Watermark.
Lima BOSC
Switch site 38 sites · 11 basins 8 open

Each site is one Watermark investigation. The page color says what opens; the campus line is the build on the ground — two different clocks.

MAU Maumee Basin 9 sites
2MI The Two Miamis 9 sites
SE Southeastern Basins 12 sites
NE Northeast Basins 8 sites
The page — what opens on Watermark Open · Draft→full record Queued · Watching→preview page Restrictedrequest access
The campus — the build on the ground Investigating Confirmed Under construction Operational
All 38 sites — the network index →
Building
The record · how to read it

How to read any record

Every record on this site — a contractor's cost estimate, an air permit, a water-discharge screen — is read the same five ways. Learn the grammar once and you can check any figure on the site yourself. Nothing below is summarized for you: each is a real record, read from the published bundle, with its evidence class on every line.

The five beats

One transformation — source → structured read → meaning → verify → connect — applied to whatever the record happens to be:

① The source — the artifact and its provenance, messiness left visible
② What we read — the structured extraction, each value's confidence shown
③ What it reveals — plain meaning, with the one load-bearing number called out
④ How to check it — the evidence tag and a link to verify it at the source
⑤ Where it connects — doors into the rest of the library

The evidence classes

Beat ④ hangs on one vocabulary. Every figure carries its evidence class — read it as a confidence signal, not decoration:

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.

One grammar, three record types

Switch between three genuinely different records below. The artifact changes; the five beats don't. Each is a real extraction — the ● in the published bundle badge marks the ones whose figures are read live from the same row the library renders.

Record type
Record Teardown

Opinion of Probable Cost

Tetra Tech · BOSC Roadwork · filed 2025-07-11
● in the published bundle
Scanned PDF · degraded OCR
① The source
PRR-01-bundle.pdf
SCAN
Tetra Tech Opinion of Probable Project Cost — summary sheet: six corridor roundabout line items with a CONSTRUCTION TOTAL of $14,223,081.
pp. 317–328 · Public Records Request 01 · summary sheet · sheet 1 of 11
The summary sheet is shown as a committed crop of the source scan (regenerated by scripts/render_walk_crops.py); the full 11-sheet exhibit is available on request.
② What we read from it
Program total
$14,223,081
Sub-estimates
6 corridors
Contingency
25%
Largest — Cole St
$3,899,800
Drainage line
$1,068,530 · 7.5%
Design-storm basis
~ not cited
Detention shown
~ none
● figures read live from the published record · ~ markers preserved as approximate
③ What it reveals

The public's road package is priced to the dollar — $14,223,081 — yet drainage is just 7.5% of it, the only one of six items detailed, with no design-storm basis and no detention shown. A number that precise on a scope that thin is the tell.

Record Teardown

Air Permit-to-Install P0138965

Ohio EPA · Division of Air Pollution Control · FINAL, issued 2026-05-28
● in the published bundle
Issued permit · text-native
① The source

Your browser can’t display this PDF inline.

Open the in-browser reader

Rendering 4132514.pdf…

Download · Open in a new tab

issued PTI + 64-item Response to Comments · Ohio EPA · Air permits (eDoc 4132514)
The 66-page issued eDocument is committed as provenance; engine make/model/size is redacted in the source itself as trade secret.
② What we read from it
Emissions units
115 · P001–P115
Data-hall gensets
114 · 3 groups of 38
Cooling towers
36 · 3 groups of 12
NOx cap
235.62 tpy
CO cap
96.06 tpy
Per-engine power (ekW)
~ redacted · CBI
Backup total
~313 MW · draft only
● figures read live from the published record · ~ markers preserved as approximate
③ What it reveals

The permit fixes the plant's shape exactly — three matched groups of generators and cooling towers, 114 data-hall gensets — yet locks the one number that sets its true scale, per-engine power, as a trade secret. The ~313 MW everyone cites is the draft public-notice figure; the issued permit will not confirm it.

④ How to check it
[open]per-engine ekW · trade secret
ekW · REDACTED · CBI Why won't the permit say how big it is?
the lock

Engine make / model / size — and so the per-engine ekW that sets the plant's true electrical scale — is claimed as a trade secret and withheld from the issued permit.

The permit fixes the plant's shape exactly — 114 data-hall gensets and 36 cooling towers in three matched groups — yet locks the one number that would let you compute its real backup capacity. The ~313 MW everyone cites is the draft public-notice figure; the issued permit will not confirm it. The scale is set; the proof is withheld.

Record Teardown

NPDES fact sheet — the 7Q10 screen

Ohio EPA · American II WWTP · permit 2PH00006 (Dug Run)
● in the published bundle
Permit fact sheet · text-native
① The source

Your browser can’t display this PDF inline.

Open the in-browser reader

Rendering oepa-2PH00006-american-ii-fact-sheet.pdf…

Download · Open in a new tab

Stream Flows table · Ohio EPA · NPDES fact sheets
7Q10 values are read from the Ohio EPA fact sheets committed in the corpus; the fact-sheet PDFs are available on request.
② What we read from it
Receiving water
Dug Run · impaired
7Q10 design low flow
0.78 cfs
1Q10 (driest week)
0.6 cfs
Summer 30Q10
0.96 cfs
Stated acute dilution
1.3 : 1
Ottawa mainstem 7Q10
0.2 cfs
Ottawa 1Q10 (driest week)
0 cfs · nearly dry
● figures read live from the published record · ~ markers preserved as approximate
③ What it reveals

Ohio EPA sizes every discharge against the stream's design low flow — and the Ottawa this project discharges into runs at just 0.2 cfs, dropping to zero in the driest weeks. The tributaries are worse: American II's own fact sheet states a dilution of barely 1.3 to 1. The receiving water is near-undiluted before this project adds a drop.

④ How to check it
[verified]fact-sheet 7Q10

A dollar total, a redacted emissions table, a low-flow water screen — three filings, one reading. Open the full record library to read any of them the same way, or each teardown's verify link to land on the exact source.