The impact study Sidney never received
The environmental and economic study that should precede any abatement or rezoning — assembled from the public record. Where the record is silent, the silence is the finding.
Under construction AWS Sidney Data Center Campus · Amazon Web Services, Inc. (operator); Amazon Data Services, Inc. (developer)
7 of 15 chapters on the record · 6 partial · 2 gaps named
Sidney’s record is unusually complete for a campus this new. Eight council resolutions, four executed agreements, a grading permit, five state permits and the approved minutes of every meeting that voted on them are all in this corpus, most of them published by the City itself. What the record is missing is not documents. It is numbers the documents were never required to carry.
Five state permits have issued or been filed for this campus and not one states a load. The megawatt figure this study most needs exists — it is in an AES Ohio construction service agreement filed with FERC — and it is filed withheld. So the IT load here is a screening bracket, and every chapter that depends on it says so rather than borrowing a number from a press release.
The fiscal record inverts that. There the instruments are exact, and what they are exact about is how little they bind. A thirty-year, hundred-percent abatement per building; a $46,000,000 PILOT schedule to the dollar; an investment figure and a jobs figure that the same agreement states, in terms, cannot be compelled; no clawback; and a company right to terminate on thirty days’ notice for any reason. The precision is real. It runs one way.
Three things this study can now say that it could not a week ago, all because the City’s own legislative portal turned out never to have been blocked. The thirty-year exemption did not exist until seven weeks before it was granted — the Community Reinvestment Area’s own ceiling was fifteen years until September 2025. One of the six votes was not unanimous: the agreement splitting revenue with the school boards passed 5–2, which is invisible in the instrument because a Sidney resolution page never records its own vote. And the City’s Planning Commission declined, 0–4, to recommend zoning this land — and Council zoned it six weeks later anyway.
Every figure here is read from a cited record and tagged: [verified] when a document states it,
[inference] when it is derived and the assumption is shown, [reference] for secondary context,
[open] when the record does not say. Where a route was blocked — the county recorder’s search, the
Power Siting Board’s docket — that is written down as a fact about a server, re-tested, and dated.
Nothing is inferred from an access failure.
This study said no member of the public could be named. It was wrong by ten months. That claim came from the Council minutes of 2026-04-27, which report only that “a number of people present raised questions and concerns” — and it was generalised into a statement about the whole record. Sidney’s Planning Commission clerk names every speaker with a street address, and did so on 2025-06-16 for nineteen residents objecting to this land. Twelve more are named at Council in February 2026. The correction is left visible in governance rather than quietly repaired, because the mistake is instructive: a silence in one document is a fact about that document, and this record contains two clerks with two different practices.
Evidence tags
- [verified]
- Read from a cited record or a live source — confirmed and reproducible.
- [inference]
- A labelled reading or derivation from cited inputs — reasoning, not a record.
- [reference]
- An outside published specification or dataset — authoritative, but not a record about this site.
- [open]
- Withheld, unresolved, or not yet checked — a question, not a verdict.
Every figure in this study wears one of these tags. A chapter's verdict — On the record · Partial · Gap · Watch — measures the record, never the project's press.
The project
What this study is, how it reads evidence, and the project on the record.
- 01 Scope & method What a real impact study is, who normally commissions one, and how this one reads evidence. On the record
- 02 The project on the record Operator, lifecycle, load, cooling, and footprint — every undisclosed field a named blank. the IT load on the record is a screening bracket, not an instrument-grounded figure Partial
- 03 Land assembly & the deal How the land was put together, by whom, and what the paper trail says about when. the record carries no conveyance instrument — the land history has not been produced for this site Partial
The environment
Water, heat, groundwater, runoff, and air — each screened where the record allows.
- 04 Water supply & consumptive use How much water the project consumes, against the receiving water's own cited low flows. cooling method undisclosed — but the water quantity is CONTRACTED, so the figures are a stated account rather than a bracketed range; what stays open is how the heat is rejected, not how much water it takes Partial
- 05 The discharge & the receiving water Dilution at the design low flows, and the burden the receiving water already carries. On the record
- 06 Heat The discharge's heat against the reach's temperature criterion at design low flow. a thermal characterization for this project — none exists on the record. Gap
- 07 Groundwater & dewatering The wells within reach of the site's pumping, and where the pumped water went. On the record
- 08 Stormwater & the built surface The runoff a campus-scale impervious surface adds to the design storm. On the record
- 09 Air The dispersion footprint of the campus's permitted generator fleet. a DRAFT permit or public notice — an air-permit application is on the record here, but the document that names the generator fleet has not issued. Gap
The economy
The labor baseline, the grid and its ratepayers, the fiscal trade, and the balance.
- 10 Labor & the local baseline What this economy is before the project — the denominator every jobs claim divides by. On the record
- 11 Power & ratepayers Whose grid this load lands on, and what it does to the households on the same wires. the campus load is a screening bracket — the demand-pressure read is withheld until an instrument grounds it Partial
- 12 The fiscal trade What tax revenue is being traded away, for how long, and what the paper trail says. the incentive instruments are on the record, but the accounting is not — no auditor's abatement report and no exemption ledger says what the trade cost Partial
- 13 Governance & the local record The ordinances, resolutions, votes, and minutes that approved — or refused — the project. On the record
- 14 The public balance What the community gives against what it is documented to get — a verdict over the chapters. synthesizes the screened chapters — 4 on the record, 2 partial, 2 gap Partial
What's missing
Every gap this study names, as one inventory — each with the record that would close it.