1/gaps/{dataset}/{zone} will tell you so. We do not fill gaps with plausible numbers.

We keep revisions, never overwrite them. TenneT restates settlement, and we store each restated value as a new revision rather than replacing what came before. You can audit our history against the upstream publication timeline.

We treat a plausible number you cannot distinguish from a real one as worse than a gap.

Questions

If this page does not cover something you need for a procurement or compliance review, email the operator. The address reaches the engineer who runs Tolven, not a ticket queue.